WIRELESS CHANNEL EQUALISATION

by Desmond P. Taylor1, Giorgio M. Vitetta2, Brian D. Hart3 and Aarne Mmmel4 1. Electrical and Electronic Eng., University of Canterbury, Christchurch, New Zealand. 2. Dept. of Information Eng., University of Pisa, 56126 Pisa, Italy. 3. Research School of Information Sciences and Engineering, Australian National University, Canberra ACT, Australia. 4. VTT Electronics, Oulu, Finland (on leave at the University of Canterbury, Electrical and Electronic Eng. Dept., Christchurch, New Zealand).

Abstract Equalisation techniques for wireless channels, in particular for those encountered in mobile wireless communications, are examined. Equalisation is broadly defined to include reception techniques which estimate the state or response of the channel and then attempt to compensate for its effects. The paper considers equalisation techniques for fading dispersive channels which include both time and frequency selectivity. In addition, brief consideration is given to the problems of blind equalisation, techniques for dealing with fast fading channels and to the problem of joint equalisation and decoding. The paper does not attempt to provide in-depth analysis or performance results. Rather, the interested reader is referred to the extensive list of references.

1. Introduction
In this paper, we review channel compensation or equalisation for wireless channels. Although the history of equalisation goes back many years to the early work of Lucky and others summarised in [1], their work was aimed almost entirely at the telephone channel which may be characterised as an essentially linear time-invariant, intersymbol interference (ISI) channel. Later work examined the line of sight microwave channel which may be considered as a very slowly time-varying channel to which most of the early theory on equalisation could be directly applied albeit at much higher transmission rates. This work is discussed in detail in [2] and its references. Finally, other work, [e.g., 2, 19, 49, 133] considered the ionospheric and tropospheric channels, both of which are time-varying wireless channels that have had a significant influence on the development of equalisers for the mobile wireless channel. The mobile digital wireless channel presents some different challenges [42], due mainly to the fact that the transmitter and receiver are mobile with respect to each other. When coupled with multipath propagation, fading results. The channel impulse response may then have an appreciable rate of variation ranging from slow to fast with respect to the signalling rate [49]. Multipath propagation is due to reflections and scattering and may cause frequency-selective fading and hence ISI. For low transmission rates, there is often significant time variability but little frequency selectivity. At higher rates, the channel is typically frequency selective but usually varies significantly more slowly with respect to the transmission rate. In the first instance, the main effect is a timevarying attenuation which affects all frequency components equally; this is known as flat fading or time-selectivity. In the second case, the channel response varies with frequency across the bandwidth of the transmitted digitally modulated signal and causes ISI between adjacent symbols. In all cases the channel may be considered to be linear. Equalisation in general consists of estimating the response or state of the channel and using the estimate to compensate the channel effects so as to improve transmission system performance. Usually, equalisation is carried out at the receiver is based only on observation of the received signal. In the time-selective case, equalisation consists of estimating the time-varying attenuation and phase of the channel and using the estimate to compensate their effects. In the frequency-selective case, it consists of estimating the response of the channel and then using this information to adjust the parameters of some form of filter to compensate for the frequency-selective effects. The filter may be linear (e.g. the transversal equaliser [2]) or nonlinear (e.g. a decision feedback equaliser [2, 173] or a maximum likelihood sequence estimator [105]). Equalisers may be either per-symbol or sequence based and are usually based on optimal receiver theory [32]. The basic theory of optimal reception over randomly time varying dispersive channels is to a large extent captured in the work of Kailath [15-18].

He developed basic theoretical structures for per-symbol receivers. The work is readily extendible to sequence based receivers and many modern receivers are based on the maximum likelihood sequence estimation approach first enunciated by Forney [105] in the context of channel equalisation. The optimal receiver structures are based on the theory of statistical signal processing which is summarised in [72, 334]. Many digital wireless systems utilise forward error correction coding techniques in order to obtain adequate error performance [177]. There is currently considerable interest in the use of trellis or signal space encoded signalling in order to maintain spectral efficiency. The receiver must then perform both equalisation and decoding. The paper briefly considers the joint equalisation and decoding process. The paper is organised as follows. In Section 2, we discuss modelling of the wireless channel. We also consider some general aspects of receiver modelling and classifications of equalisers. Section 3 focuses on equalisation or compensation of timeselective or flat fading channels. Section 4 considers the equalisation of frequencyselective fading channels. Section 5 briefly considers the problem of joint equalisation and decoding and finally Section 6 provides some conclusions.

2. The Communications System

Here we develop a mathematical description of the physical communication system. We consider the transmitter, channel and receiver, with particular emphasis on characterising the channel. A diagram of the basic system is shown in Fig. 1. 2.1 The Transmitter We consider two classes of transmitted signals, s(t ) : (a) linearly modulated signals [19]; and (b) Continuous Phase Modulated (CPM) signals [20]. Linearly modulated signals may be expressed in complex baseband form as s(t ) = ck p(t kT )
k =1 N

(2.1)

where ck is the k-th transmitted symbol; c N = [c1 , c2 , , c N ] is the symbol sequence of length N; T is the signalling interval or symbol period; and p(t ) is the transmitter impulse response or pulse shape. The symbols, ck , are taken from an M-ary complex constellation, where M is normally a power of two. Possible constellations include Mary amplitude shift keying (ASK), where ck {0,1,2, ,( M 1)} , M-ary phase shift 2 2 ( M 1) keying (PSK), where ck 1,exp( j ), ,exp( j ) , and M-ary quadrature M M

amplitude modulation (QAM). For square QAM constellations, M is a power of 4 and the symbols have the form ck 1,3, , M 1 + j 1,3, , M 1 . CPM signals may be expressed in complex baseband form on 0 t NT as [20] s( t ) = t 2Es exp j ( , c N )d , T 0

{{

} {

}}

(2.2)

alphabet, { 1,3, , ( M 1)} . We primarily consider linearly modulated signals, but the most of the concepts apply to CPM signals. We note that while we are primarily concerned with equalisation, some modulation methods are robust to delay spread [212,213]. Also, noncoherent detection of orthogonal FSK is relatively unaffected by frequency-selectivity [200]. 2.2 Wireless Channels In wireless communications, the transmitted signal is modified by three physical mechanisms: inverse distance power loss, shadowing and multipath propagation. It is also corrupted by additive noise. Inverse distance power loss causes the received signal strength to decrease with increasing distance from the transmitter, typically according to an inverse second to fourth power law [163]. Shadowing accounts for slow bulk signal strength variations, as when the receiver is obscured from the transmitter by buildings, hills, or tunnels [117, 121]. It is typically modelled by representing the envelope of the line-of-sight component of the received signal as a random variable having a log-normal probability density function [117, 121, 163]. Both power loss and shadowing merely attenuate the received signal and have little influence on equaliser design. On the other hand, multipath propagation, due to the presence of multiple paths between transmitter and receiver, can severely distort the transmitted signal. Moreover, it is usually time-varying and causes fading. The distortion due to multipath changes appreciably over one wavelength, a distance that is at least two orders of magnitude smaller than the distance over which either inverse power loss or shadowing effects change significantly [132, 210]. High performance equalisers can be designed only if adequate models of the channel are provided [19, 134] to represent the distortion of the signal due to multipath propagation. Given a channel model an equalisation strategy can be developed. Different channel models lead to significantly different equaliser structures. In the following subsections, we consider multipath channel characterisation and modelling.

where E s is the transmitted energy per symbol interval, (t , c N ) is the information bearing phase and the k-th transmitted symbol, ck , belongs to the M-ary real

2.2.1 Multipath Channels For simplicity, we assume a stationary transmitter (base station) and a mobile receiver. The transmitted signal is reflected and diffracted by scatterers, such as hills, buildings, trees and vehicles. Some of the signal reaches the receivers antenna, usually via several paths. The paths exhibit differing attenuations and have different lengths, so that the receiver observes several relatively delayed and attenuated versions of the signal. Each path delay may be conceptually divided into two parts: the so-called cluster delay [3], which is on the order of a symbol interval, and the fine delay, which is on the order of the carrier period. The former depends on the relative positions of the large scale scatterers and is preserved in the channel model. The latter can be modelled as a random variable, affecting only the carrier phase. Hence, path attenuation and fine delay are lumped together as a complex gain. Due to change in fine delay, this gain changes markedly over a carrier wavelength (0.3m at 1GHz). The superposition of the arriving paths at any value of delay induces destructive and constructive interference, varying according to position. As an antenna moves through this interference pattern, its spatial variation appears as a time-variation in the received signal. In addition, due to the motion of the antenna, the signal on each path undergoes a Doppler shift that depends on the path arrival angle but does not exceed some maximum, f D . We call f D the onesided Doppler spread. It equals the maximum relative speed of the transmitter, channel scatterers and receiver divided by the carrier wavelength. Thus the received signal is the sum of many Doppler shifted, scaled and delayed versions of the transmitted signal [19, 42, 49, 152]. Due to the linearity of the channel, the received signal can be modelled as [134] r ( t ) = z ( t ) + n( t ) where z(t ) =

(2.3)

s ( t ) a ( t , ) d .

(2.4)

Here z (t ) is the noiseless received signal, a (t , ) is the instantaneous time-varying channel impulse response and n(t ) is additive white Gaussian noise. Physically, the channel can be visualised as a densely tapped delay line, with delay index , so a(t , ) represents the time-varying tap gain at delay . In terms of the various paths, a (t , ) is the sum of the complex gains of all paths with delay , measured at the current location of the receivers antenna. Since the antenna is moving, the sum is time-varying. A linear time varying channel may also be characterised by the Fourier transform H ( f , t ) of a(t , ) with respect to the delay variable . This function is known as the time-variant transfer function and allows the use of frequency domain techniques [134].

2.2.2 Characterising Multipath Fading Channels The system functions a(t , ) and H ( f , t ) describe any time-varying channel. In the multipath fading case, they represent realisations of a stochastic process since, in practice, the receiver has no knowledge of the instantaneous scatterer geometry. Thus, statistical characterisation is necessary [134]. If we model the channel as clusters of many independent scatterers [3], the Central Limit Theorem applies and we can assume Gaussian statistics [49]. The channel is then characterised by the mean and correlation functions of one of the time varying system functions. When the complex gains due to the different scatterers have similar amplitudes, the functions a (t , ) and H ( f , t ) have zero mean and their envelope obeys a Rayleigh distribution. When there is a dominant path (e.g. a line-of-sight path), or more generally, a dominant path per delay value, they have a non-zero mean and their envelopes have a Rician distribution [19]. These models are justifiable mathematically, but other distributions, such as the Nakagami-m distribution [154, 201], fit some experimental results more closely. For equaliser design, the Rician channel model is usually sufficiently general. Many designs consider Rayleigh fading only, since an equaliser designed for Rayleigh fading generally performs better in a Rician fading channel. As a first order statistical description of the fading, we can decompose a (t , ) into a specular and diffuse component, corresponding to the dominant paths and the remainder. The specular component, defined by a s (t , ) E {a (t , )} , (2.5)

is known as the channel mean. It may include a Doppler shift. The diffuse component is given by a d (t , ) a (t , ) E {a (t , )} and is Rayleigh faded. A sufficient second order statistical description of the channel process a (t , ) , under the assumption of Gaussianity, is given by the correlation function [19, 134, 186] Raa (t 1 , 1 , t 2 , 2 ) = E {a (t1 , 1 )a * (t 2 , 2 )} (2.7) (2.6)

This is known as the tap gain cross-correlation function, as it represents the crosscorrelation between scatterers at different delays. If the channel can be modelled as a collection of Wide Sense Stationary (WSS) scatterers, it simplifies to Raa ( , 1 , 2 ) E {a (t + , 1 )a * (t , 2 )} (2.8)

In addition if the WSS scatterers are assumed to be uncorrelated (uncorrelated scattering or US) the channel is said to be WSSUS and Raa ( , 1 , 2 ) can be rewritten as

Raa ( , 1 , 2 ) = Paa ( 1 , ) ( 1 2 ) ,

(2.9)

where Paa ( , ) is called the tap gain correlation function. Of physical importance is the function Paa ( ,0) which is proportional to the average power received from scatterers at delay . Knowledge of Paa ( ,0) allows evaluation of the multipath delay spread d which is the interval in over which Paa ( ,0) is effectively non-zero. Another function commonly used to characterise a fading dispersive channel is the scattering function S aa ( , ) which is defined as the Fourier transform of the tap gain correlation function Paa ( , ) with respect to the correlation lag , that is S aa ( , ) Paa ( , ) exp( j 2 )d

(2.10)

It is meaningful only for WSSUS channels, and is proportional to the power scattered by the medium at delays ( , + d ) in the Doppler shift interval ( , + d ) . The width of S aa ( , ) in is the multipath spread, and its width in is the two-sided Doppler bandwidth B D (or Doppler spread) and is equal to 2 f D . Finally, another function used to characterise a WSSUS channel is the time frequency correlation function Qaa (, ) which is defined as the Fourier transform of the tap gain correlation function with respect to the delay variable , as Qaa (, ) Paa ( , ) exp( j 2)d

(2.11)

It represents the cross-correlation between received frequencies spaced by Hz. Two other parameters, often used to provide a description of the time- and frequency-selective properties of the channel, are the coherence time c and the coherence bandwidth Bc . The first represents the interval over which the received signal can be considered coherent and is roughly equal to the inverse of the Doppler spread. The second represents the frequency band over which the multipath fading can be considered frequency-flat and is approximately the inverse of the multipath spread. 2.2.3 Properties of the Channel Channel delay spread arises from the variations in path length and can produce deep notches in the time varying frequency response H ( f , t ) . Since the signal

bandwidth is usually on the order1 of 1/T, the normalised delay spread, d/T, is a measure channel frequency-selectivity. When d/T << 1, the channel is highly correlated across the signal bandwidth - i.e. frequency-nonselective or frequency-flat. Any fading is then almost uniform across the signal bandwidth. For larger values of d/T, the channel correlation is reduced. A fade at one frequency does not then imply fades at more distant frequencies, so that spectral notches arise, and equalisation is required [49]. Time-variation or Doppler spread arises from relative motion of the transmitter, scatterers and receiver. To illustrate its effects, consider the amplitude and phase of a single time-varying channel tap, a d (t , 0 ) and a d (t , 0 ) . At any values of delay and time a d (t 0 , 0 ) is a Rayleigh random variable and is commonly faded by 20dB relative to the mean signal strength2 and reliable communication is difficult [42]. a d (t 0 , 0 ) changes smoothly in time, exhibiting quasi-periodic fades, where the amplitude drops rapidly but relatively briefly [42]. The uniformly distributed phase also changes smoothly, except during deep fades, where it may change by 180 in a fraction of a symbol interval. A measure of time-selectivity is the normalised Doppler bandwidth B D T . When BDT << 1, the channel is highly correlated in time (essentially time-nonselective or time-invariant). The channels time variation can then be largely ignored in equaliser design as long as it has some tracking ability, since a fade lasts considerably longer than the duration of a transmitted pulse. However, for larger values of BDT, as in some TDMA radio applications [194], an equaliser specifically adapted to fast fading channels is required. In considering the time-selective behaviour of a dispersive channel, we note that the tap gain correlation function may often be written in product form [e.g., 198] as Paa ( , ) = Pa ( ) Ra ( ) (2.12)

where Pa ( ) is the delay power density profile (same physical meaning as Paa ( ,0) ) and Ra ( ) is the normalised autocorrelation function ( Ra (0) = 1 ). The Fourier transform S a ( f ) of Ra ( ) is the Doppler spectrum of the channel fading process. A common mathematical model for it is Ra ( ) = J 0 (2f d ) , (2.13)

( J 0 ( x ) is the zero-order Bessel function) which results from the assumption of isotropic scattering [152]. The corresponding fading power spectrum is

We do not consider spread spectrum systems, where the bandwidth is >> 1 / T ; however, many of the concepts considered here may be directly extended. Deeper fades occur but more rarely.

1 f2 Sa ( f ) = 1 2 fD 0

f < fD . f fD (2.14)

This expression is best regarded as the average of all possible spectra. It is usually a poor model for any particular channel. Other models are often more useful in assessing system performance and some are found in [33]. Finally, we note that when Doppler spread arises due to terminal motion, the Doppler bandwidth is the same for all values of multipath delay. 2.2.4 Classifying Channels Wireless channels may be classified by the spread factor, BDd [19] which plays an essential role in defining the measurability of a channel [183-186]. When BDd < 1, the channel is called underspread [125], and its impulse response may be estimated (e.g. from pilot tones or pilot symbols), although the difficulty increases as BDd nears unity. When BDd > 1, the channel is overspread, and cannot be estimated3. A channel may also loosely be regarded as belonging to one of four channel classes, according to the values of d/T and BDT. In the first, the channel is essentially both time- and frequency-nonselective (d/T << 1, BDT << 1). The received signal is scaled by a complex gain, a so equalisation consists merely of estimating the channel gain and phase, where a (t , ) = a (t ) . The second class comprises essentially timeinvariant, (BDT << 1), frequency-selective channels. The channel response varies with frequency across the bandwidth of the transmitted modulated signal, but it changes very slowly compared to the symbol rate. Its impulse response may be regarded as that of a linear, time-invariant filter, a (t , ) = a (0, ) , which causes quasi-constant ISI between adjacent symbols. Equalisers have been historically developed for such channels [2]. In the third class, the channel is time-selective and essentially frequency-nonselective (d/T << 1). Its main effect is a time-varying complex attenuation, a (t ) , which affects all frequency components equally and a (t , ) = a (t ) ( ) . The fourth class comprises channels that are time- and frequency-selective (known as doubly selective channels). The response varies appreciably across the signal band, and is time varying. This class is the most general with the other three being special cases. We will be principally concerned with the third and fourth classes. The values of d / T , BD T and BD d significantly affect the transmitted signals and equalisers that can be employed.

This criterion was introduced by Kailath [183]. A more accurate criterion replaces the spread factor with the area under the Doppler delay spread function [185].

2.2.5 The Tapped Delay Line Channel Model The continuous time channel representation contains more information than is required in a receiver, since the transmitted signal is bandlimited, and the channel response outside the signal bandwidth is irrelevant. Moreover, a discretised or sampled representation of the channel is more amenable to digital implementations. Therefore, we consider a discretised, version of (2.3). The transmitted signal can be represented as a weighted sum of its samples, sk s( kTs ) , as t kTs , s(t ) = sk sinc k Ts

(2.15)

where 1/Ts is chosen to be at least the Nyquist rate for s (t ) . Substituting this into (2.4), we obtain
nT z n = sk sinc r k a (nTr , )d = sn k a n ,k , k k Ts

(2.16)

where z n z (nTs ) , the channel tap gains a n ,k are given by

nTr k a(nTr , )d a n ,k = sinc Ts

(2.17)

and the sampling rate 1/Tr is chosen to be faster than the Nyquist rate for the noiseless received signal. Since the pulse shapes of interest have bandwidth of at least 1/2T, it is incorrect4 to employ symbol-spaced taps as, noted in [142], when matched filtering is not employed. There are an infinite number of non-zero channel taps in general, due to the infinite duration of the sinc(.) functions in (2.16). The taps at large nTr / Ts k diminish, so the infinite summation in (2.16) can normally be truncated without appreciable error. It is simplest to use a common sampling period, Tr throughout. In addition, Tr, is normally chosen for convenience such that T is an integral multiple of Tr: i.e. T = rTr, where usually r = 2 is sufficient. Thus we can define the vectors rrN [r1 , r2 , , rrN ] , z rN [ z1 , z 2 , , z rN ] , s rN [ s1 , s2 , , srN ] and n rN [n1 , n2 , , nrN ] . The received signal actually extends beyond t = [0, NT ] due to the tails of the transmitted pulse shape and the delay spread, but we can ignore these edge effects for N large.

An equaliser with symbol-spaced delay taps suffers a performance penalty relative to one designed assuming with its taps at the Nyquist spacing.

10

2.2.6 The Power Series Expansion Channel Model The delay line model expands the channel impulse response as the weighted sum of time-shifted sinc(.) functions as a function of delay, . Other choices of basis function are available [134]. If the channel can be represented accurately using fewer basis functions, fewer parameters need to be estimated, potentially leading to simpler equalisers. The optimal set of basis functions is obtained from the Karhunen-Love expansion of the channel autocovariance [125]. However, this must be known or estimated, to obtain the basis functions. Thus, there is more interest in choosing basis functions known a priori to be good. A Taylors series expansion was proposed in [134], so that the basis functions are polynomials. This is most appropriate for smoothly changing functions, such as the variation of a (t , ) or A(t , f ) in t or f and has recently resulted in a Reduced Dimensionality Model [151,164,166] for doubly selective channels and a linearly time-selective distortion model for rapidly varying flat fading channels [48,100,129,130]. 2.3 The Receiver The receiver must detect the information sequence while compensating the channel distortions. At its core is the equaliser, which we discuss more thoroughly in subsequent sections. Here, we briefly discuss some common receiver properties.

2.3.1 The Receiver Front End An optimal receiver obtains a set of sufficient statistics [32] for recovering the transmitted symbol sequence. When the channel is known, the output of a filter matched to the received pulse can be sampled at the symbol rate to provide a set of sufficient statistics [105,138], as mk =

r (t ) p

(t kT )a * (t , )ddt

(2.18)

For this case, the channel may be modelled as a delay line with symbol-spaced taps. However, this receiver usually cannot be implemented because the time-varying channel is unknown a priori. Hence, a set of sufficient statistics is only easily obtained by sampling the received signal at a rate, 1/Tr, that is at least its Nyquist rate. A low pass filter to limit the noise bandwidth is needed before sampling. There are other front-ends, which can make acceptable trade-offs between performance and complexity. For slowly time-varying channels, the noise-limiting filter is often chosen to be a filter matched to the transmitted pulse. Although near optimum for the frequency non-selective channel [142], it is sub-optimum for the time-varying

11

and/or frequency selective case. The received signal is Doppler-spread by the channel, so such a filter has insufficient bandwidth. Another simplification is to sample the filtered received signal below its Nyquist rate, typically at one sample per symbol. This entails a power penalty, and in fast fading channels, leads to an error floor [71,160]. 2.3.2 Classifying Equalisers Many different equaliser structures have been investigated; however, most are based on similar ideas or share similar properties. Hence, it is instructive to identify equaliser groupings: One grouping is according to the class of channel for which the equaliser is designed. This grouping is hierarchical, in that equalisers for the general time- and frequency-selective channel have simplified counterparts in the time-invariant and frequency-flat classes of channels. A second grouping is according to the statistical basis of their decision rules. In particular, in designing an equaliser according to some optimality criterion [32], additive noise may be considered or ignored and the multipath channel may be treated either as known or as a stochastic process.

We consider two examples in the second group in order to provide a first insight into the practical problems encountered in optimal design. For the first, the channel is assumed known and Gaussian noise is included. Then the probability density function of the vector rrN of received samples, conditioned on the symbol sequence, ~ c N , is given by [72] p(rrN | ~ cN ) = cN ) N C r (~ 1
H 1 ~ exp (rrN rN ( ~ c N )) Cr c N )) . (cN ) (rrN rN ( ~

(2.19) where indicates the determinant, and where rN (~ c N ) E {rrN | c N = ~ c N } and

H C r (~ c N ) E (rrN rN ( ~ c N )) (rrN rN ( ~ c N ))| c N = ~ c N are the mean vector and covariance matrix of r conditioned on c = ~ c . Since the channel is known, the rN N N

expected value of the received signal is its noiseless version, corresponding to cN = ~ c N . Hence, rN (~ c N ) = z rN (~ c N ) and the covariance matrix is the noise
H autocovariance, C r = E {n rN n rN }, so the determinant in (2.19) is independent of the hypothesised symbol sequence. Assuming that n rN to be white noise, the loglikelihood function of the received vector rrN conditioned on c N = ~ c N is found from

(2.19) as

12

2 c N ) = rk z k (~ cN ) (rrN | ~ rN k =1

(2.20)

 N [32] of the transmitted symbol sequence is The Maximum Likelihood (ML) estimate c then arg max (rrN | ~ N . cN ) c
~ cN

(2.21)

In treating the channel as known, it is realised that in practice it must be estimated a priori and the estimate used in computing the sequence metric (2.20). This class includes adaptive MLSE receiver structures [135,142,144,182] and the differential detectors used in time-varying, frequency-flat channels. In the latter case (at least conceptually), the received signal during one symbol interval is divided by the detected symbol and the result used as a channel estimate for the next symbol interval [19]. In the second example, both the Rayleigh fading channel and noise are treated as stochastic processes. The probability of observing the vector rrN conditioned on ~ c N , is again given by (2.19). However, the channel is unknown and purely random, so the expected value of the received signal vector is the null vector, rN (~ c N ) = 0 rN and the received signal autocovariance C r (~ c N ) depends on ~ c N . Then the log-likelihood of the ~ received signal vector r conditioned on c = c is from (2.19),
rN N N H 1 ~ c N ) = ln C r (~ c N ) + rrN (rrN | ~ Cr ( cN ) rrN

(2.22)

It consists of a bias term and a quadratic form. The ML detection strategy is again given c N ) given by (2.22). However, evaluation of the metric of by eq. (2.21) with (rrN | ~ (2.22) is much more complex than that of (2.20). This metric characterises the second type of equaliser which is most suited to fast randomly time-varying channels. Finally we note that Rician channels have both specular and diffuse components. If we assume the specular component to be known, we may conclude that the optimal receivers should combine the equalisers from both categories above.

3. Frequency Non-Selective Channel Equalisation

Here we consider equalisation for the time-selective or frequency-flat fading channel. As noted in Section 2, a fading channel is frequency-flat when its delay spread is so small that the multipath effect results in a complex time-varying multiplicative distortion.

13

The first relevant studies concern optimal diversity detection of digital signals on time-selective fading channels and date back more than forty years [3-18]. In particular, a clear understanding of the general problem of Maximum A Posteriori (MAP) detection of digital signals transmitted through a Gaussian random channel was provided by Kailath [15-18]. He found the MAP receiver for an M -ary set of digital signals {qi (t ), i = 1,2,, M } transmitted on a purely random channel to consist of M branches, each an estimator-correlator structure5. The optimal receiver operates by making in its ith ( i = 1, , M ) branch a Minimum Mean Square Error (MMSE) estimate of the faded useful signal. This is provided by a time-varying, unrealisable filter [32] designed under the assumption that qi (t ) is the transmitted signal. A decision metric is derived by

correlating this estimate with the received signal and adding a bias term6 to the correlation result. If the l-th branch output produces the largest metric, the receiver decides that the signal q l (t ) has been transmitted. The estimator-correlator is illustrated in Fig. 2 for a binary communication system. If the channel is not purely random and its mean response is known a priori, the metric computed in the l-th branch of the receiver is the sum of two terms: one evaluated by an estimator-correlator processing only the random component of the received signal and the other evaluated by a correlator extracting information from the deterministic component [15]. We also note the following: a) Implementation of the estimator-correlator detector, employing realisable timevarying filters is discussed in [32] and [41] for discrete and continuous time signals (see also [21-24]), respectively; Since the optimal receiver computes M MMSE estimates of the fading distortion, one for each hypothesis, we can interpret the MAP receiver as either: (1) a form of equaliser because the time-varying channel gain is estimated and its effect is compensated by correlating the fading estimate with the received signal; or (2) a form of partially coherent detection since the receiver, in correlating the channel estimate with the received signal, tries to compensate for the channel phase effects [25,26]. Coherent detection is achieved in the limit of perfect estimation of the fading process; MMSE estimation of the fading requires the mean and autocorrelation function of the random channel. In general, these are not known and must be estimated. The receiver can employed for both one-shot and sequence detection [32]. However, receiver complexity increases exponentially with sequence length and leads to unacceptable complexity in practical applications.

b)

c) d)

This interpretation of the optimal receiver was first proposed by Price for a fading channel with a single scatter path [4]. Kailath extended Price's ideas to an arbitrary random channel.
6

In general the bias term depends on the fading and noise second order statistics and on the hypothized signalling waveform.

14

Following the above, considerable research has been directed to the analysis of simple noncoherent one-shot receiver structures (see, for instance, [28-31]) because of their practical importance. Numerous solutions of this type have been proposed for the detection of digital signals transmitted through frequency flat fading channels. They can be roughly categorised as: a) b) c) Noncoherent detectors; Coherent detectors employing pilot tones or symbols as a phase reference. Sequence detectors.

In the remainder of this section these will be examined. A final subsection is devoted to receiver structures for fast fading channels. 3.1 One-Shot Noncoherent Detectors Channel estimation is difficult in fading channels [42]. Noncoherent receiver structures allow detection of a signal in the absence of an explicit channel estimate and offer the advantage of simplicity. Strictly speaking these are not equalisers since they make no attempt to estimate the channel. However, they are widely used in wireless transmission and their analysis provides a basis for approaching that of more complex equalisation structures. If the linearly modulated signals of (2.1) are considered, a differential receiver can be employed, provided that differential encoding is accomplished at the transmitter [19]. An analysis of the error performance of differential PSK receivers on fading channels can be found in [14, 29] and [33-37]. Noncoherent detectors are also available for the CPM signals of (2.2) and comprise differential detectors, discriminators and matched filter & envelope detectors [4]. An analysis of their error performance is provided in [38-40] and [46, 94] for the differential receivers, in [43-47] for the discriminators and in [3, 7, 13, 14, 48] for the matched filter & envelope detectors. These detectors all suffer from two drawbacks: (1) there is a Signal-to-Noise-Ratio (SNR) loss with respect to coherent detection; (2) if the fading is fast (changes appreciably in a symbol interval), the detector error performance will exhibit an error floor [14, 95]. This is largely due to the quick phase changes which the signal experiences during deep fades [49]. 3.2 Reference Based Techniques for Coherent Detection Coherent detection is possible if a reference (or sounding) signal [50, 51] is transmitted with the information bearing signal. In practice an accurate phase reference cannot be generated by a Phase-Locked-Loop (PLL) because a PLL cannot track the rapid phase changes of the channel fading [52]. A coherent reference can be made available to the receiver by transmitting a time-continuous sounding signal (pilot tone) [53-62] or by transmitting a sequence of known symbols (pilot symbols) interspersed with the data symbols [62-67]. Several pilot tone techniques have been proposed as follows: 15

a)

b)

c)

d)

That described in [53] which consists of sending a continuous wave sounding signal together with a data BPSK signal. The data and the sounding signals can be separated since they are kept orthogonal in phase; The Transparent-Tone-in-Band (TTIB) technique [57], where the baseband spectrum is split into two segments. The segment in the upper frequency band is translated up in frequency by an amount equal to the 'notch' width and a reference pilot tone is added at the center of the resulting notch. The Tone Calibration Technique (TCT) [55] creates a spectral null in the data signal by means of a zero DC encoding technique (e.g. Manchester coding [19]) and inserts a pilot tone in the null. The TCT scheme is illustrated in Fig. 3 together with the baseband spectrum of the transmitted signal. The Dual-Pilot Tone Calibration (DPTC) Technique [54], where two pilots are symmetrically located outside the data spectrum near the band edges. DPTC provides better bandwidth efficiency than TCT at the price of increased sensitivity of the pilots to frequency shifts [54].

Pilot-tone techniques lead to robust and simple receiver structures, as evidenced by Fig. 3. The pilot tone can be separated by relatively simple circuitry from the received signal. The use of coherent detection substantially lowers the error floor level of the receiver. Their main disadvantage (and also of the pilot symbol techniques) is that a fraction of the transmitted power is wasted in transmitting reference signals. Simpler transmitter and receiver processing is achieved by pilot symbol assisted modulation (PSAM) [62-67], although frame synchronisation is required at the receiver. In PSAM transmission the transmitter periodically sends known symbols, from which the receiver derives its amplitude and phase reference. The PSAM transmitter and receiver schemes are shown in Fig. 4 together with the transmitted data format. Here, the data symbol rate is equal to ( K 1) / KT , 1 / ( KT ) being the pilot symbol rate7. Like pilot tone modulation, PSAM suppresses the error floor and offers the further advantage of enabling multilevel modulation without requiring a change of the transmitted pulse shape or of the peak to average power ratio. A comparison of PSAM with TTIB [62] has shown it offers substantially better energy efficiency for any practical power amplifier. Finally, we note that reference-based techniques for coherent detection were first proposed for linearly modulated signals. Recently, Ho et al. [68] have shown that a pilot symbol assisted detection strategy can be implemented for CPM signals.

The pilot symbol rate should be at least 2( B D ) MAX , ( BD ) MAX being the largest value of the Doppler bandwidth BD [63].

16

3.3 Sequence Detectors This subsection examines sequence detection techniques. In order to present a unified framework, a recursive formulation [70] of the MAP detection strategy [15,69] for the linearly modulated signal of (2.1) on a slow fading channel is derived. It is then shown how the different solutions can be related to the optimal strategy. In a MAP receiver for the linearly modulated signal of (2.1) transmitted over a slow fading channel [70] the received signal is passed through a matched filter. The symbol-rate samples at the filter output can be expressed as [70] y k = a k ck + wk (3.1)

for k = 1,2, , N , where a k represents the fading distortion, ck is the k-th transmitted symbol and wk is AWGN8. Note that y k is a sufficient statistic if it is assumed that a (t ) is constant (and equal to a k ) over the duration of the transmitted pulse p(t kT )  N of the symbol sequence is found (slow fading assumption). Then the MAP estimate c as c N [c1 , c2 , , c N ]  N = arg max c p( ~ cN | y N ) ~ cN (3.2)

c N | y N ) is the where y N y1 , y 2 , , y N , ~ c N is a hypothesised data sequence and p(~ ~ probability density function of c N conditioned on y N . The solution of (3.2) can be found by exhaustive search over the set of hypothesised sequences {~ c } . This entails a
N

computational burden increasing as M . The evaluation of the conditional density c N y N over the set {~ p~ c N } becomes numerically simpler if a recursive formula is

derived. To begin, we observe that [70] p( y k | y k 1 , ~ ck 1 ) ~ ~ p(~ p(ck | ck 1 ) ck | y k ) = p( y k | y k 1 ) (3.3)

~ |~ In (3.3) p(c k ck 1 ) represents a transition probability and equals 1 / M if the sequence ~ |~ is uncoded (if ~ c is coded p(c c ) depends on the code structure). The MAP strategy then becomes the ML strategy. Moreover, p( y k | y k 1 ) does not depend on the trial sequence ~ c N . Taking this and the recursive formula (3.3) into account, the MAP
k k k 1

strategy (3.2) can be rewritten as

8

A similar signal model exists for the samples of a CPM signal taken at the output of the receiver front end filter (for instance [71]). Thus the developments of this section can be easily extended to the problem of CPM detection.

17

N ~ c N = argmax ck ) , (y k , ~ ~ c N k =1

(3.4)

where ~ |~ ~ ck ) = ln p(c (y k , ~ k ck 1 ) + ln p( y k | y k 1 , ck 1 ) .

} {

(3.5)

ck 1 ) is Gaussian [72] and from The conditional probability density function p( y k | y k 1 , ~ (2.19) and [70] may be expressed as y ( k | k 1) 1 y k ~ p( y k | yk 1 , ck ) = exp 2 2 y ( k | k 1) y ( k | k 1)
2

(3.6)

where the mean y ( k | k 1) and the variance y2 ( k | k 1) are given by

y ( k | k 1) = E {y k | y k 1 , ~ ck } ,
2 ~ 2 y ( k | k 1) = E y k y ( k | k 1) | y k 1 , ck .

(3.7)

(3.8)

Moreover, the mean y ( k | k 1) can be rewritten as [70] ~a ~ y ( k | k 1) = c  k ( k | y k 1 , c k 1 ) , (3.9)

Wiener filter [72,73]. However, if the process {a k } can be characterised by a Gauss-

 ( k | y k 1 , ~ where a ck 1 ) is the MMSE one-step prediction of the fading sample a k assuming that the sequence ~ ck 1 has been transmitted. In general the evaluation of  (k | y , ~ a c ) and the variance 2 ( k | k 1) can be accomplished by a time-varying
k 1 k 1 y

Markov model [72,74], both quantities can be computed recursively by means of a Kalman predictor9 [73] for a given sequence ~ c N . M N Kalman filters are needed, one for each possible symbol sequence. In general, the complexity of the MAP (or ML) receiver is large and increases exponentially with sequence length. Some studies [71,72,75,77] have indicated that it can be simplified if one of the following two assumptions hold: A.1) The sequence {a k + wk } of fading plus noise samples is an Auto Regressive (AR) process of finite order L [71,72,77].

The recursive evaluation of likelihood functions was first described by Schweppe [76].

18

A.2) The autocovariance function of the fading process has finite support i.e. Ca ( kT ) = 0 for k > L [75]. In both cases the channel is said to have finite memory L [75,77]. However, it can be shown that only (A.1) leads to a simplified optimum structure [71], whereas a reduced complexity receiver based on (A.2) is not optimum. Under (A.1), evaluation of y ( k | k 1) and y2 ( k | k 1) requires only the fixed-length vectors ~ , , c ~ ~ k 1 , c y ( L) y , y , , y and ~ instead of the variable c ( L) c
k 1

k 1

k 2

kL

k 2

kL

length vectors y k 1 and ~ ck . This reduces the number of estimation filters from M N ~ ,c ~ , , c ~ to M L (the number of possible data vectors ~ ck 1 ( L) = c k 1 k 2 k L ), independent

of the actual sequence length. Under this assumption, the Kalman filters can be replaced by time invariant Wiener predictors [71, 75, 77, 86]. This leads to substantial complexity reduction. Under (A.2) or other similar assumptions, optimality, strictly speaking, requires infinite-length predictors and complexity reduction is not possible. Finally, we note that: a) The sequence e y ( k | k 1) y k y ( k | k 1) (see (3.6)) represents a discretetime innovations process [76,77]. Thus the optimal sequence receivers are in fact innovations-based receivers [77]; b) The optimal receiver minimising the symbol error probability is closely related to the optimal sequence estimator. In fact, the optimal symbol decision ck on the k-th symbol ck (1 k N ) , given y N , is given by [72] ck = argmax  k c ~ p c | y ( ) N N  ~ c N ( ck ) (3.10)

  ~ =c where (ck ) denotes the set of all trial sequences ~ c N such that c k k . The c N | y N ) of (3.10) are also evaluated from the sequence conditional probabilities p(~

detector (see (3.2)). c)

y ( k | k 1) in (3.7) or (3.9) is an MMSE estimate of the faded signal component sk a k ck in (3.1). Thus, an optimal sequence detector implicitly evaluates multiple channel estimates, i.e. as many channel estimates as the hypotheses on the transmitted sequence. However, an unambiguous phase reference can be computed only if some symbols are known (as in PSAM transmission) or if the transmitted sequence is coded [86].

3.3.1 Symbol-by-Symbol Receivers Both optimal sequence and symbol-by-symbol receivers are complex structures even when assumption (A.1) holds. An alternative is the class of sub-optimal decision

19

feedback receivers [25-27], [70] and [78-83]10. These are based on the idea that in order to detect the k-th symbol ck coherently, an estimate of the fading distortion sample a k is required. If the data decisions on previous symbols are reliable, they can be used to remove the modulation from the corresponding received signal samples and to predict a k . A general scheme for such a receiver is shown in Fig. 5. Decision feedback leads to a reduction in the number of channel predictors from M N (or M L ) to one. The single channel estimate can be computed by a Wiener predictor [25,26,81], by a Kalman filter [70,78,79] or by an Extended Kalman filter11 [80]. A drawback of these receivers is that a periodic refresh of their memory with a string of known symbols is required to prevent loss of channel tracking (receiver runaway) [26] and to solve the phase ambiguity problem. The error performance of a decision feedback receiver can be improved (with increased detection latency) if the two-stage architecture of [83] is used (see Fig. 5). In this case, the first stage consists of a symbol-by-symbol detector with a MMSE channel estimator. The data decisions of the first stage are delivered to the second stage which generates an improved channel estimate by means of an optimal smoother. Finally, this estimate is used to produce new (more reliable) data decisions. A similar architecture has been proposed by Kam in [118]. K-lag symbol-by-symbol MAP estimation has been investigated by Seymour and Fitz in [82] for PSK and QAM signals. They derive a suboptimal receiver by resorting to decision feedback, to a thresholding technique for discarding unlikely hypotheses on the past data decisions and to assumption (A.2) on the finite memory of the channel. A feature of their solution is that it produces soft information for the data decisions i.e. an ~ | y ) for each possible symbol c ~ . This estimate of the a posteriori probability p(c kK N kK can be used as a decoding metric in interleaved coded modulation systems [82] and in iterative decoding schemes. 3.3.2 Block Receivers In detecting a long data sequence, the sequence of received samples can be partitioned into blocks of length N and a block-based algorithm can be employed at the receiver. Block detectors can be roughly divided into two classes: 1) 2)
10

multiple-symbol ML detectors [88-90] and [94,95]; ML detectors employing the Expectation-Maximisation (EM) algorithm [91-93].

Most decision feedback receivers are designed for linearly modulated signals and, in particular, for PSK. However, they can be designed for CPM (see, for instance [84, 85, 87]). An Extended Kalman filter allows estimation of both the fading distortion and other randomly-varying system parameters even if the joint estimation problem is nonlinear.
11

20

Multiple-symbol ML detectors were proposed by in [90] for block detection of differentially encoded MPSK sequences transmitted over Rayleigh fading channels. This work showed that ML detection can be interpreted as a multiple-symbol differential detector [173,89] and that it provides an appreciable reduction of the error floor in fast fading with respect to a conventional differential receiver. The main drawbacks with respect to a conventional differential receiver are: (a) a complexity increase as M N metrics (one for each possible data sequence of length N) must be computed and compared, and (b) the receiver must estimate the second order channel statistics. Multiple-symbol differential detection algorithms have been also investigated in [88, 171, 95]. The work of [171] has provided an interesting interpretation of the ML block decoding algorithm for QAM signals and has investigated the error performance of suboptimal algorithms in the presence of coding and diversity. The EM algorithm [96] was proposed in [91, 93] as a solution to the problem of ML estimation of linearly modulated data sequences. Its application leads to a two step iterative procedure embedding a Kalman filter for channel estimation. The EM receiver needs a start-up estimate of the fading channel over the whole received block. This can be easily obtained using the PSAM technique. It has been shown [93] that the algorithm converges to the ML data solution in two or three iterations. 3.3.3 Viterbi Algorithm-Based Receivers Sequence detection receivers employing the VA [19] in their demodulation algorithm [71, 75, 86] and [97-101] provide an alternative approach. For optimality, these require assumption (A.1) concerning the channel model. Practical VA-based receivers for M-ary linearly modulated data sequences12 employ a M L -state trellis. Each state k , for the k-th signalling interval, corresponds to a hypothesised vector of L symbols, ~ ,c ~ k L +1 ,  , c ~ k 1) . k = (c kL (3.11)

There are M transitions emerging from each state k going to M different states k +1 ~ (a trellis each corresponding to one of the M possible choices for the symbol c k diagram is illustrated in Fig. 6 for M = 4 and L = 1). For any transition from k to k +1 , the conditional mean y ( k | k 1) of (3.9) and the conditional variance

y2 ( k | k 1) of (3.10) can be evaluated by means of an L -tap Wiener predictor [72] on ~ (the symbol labelling the transition), the L symbols comprising and the basis of c k k
the received samples y k ( L) ( y k L , y k L +1 , , y k 1) 13. The branch metric is associated

12 13

The same algorithms can be extended to CPM signals [71].

The number of predictors equals M L in a receiver for linearly modulated signals that processes one sample per signalling interval.

21

~ and is easily derived from (3.5) with the transition emanating from k , is labelled by c k as ~ ) = ln p( y | y ( L), k , c ~ k ) + ln p(c ~k | k ) , ( k , c k k k (3.12)

~ k ) is the ~ k | k ) denotes the transition probability and p( y | y ( L), k , c where p(c k k Gaussian probability density of (3.6). Finally, the VA is applied to the decoding trellis defined by (3.11) with the branch metric (3.12). Data decisions are provided by the VA with a fixed delay depending on L , the memory of the channel [86]. The use of the VA for ML detection of signals transmitted on fading channels was initially illustrated in [75] which derived the ML receiver assuming finite channel memory. Later, a receiver was developed in [71, 97] for ML detection of CPM signals assuming that the fading can be modelled as an AR process. The approach was extended in [86], [98-100] and [101] to the problem of ML detection of PSK signals. VA-based receivers provide better error performance than one-shot noncoherent detectors at the price of larger complexity. In practice, significant complexity reduction can be achieved. The number of symbols comprising each trellis state can be reduced from L to Q (where Q < L should be properly selected [86]). By predicting the fading process with L-tap optimal filters, the survivor path symbols can be used in place of the missing state symbols [86]. This represents a form of decision feedback and has been exploited in [102-104] to reduce the complexity of the ML receiver [105] for frequency selective channels. Finally, we note that the VA-based receiver effectively computes as many carrier references as the number of trellis states. Each fading estimate is evaluated conditioned on the channel symbols corresponding to a survivor path in the trellis. These detection algorithms employ the Principle of Per-Survivor Processing (PSP) [106], which provides a general solution to the implementation of MLSE whenever the received signal, due to imperfect knowledge of channel parameters, is affected by some degree of uncertainty. Its application consists in embedding the data aided estimation of the channel parameters in the data detection algorithm. In the detection of signal sequences transmitted over frequency-flat fading channels, the unknown parameter is the fading distortion. If the state trellis of (3.11) is defined and PSP applied to it, VA-based algorithms can be heuristically derived. 3.4 Receiver Structures for Fast Fading Channels Most of the preceding receiver structures have assumed slow time selective fading [124], where the fading has been approximated as a stepwise function, that does not change appreciably in any signalling interval for the CPM signal of (2.2) or over the signalling pulse duration for the linearly modulated signal of (2.1). Otherwise, the fading is deemed fast. In fast fading, these receivers exhibit an error floor (an irreducible error-

22

rate value as the signal-to-noise ratio increases). If coding [107, 177] and/or explicit diversity techniques [49] (e.g., space, time or frequency diversity) cannot be employed in a digital communication system operating in fast fading, reduction of the error floor can be pursued by exploiting the implicit time diversity of the channel [125]. This can lead robust receiver structures for fast fading channels and two different approaches have been proposed; namely, 1) 2) the multisampling approach [71, 126, 127]; the double-filtering approach [100, 129].

The first stems from the fact that, since the channel is rapidly changing, accurate estimates of the fading process can be obtained only if closely spaced samples of the received signal are available. This requires processing more than one sample of the received signal per signalling interval. Multisampling14 was employed in [78,79] and later in [127]. These used decision feedback symbol-by-symbol receivers for PSK signals assuming a rectangular signalling pulse p(t ) of duration equal to the signalling interval. A multisampling receiver for CPM signals was developed in [71] and one for bandlimited PSK signals in [126]. The second approach, developed in [100] for PSK and in [129] for CPM assumes a linearly time-selective channel model [134]. This consists of approximating the fading distortion by a straight line over the duration of the transmitter pulse p(t ) for linearly modulated signals and over the signalling interval for CPM signals, thus allowing for linear changes in the fading with time. The implicit time diversity can then be extracted by two matched filters (for a linearly modulated signal) [100] or two filter banks [129] (for CPM signals). In the PSK case [100], the symbol-rate samples at the outputs of the two filters are expressed by
0 0 = ak yk ck + wk 1 yk = a1 k ck + wk

(3.13) (3.14)

0 denotes the value of the fading distortion at the center of the k-th signalling where a k 1 interval and a k the slope of the straight line approximation in the same interval. The 0 sequences {y k } and {y k1 } can be processed using the same techniques as in the MAP

receivers processing one sample of the matched filter output per signalling interval. An example is the VA-based receiver of [100]. A variation of this receiver results in a so-called blind detector. FIR channel estimators are employed and their tap coefficients are evaluated based on geometrical considerations, independent of the statistical properties of the fading and the received Theoretical consideration of multisample processing in optimal detection can be found in [128].
14

23

signal to noise ratio. The resulting receivers are known as blind detectors because they ignore the statistics of the fading process. Blind VA-based algorithms have been derived for CPM signals and PSK signals in [129] and [100]. They provide good error performance and fast acquisition, usually with no training. When linearly modulated signals are employed, the double filter receiver requires a transmitter pulse with a spectrum equal to the square root of a raised cosine function with 100% roll-off in order 1 to avoid ISI in the filter output y k [100]. This reduces the bandwidth efficiency with respect to receivers operating at one sample per symbol, for which smaller values of pulse roll-off are usually selected. Finally, we note that the double-filtering approach can be used to design noncoherent detectors for PSK [130] and FSK [131] that provide a low error floor in fast fading.

4. Equalising the Doubly Selective Channel

Although the frequency-flat channel model is simple and low-complexity equalisers have been designed for it, it is often a poor model for actual wireless channels. Under some circumstances, the delay spread can reach 20s [211]. Systems with a channel symbol rate exceeding a few thousand symbols per second are adversely affected unless this dispersion is equalised. Therefore, we now consider equalisers for the time- and frequency-selective channel, also known as the doubly spread (delay and Doppler spread) channel. Equalisation for doubly spread channels is a challenging problem. Instead of one random process to estimate as in the flat fading case, there are many, one for each tap in the tapped delay line model of (2.16). However, the problem of frequency-selectivity arose first in the telephone channel [2], and the HF channel is doubly spread [49, 176], so a considerable body of theory and practice is already available. We categorise equalisers according to the statistical basis of their decision rules (rather than the more superficial historical or structural classifications). The most successful schemes account for both the channels double spreading and the additive noise. These are described in later subsections, initially addressing equalisers which treat the channel as a deterministic or known process, and subsequently describing equalisers which treat the channel as a random process. When the Doppler and delay spreads are small, they may be ignored and the performance of low-complexity non-coherent, differentially coherent or coherent receivers (without equalisation) is satisfactory [198-201]. Some of these are described in Section 3.1. However, for larger values of delay spread, or at higher SNRs, ISI leads to an error floor, where an increase in transmitter power does not improve the BER. If the floor is too high, then a more complicated receiver incorporating an equaliser is required. As a general rule, the delay spread becomes significant when its normalised value, d / T , exceeds approximately 10% [198-200].

24

4.1 Inverting Equalisers The doubly spread channel may be considered as a linear time-varying filter and linear equalisers are applicable (we discuss decision feedback equalisers (DFEs) in section 4.2.8). Linear equalisers are usually transversal filters, where the tap gains are either symbol spaced (and operate on the output of a symbol-rate sampled matched filter) [1, 2] or fractionally spaced (so the equaliser incorporates the matched filter) [2, 178, 214]. The tap coefficients are calculated to invert (at least approximately) the channels transfer function so as to eliminate or to reduce ISI, according to either the zero forcing or minimum mean square error criterion [2]. A hard decision is made on the output resulting in low equaliser complexity. At high SNR, under most criteria of optimality, these equalisers have the form of a filter with the sampled impulse response hi ,k at the ith symbol interval designed to satisfy

r
k

ir k

hi ,k ci

(4.1)

Even for a known channel, this is not straightforward [153]. When the channel is unknown, the coefficients, hi ,k , must be estimated, and continually adapted to track the changing channel [e.g., 162, 194]. This requires knowing the transmitted symbols, and so a training sequence is often transmitted. The tap weights are acquired during this. They are then tracked in a decision-directed mode, as in Fig. 7 [189, 215]. The receivers decisions should be highly reliable, and decision-directed adaptation is normally successful. However, a string of decision errors (e.g., during a deep, wideband fade) can cause decision-directed estimators to fail [189]. Thus regular retraining is often needed, at a frequency several times that of the Doppler spread [176]. An alternative strategy, when the fading is sufficiently slow [191, 216], is to adapt the tap coefficients weights during the training sequence only. Over long intervals or when the fading is faster, training sequences must be regularly interleaved with segments of data [195, 196]. The common adaptation algorithms are the Recursive Least Squares (RLS) and Least Mean Square (LMS) algorithms [2, 73]. RLS acquires the channel taps rapidly [73], and is a good choice during the training sequence [217]. The LMS algorithm is substantially less complicated than RLS or its variants [2, 73, 218, 219, 220], and therefore, is often preferred as a decision-directed estimator. For uncorrelated inputs, the tracking performance of the LMS [e.g., 73, 155] and RLS [e.g., 73, 179] algorithms are comparable. Unfortunately, LMS is slow to converge and causes significant noise enhancement when the eigenvalues of the input autocovariance are widely spread (i.e. the input is correlated) [2, 73, 155, 156]. This is typical in linear equalisation due to the ISI [221] from delay spread. Thus, computing the tap coefficients from an estimate of the channel impulse response (equalisation through channel identification) offers faster convergence.

25

Both the RLS and LMS algorithms track the channel, in the sense that their estimate of the tap coefficients is computed to equalise the channel as it was in the past. This lag error may be only partially diminished by increasing the LMS step-size [155] or by decreasing the RLS forget factor [179, 222], since then the estimation noise dominates [179]. Thus linear equalisers are best suited to quite slowly varying channels [175]. The performance of linear equalisers in specific situations is addressed in [223, 224]. Novel approaches to linear equalisation are described in [225-227]. In [225], it is the real bandpass signal that is equalised, rather than the complex baseband signal. Linear equalisers can eliminate the phase distortion of the channel satisfactorily, but amplitude distortion cannot be ameliorated without noise enhancement [2]. Since deep frequency selective fades are characteristic of wireless channels, DFEs are generally preferred to linear equalisers since their complexity is comparable and their performance suffers less under amplitude distortion. Linear equalisers, DFEs, and adaptive MLSDs are compared in [175, 194, 195, 196, 217, 228, 229]. 4.2 Equalisers for the Deterministic Channel Here, we consider equalisers derived under the assumption that the channel is known a priori. All equalisers in this section share a common metric, the squared Euclidean distance of (2.20), although they may compute it in different ways. We include blind equalisers in this section, since they also try to estimate the channel. As discussed previously, statistical detection theory offers two main criteria of optimality: minimum probability of sequence error through maximum likelihood sequence detection (MLSD or MLSE), and minimum probability of symbol error through maximum a posteriori symbol detection (MAPSD). Both approaches are discussed. Reduced complexity variants of each are also described: reduced state sequence detection (RSSD or RRSE) and Bayesian decision feedback, respectively. The assumption of a known channel is reasonable in the slowly time-varying case, where an adaptive estimator can readily track the channel. However, channel estimation is much harder in fast fading channels, and impossible in overspread channels. 4.2.1 MLSD with an Actually Known Channel Here, the ideal case is discussed, where the channel is assumed known, (e.g., a genie-aided equaliser). Although there is much earlier research [e.g.15,105], it was not until the work of [105] that a finite complexity ML sequence detector for time-invariant channels was synthesised. The receiver front-end has the form of a whitened matched filter, sampled at the symbol rate. Its outputs, y n , are compared with noiseless hypothesised signals using a squared Euclidean metric, thus creating a number of branch metrics15,
15

This metric is not well-defined as r . 26

i =

(i +1)r 1
n = ir

yn

k k = L +1+ i

c h

n ,n kr

(4.2)

where hn ,n kr is the sampled received pulse shape defined by substituting (2.1) into (2.4), L is the (assumed) finite duration of the overall channel response, and we assume the receiver front-end to comprise a noise-limiting filter and a fractionally-spaced sampler. There are ML branch metrics due to the ISI combinations. Sequence metrics similar to (2.20) are computed by recursively summing the branch metrics of (4.2), and the search for the maximum sequence metric is efficiently conducted by the Viterbi algorithm. In [135], a structure is proposed that replaces the whitened matched filter with the traditional matched filter, and [142, 157] describe two other variations. MLSD performance for arbitrary channels is bounded in [105, 135, 230, 231]. For fading channels, it is necessary to average over the pdf of the fading process [19]. For WSSUS channels, the averaging must be conducted over each value of delay, although more convenient mathematical techniques are available [133]. In addition, it has been found that the BER can be significantly improved by exploiting the implicit delay diversity [19,232,233], as in CDMA systems, through RAKE detection [234]. It is instructive to consider MLSD for known, time-varying channels. Although a channel estimator for this case is much more difficult to devise, [15] alludes to the solution, and it is described in [136, 138]. The two path case is addressed in [235], and joint equalisation and decoding are considered in [236]. With linearly modulated signals, the receiver front-end is a filter matched to the received pulse (the transmitter pulse convolved with the time-varying channel) [138]. The filter output is sampled at the symbol rate, and processed as in the time-invariant case of [135]. Analysis shows two forms of implicit diversity: due to delay and Doppler spreading [138]. 4.2.2 Pilot Information Pilot signals allow for channel estimation prior to data detection. In the timeinvariant channel only a single training sequence is required; in the frequency-flat fading channel a single pilot tone or a sequence of pilot symbols suffice, as in section 3.2. In the doubly spread channel, a single training sequence cannot track the time-varying channel [191, 208, 237], nor can a pilot symbol sequence efficiently measure frequencyselectivity [198, 238], since the adjacent data symbols overlap the pilot symbols and obscure the channel information in the known pilot symbol. More general methods are needed, so as to maintain orthogonality between the pilot and data-bearing signals at the receiver, and to allow the pilot information to be extracted before detection [51, 138]. In general, a comb of pilot tones is required to characterise the channel in frequency as well as in time [138].

27

In many systems, the channel is slowly time-varying but highly frequencyselective. Then, the method of [180, 197, 239] is superior, where training sequences of pseudo-random symbols are transmitted periodically. The equaliser estimates the channel response from each. By interpolating [180, 239] or Wiener filtering [197] between training sequences, the channel may be estimated over the whole transmission. The sequences are usually several times the length of the received pulse, and must be spaced at the Nyquist rate for the channels Doppler spread. Thus throughput is low except for small Doppler spread. For faster fading, very short training sequences may be used as when isolated pilot symbols are employed [138].

4.2.3 Adaptive MLSD The MLSD of section 4.2.1 was derived assuming a known channel. In practice channel estimation is required. Adaptive MLSD provides the simplest structure. A single channel estimator is employed. The transmitted symbol sequence is detected using the estimated channel impulse response (i.e. the known channel impulse response,  hn ,n kr , at time t = nTr, in (4.2) is replaced by its estimate, h n ,n kr ). This estimate is updated according to the detected sequence, as shown in Fig. 8 [135, 144]. When a matched filter is used as a front-end, it too must be updated [135, 240]. Transmission normally starts with a training sequence to initialise the channel estimator. The receiver is then switched to a decision-directed mode, where tentative decisions from the survivor sequence with best metric [241] are fed-back to the estimator from the Viterbi processor after some delay [182, 242, 243]. The Viterbi decision delay is on the order of five times the channel memory [19], so the channel estimator has only out-dated information available, and the estimate suffers a lag error [140, 241]. For time-varying channels, this must be traded off against the accuracy of the tentative decisions, so the tentative decision delay may be chosen to be less than the Viterbi decision delay [144]. The lag error may also be diminished by predicting the channel estimate [188, 203, 209]. As with linear equalisers, a fast acquiring algorithm is desired [188]. For tracking, the low complexity LMS algorithm is invariably favoured in adaptive MLSD and adaptive PSP MLSD (see subsection 4.2.4) [135, 140, 144, 146, 243]. The estimator inputs are the MLSDs tentative decisions. When the input data correlation is low, the tracking ability of the LMS [73, 155, 156, 244, 245] and RLS [73, 179, 246-250] algorithms approximately match. The BER performance of the LMS-adaptive MLSD is analysed in [203]. A floor occurs when the fading is too fast (e.g. a BER floor of 10-3 was observed for a channel with two taps, of equal mean power, spaced by one symbol interval, for a normalised Doppler spread, fDT, of 10-3). Thus, adaptive MLSD is suited only to very slowly fading channels.

28

Performance is simulated for GSM-like systems in [182, 242]. Trellis-codedmodulation over doubly spread channels is addressed in [243, 251], where the need for effective, low complexity equalisers compatible with interleaving is identified. Adaptive MLSD is compared with other equalisation strategies in [194, 195, 252]. 4.2.4 Adaptive PSP MLSD The tentative decision delay of adaptive MLSD is unsatisfactory for timevarying channels and to avoid this, Per Survivor Processing (PSP) [106] may be applied [140, 141, 142, 146, 253, 254]. Each survivor in the trellis has an associated channel ~  estimator and an estimate, h n ,n kr ( ck ) , which may be updated with no lag using the survivors hypothesised symbols [106]. A generic block diagram is shown in Fig. 9. Adaptive PSP MLSD is motivated by the inadequacy of adaptive MLSD in timevarying channels, but it is truly optimal only in the time-invariant channel, using the RLS algorithm for channel estimation [141, 142]. When used in time varying channels, the LMS algorithm is preferred [140, 141, 142, 146]. Error floors can still occur, even in relatively slow fading. In [241], adaptive MLSD and adaptive PSP MLSD are studied. For the non-fading but Doppler shifted channel, the additional complexity of adaptive PSP MLSD is found to be unwarranted, although in [140], it outperformed adaptive MLSD. As a performance-complexity trade-off, the number of channel estimators may be reduced [149]. 4.2.5 Reduced Complexity MLSD Receiver complexity is governed by the number of branches multiplied by the complexity of the branch metric. There are ML branches, where M is the constellation size and L is the length of the received pulse. The trellis is usually infeasibly large, and thus there is considerable interest reducing it. In many cases the energy in the tails of the received pulse is small, and can be neglected without significant penalty. This leads to truncated sequence detection [255, 256, 257]. Further simplification is obtained by observing that reliable decisions can be made once most of the pulse energy (i.e., its main lobe) has been received. The postcursors can then be dealt with by decision feedback [104, 188, 258, 259]. Since insufficient energy is available to make reliable decisions on the precursors, all combinations must be hypothesised through the MLSD. This is called delayed decision feedback sequence detection (DDFSD or DDFSE) [104]. Reduced state sequence detection (RSSD or RSSE) refines this idea. Instead of a full trellis for the precursor and a single decision history per survivor for the postcursor, set partitioning principles [109] are applied to steadily reduce the number of hypothesised symbols as more of the received pulse arrives [102, 103]. In [260], finer control of the number of states is achieved. One analysis of RSSD is presented in [102] and tighter performance bounds are provided in [261].

29

The span of the received pulse may also be reduced by adaptive prefiltering to obtain some desired impulse response [159, 262], using a linear equaliser [145, 159] or a decision feedback equaliser [158, 161, 263]. MLSD is applied to this pre-filtered signal. However, the pre-filter colours the additive noise, thereby reducing performance if it is not taken into account [157]. Also a DFE pre-filter exhibits error propagation, so a hybrid structure that delivers the MLSDs soft outputs to the DFE is superior [161, 181, 188]. This structure is closely related to the DDFSD of [104]. Another approach is to retain the full trellis, but to search it more intelligently. The Viterbi, single stack, Fano, 2-cycle, stack, merge, bucket and M- algorithms are reviewed and compared in [264]. More recent algorithms are presented in [265, 266]. Equaliser performance using the M-algorithm [187, 267, 268] and the Fano algorithm [269] has been studied. It is clear that exhaustive trellis search via the Viterbi algorithm is wasteful, as the other algorithms attain excellent performance at reduced complexity. The complexity of computing the branch metric may be reduced, by adopting a reduced complexity channel model. Provided the model is sufficiently accurate, little degradation results. For channels causing sparse ISI, complexity savings may be obtained without loss of optimality [270]. In [151, 164, 166], the channel transfer function is modelled as a power series in frequency with time varying coefficients [134], instead of the usual tapped delay line. The latter requires a large number of taps if path delays do not match the model delays (which happens invariably). The work on series models has shown that only a modest number of terms (i.e. a low-order polynomial) is required to obtain excellent performance for a large class of channels. Computing the branch metric is simpler, and the number of channel parameters to be estimated is much reduced. A dual approach is adopted in [48, 100, 129, 130, 131], where a polynomial model of the channels time-variation is adopted. 4.2.6 Known Channel MAPSD When the criterion for optimality is the maximum likelihood sequence, the equaliser is efficiently implemented by the Viterbi algorithm [105]. However, BER is normally used to compare equalisers, so a better criterion is to detect the symbols with maximum a posteriori probability [271]. Little attention has been paid to MAP symbol detection (MAPSD), since its implementation requires a two-way recursion [272, Appendix] unsuited to long transmissions. There is renewed interest in MAPSD, since the soft symbol probabilities preferred by decoders are directly computed and its low SNR performance is utilised in Turbo decoding [273]. In doubly spread known channels with a noise-limiting receiver front-end followed by a fractionally-spaced sampler, an inefficient implementation of the MAPSD branch metric is [274]

30

i =

(N

1
0

Tr )

i y n ck hn ,n kr ( ) 1 1 + i r k = L +1+ i exp N 0 Tr n =ir

(4.3)

4.2.7 Reduced Complexity MAPSD A D-lag MAPSD or MAP symbol-by-symbol detector is proposed in [275]. It makes a MAP decision on a symbol, given all past samples but only the next Dr samples beyond. Under this constraint, only a forward trellis pass is required [276]. The trellis size is exponential in D, and for near-optimal performance, D must be at least as large as the received pulse length, L, and possibly as large as 5L [277]. Problems, such as the need for multiplication and exponentiation, and the potential for underflow, have been largely overcome [278]. In the doubly spread channel, MAPSD is complicated by the need to track the channel. Kalman filtering is used in [193, 279], and LMS adaptation is considered in [279]. Bayesian decision feedback equalisers (BDFEs) are analogous to DDFSDs or RSSDs, since both use decision feedback to reduce the complexity of the optimal structure [280]. However, Bayesian DFEs adopt a structure that is quite different to the recursive D-lag MAPSD; namely that of a DFE, but with the feedforward and feedback tap coefficients replaced by a nonlinear structure [205]. The BDFE treats detection as a nonlinear classification problem [281]. In decoding a symbol, a fixed-length vector of received samples in the vicinity of the symbol identifies a point in a multidimensional space. The space is divided into M nonlinear regions, one per possible decision. These regions are each the union of smaller decision regions, one for each of the ML ISI combinations. The MAPSD makes a decision by locating the decision region in which the received sample vector lies. Computational savings are achieved in BDFE over the full MAPSD implementation, since past decisions reduce the number of allowed ISI combinations, and thus the number of smaller decision regions that exist [205]. A block approach is presented in [204]. Neural networks are effective nonlinear identification algorithms. Several models have been applied to BDFE [281, 282, 283], to achieve the equalisation of nonlinearly distorted signals [284] and to suppress adjacent channel interference (ACI) [282] and co-channel interference (CCI) [283]. A promising technique for doubly spread channels is pursued in [202, 205, 285, 286]. In [202], the LMS algorithm is used to adapt a complex radial basis function network in a GSM-like channel. BDFE outperforms the adaptive MLSD of section 4.2.3.

31

4.2.8 Decision Feedback Equalisation In the extreme case, the nonlinear boundaries of MAPSD become hyperplanes, and it simplifies to the well known decision feedback equaliser (DFE) which comprises feedforward and feedback transversal filters connected to a decision device. In the timevarying channel, their impulse responses are time-varying. The feedforward filter, { f i , k } , processes the input signal so as to minimise the channels phase and amplitude distortion, and the feedback filter, {g i ,l } , processes the decision devices output so as to subtract postcursor ISI from the decision devices input (assuming correct past decisions). The input to the decision device equals i l c i f i ,k yir k gi ,l c
k l >0

(4.4)

i is its output. A DFE performs better than a linear equaliser on channels with where c severe amplitude distortion. When an decision error is made, its effect lingers due to the feedback filter, potentially causing error propagation. However, for most practical channels, the DFE recovers quickly [175, 287-289]. Its BER performance is analysed in [290, 291]. Analysis is simplified when the effect of incorrect past decisions is neglected. Under this assumption, the mean square error is calculated in [174], and for two path diversity outage probabilities are calculated in [292] and the BER is bounded in [293, 294]. DFE filters are normally designed under the mean square error criterion [19, 174, 175], and may be separately optimised [295]. The filter coefficients may be adapted directly from the error between the decision devices input and output, normally by the LMS [174, 176, 190] or RLS [176, 190] algorithms. Other schemes include equalisation by channel identification [191, 296], a more robust approach in time-varying fading. Interpolating channel estimates or tap coefficients obtained from periodic training sequences is one means of channel identification [180, 297]. Others are described in subsections 4.2.2 and 4.2.3. A fractionally spaced adaptation scheme is described in [168] and [259] presents a variant for channels with long delay spread. The consequence of a poor channel estimate is studied in [191], where again an error floor is observed in time-varying channels. The DFE offers a nice balance of complexity and performance. As with RSSD, best performance is achieved when the channel is minimum phase, yet this is difficult to assure in a time-varying channel. Reversing time may convert a channel with a undesirably large precursor into one that can be equalised more easily [189, 298, 299]. When DFEs are used with interleaved coded modulations, more reliable decisions are available from the decoding algorithm, and it is advantageous to use them in the DFE feedback loop [300]. A modified DFE structure is required [173], and [300]

32

shows that this relatively simple equalisation scheme can approach channel capacity. In [301], a similar approach is adopted, and the fading frequency-selective channel is addressed. DFEs are also used in equalising CPM signals, where ISI is due in part to the memory in the modulation scheme rather than a dispersive channel [44, 84, 97], and in equalising differentially demodulated signals [302]. Block equalisers, for short transmissions between training sequences, may also exploit past decisions [195,196]. 4.2.9 Blind Equalisers for Frequency Selective Channels Thus far, some form of reference signal has been explicitly or implicitly assumed: training sequences, pilot tones or pilot symbols. These reference signals are used to estimate the channel impulse response, so equalisation may commence. The reference signals lead to reduced complexity equalisers and ensure fast acquisition, but they are an unnecessary overhead as the blind equalisation literature shows. Blind equalisation potentially has many benefits in wireless channels, which are characterised by unpredictable deep fades. An equaliser which can recover automatically from these is indeed worthwhile. Most of the literature on blind equalisation is directed at wireline applications and other very slowly varying channels. The early schemes of [303-307] converge slowly. In [308], it is shown that second order cumulants of symbol-spaced sampled signals are inadequate to identify non-minimum phase channels, and thus many blind equalisers have been proposed based on higher order cumulants, as in [309-315]. These require significant time for accurate computation, and are generally unsuited to the short packet lengths and time-varying channels typical of wireless systems. These deficiencies have been realised [312, 316-318], and recently, blind equalisers with fast acquisition times have been developed [316, 317]. An effective method is to employ sequence detection [141], fractional sampling and second order cumulants together [142, 316, 318-320], although fractional sampling is not always sufficient [321] (a deficiency avoided in [322]). Such blind equalisers acquire in approximately a hundred symbol periods [141, 322, 323], so packets may be processed in two passes: one to estimate the channel, and a second to make decisions. However, this does not attain the goal of swift, automatic recovery after a fade, when it is desirable to re-acquire within a few tens of symbols. One approach is to account specifically for the channels time-variation in the blind equalisers channel model [317]. Other blind equalisers are presented in [279, 318, 324-326]. Blind equalisation in conjunction with antenna arrays is investigated in [328, 329]. 4.3 Equalisers for the Random Channel It is surprising that little attention has been paid to this class of equaliser, since they are explicitly designed for doubly spread channels. Most others have been ad hoc 33

solutions employing the LMS or RLS algorithms to provide adaptivity. Early work was aimed at the ionospheric and tropospheric scattering channels [4, 5, 76, 147]. The problem was identified as detecting Gaussian signals (i.e. Rayleigh or Rician faded signals) in Gaussian noise, through their different autocovariances [5, 15, 22, 24, 76, 147]. Only the time-varying channel and noise autocovariances need to be calculated, and the time-varying channel impulse response is not required, although in some special cases it is actually estimated [77, 86, 98, 171]. The detection of non-Gaussian signals has also been investigated [21, 23, 143]. Kailaths estimator correlator structure is well known and theoretically elegant [15-18]. In coloured noise, a non-causal filter is required [15], but this case is generally unimportant. A disadvantage of these structures is that sequence detection is undertaken using a brute force search [15, 32]. As discussed in section 3.3.3, the Viterbi algorithm may be introduced under the assumption of finite memory [19, 20]: that is, the minimum mean square estimate, y ( k | k 1) , of yk, conditioned on a data sequence and the past received samples, requires only a finite number of past samples16. This is shown more fully in [77] for realistic pulse shapes and Rayleigh fading channels. Using Bayes theorem, coupled with the Innovations process or the Cholesky decomposition [71, 77], the sequence metric of (2.10) can be decomposed into the conditionally independent branch metrics (also derivable from (3.6)),
(i +1)r 1
k = ir

i =

y k y ( k | k 1)

2 2 ( k | k 1) + ln y

y2 ( k | k 1)

(4.5)

where 2 y ( k | k 1) is the conditional variance of the prediction. This metric may be interpreted as the squared Euclidean distance between the received signal and the expected received signal, normalised by its variance. The bias term, ln y2 ( k | k 1) , may be neglected for some channels and modulations [77, 86, 98, 171]. The conditional mean, y ( k | k 1) , is computed as the weighted sum of past received samples. The metrics data dependence arises in the conditional mean and variance, and their computation requires the channel and noise autocovariances. Several methods for estimating these have been proposed [127, 139, 330, 331]. The error performance of the innovations-based MLSE scheme, described in [77] for a Rayleigh fading doubly spread channel is analysed in [331]. Alternatives to the MLSE solution are the ML-based sliding block detection scheme of [332] and the Kalman filter based MLSE receiver of [148, 164], where the received pulses time variation and the channel are modelled as ARMA and AR processes respectively. MLSE

Estimates based on a finite number of samples are, strictly speaking, optimal only under assumption (A.1).

16

34

employing Kalman filter channel estimation is analysed in [169]. Another sequence detector for unknown channels is the MAP receiver of [193] which employs a bank of extended Kalman filters to jointly estimate the channel, data and symbol timing. Finally, it can be shown, given a high enough sampling rate (Nyquist rate for the received signal) and sufficiently accurate predictors, the error floor in fast doubly selective fading can be lowered arbitrarily [71, 160].

5. Combined Equalisation and Decoding

In order to achieve the required error performance, most digital wireless systems require some form of forward error correction coding [177]. However, wireless systems usually operate in a tightly band limited environment which must be used in a spectrally efficient manner. Conventional error correction techniques require additional bandwidth to maintain the required data rate and this is often unacceptable. Consequently, there is much interest in the use of spectrally efficient coding techniques such as trellis coded modulation and multi-level coded signalling for digital wireless transmission. The use of these bandwidth efficient techniques, where the coding is embedded in an expanded signal set leads to a requirement for combined equalisation and decoding techniques. In the time-invariant ISI channel, two major approaches, reduced state sequence estimation (RSSE) [102] and [103] and delayed decision feedback [104], have been investigated. To date, there have been a few attempts to develop similar techniques for use in the digital wireless environment. In the case of trellis coded modulation, researchers have considered receivers which perform joint MLSE and decoding on an enlarged trellis structure often resulting in unacceptable system complexity [86, 103]. One of the problems with coded transmission is that in slow frequency selective fading, interleaving is required in order to randomise the effects of the channel at the receiver. This makes it difficult to develop joint equalisation and decoding approaches due to the requirement to de-interleave the received signal stream following channel equalisation but preceding the decoding algorithm. All the equalisation techniques considered for uncoded transmission can be easily extended to coded signalling [107] in the absence of interleaving. In particular, both block [88, 171, 95, 108] and VA-based [75, 86] detection algorithms have been proposed for the detection of uninterleaved coded sequences and, in particular, of TCM signals [109]. If interleaved channel codes are used [110-115] other receiver architectures are needed. In fact, when an interleaver is employed, channel estimation must be performed before deinterleaving in order to exploit the correlation of the fading process. Then the channel estimation can no longer be embedded in the data decoding as in the receiver structures considered earlier and leads to sub-optimum two-stage receiver structures [67,86,101] and [116-119], and illustrated in Fig.10. Here the first stage accomplishes channel state estimation ignoring the effects of coding, and the second stage carries out

35

the decoding of the coded signal using the deinterleaved fading estimates generated by the first stage. The major difference among the proposed two-stage receiver structures for coherent detection of interleaved coded linearly modulated signals is in the design of the channel estimator. Both [116] and [67] suggest employing only the pilot symbol samples to estimate the fading distortion at the receiver. This strategy is called TCMP17 [67]. It can be improved if a fraction of the channel state information is extracted from the data bearing symbols, as suggested in [117] and in [119]. A conceptually similar detector is suggested in [118]. A two-stage receiver structure has also been proposed in [86] and in [101]. In both papers a VA-based receiver is employed in the first stage and its symbol estimates feed a channel estimator which provides a smoothed estimate of the fading process. Simulation results reported in [86] indicate superior performance compared to other two-stage receivers at the price of increased complexity. Finally, we make the following observations: 1) In interleaved coded transmission an alternative to channel equalisation is the use of differential encoding in the transmitter and differential detection before deinterleaving at the receiver. This solution has been proposed in [120] and yields poorer performance than the above two-stage structures [86]. The cited references describe receiver architectures for interleaved coded PSK signals. However, interleaved trellis-coded CPM signals represent an alternative to PSK signals [122]. The detection of interleaved trellis encoded CPM signals has been investigated in [121, 123] using two-stage receiver structures. The work of [123] suggests using a noncoherent detector in the first stage, whereas a coherent detection scheme is proposed in [121]. However, [121] does not provide any practical solution to the problem of carrier recovery.

2)

In general, optimal joint equalisation and decoding for wireless communications remains an open problem, primarily due to the complexity of the required signal processing. There is some suggestion that the use of turbo codes and iterative decoding structures may lead to solutions that have a manageable complexity [333]. However, a generally applicable solution remains to be found.

6. Conclusions
In this paper, we have investigated the state of the art in wireless channel equalisation techniques. The paper has focussed specifically on the equalisation problem and has not considered some important and interesting related topics including adaptive transmitters, joint transmitter and receiver optimisation, nonlinear channels or synchronisation issues. The paper has not attempted to provide any in depth analysis, but
17

TCMP results from the application of the PSAM technique to an interleaved TCM transmission.

36

has attempted to provide a sufficient survey to be useful to researchers either entering the field or to those workers who wish a convenient source of reference material. The interested reader is referred to the extensive list of references. After a brief consideration of channel modelling, classification of channel types and some aspects of receiver structures, we have described techniques for the equalisation or compensation of time-selective or flat fading channels. This has included a discussion of conventional non-coherent reception techniques in which no attempt is made to estimate and/or compensate for the channel. Strictly speaking such receivers are not equalisers; however, their analysis provides the basis for the more complex receivers which estimate the channel state and compensate for the fading effects. We have found that the most successful of these incorporate maximum likelihood sequence estimation (MLSE) algorithms. These usually employ the VA and often use per-survivor processing to estimate the channel. In addition, structures which are effective in compensating for fast fading channels have been considered. Section 4 concerns itself with equalising the time and frequency selective wireless channel. Again the most successful structures are based on the theory of optimal receivers and usually include MLSE algorithms again implemented by means of the VA. In all cases it has been found that the error floors, present when conventional receivers are employed, can be substantially lowered. We have also considered the influence of the channel model on the resulting equaliser strategy. We have seen that, through the use of reduced complexity channel models, it is possible to arrive at simplified receivers that yield optimum or nearoptimum performance provided that the reduced complexity channel model is sufficient. Finally, we have briefly discussed the problem of joint equalisation and decoding. We have noted that most of the receiver/equaliser structures are easily extended to decode trellis and other signal space codes, provided that interleaving is not employed. When interleaving is used, then other strategies such as two stage receivers are required in order to both estimate and compensate the channel and to decode. It has been noted that the general problem of joint equalisation and decoding remains an open problem in the fading, dispersive wireless channel environment.

References
1. 2. 3. R.W. Lucky, J. Salz, & E.J. Weldon, Jr., Principles of Data Communication, New York, NY: McGraw-Hill, 1968. S.U.H. Qureshi, Adaptive Equalization, Proc. IEEE, vol. 73, no. 9, pp.13491387, Sept. 1985. G. L. Turin, Communication through Noisy, Random Multipath Channels, Proc. of the IRE National Convention Record, pt. 4, pp. 154-166, March 1956.

37

4. 5.

6. 7. 8.

9. 10. 11. 12. 13. 14.

15. 16.

17. 18.

19. 20. 21. 22.

R. Price, The Detection of Signals Perturbed by Scatter and Noise, IRE Trans. Inf. Theory, vol. 4, pp. 163-170, Sept. 1954. R. Price, Optimum Detection of Random Signals in Noise with Application to Scatter-Multipath Communications -I, IRE Trans. Inf. Theory, vol.6, pp. 125135, Dec. 1956. H. B. Law, The Detectability of Fading Radiotelegraph Signals in Noise, Proc. of the IEE, vol. 104, pt. B, March 1957. J. N. Pierce, Theoretical Diversity Improvement in Frequency-Shift Keying, Proc. IRE, vol.46, pp. 903-910, May 1958. G. L. Turin, Error Probabilities for Binary Symmetric Ideal Reception through Nonselective Slow Fading and Noise, Proc. IRE, vol. 46,pp. 1603-1619, Sept. 1958. H. B. Voelcker, Phase-Shift Keying in Fading Channels, Proc. of the IEE, vol.107, pt. B, pp. 31-38, January 1960. J. N. Pierce and S. Stein, Multiple Diversity with Nonindependent Fading, Proc. IRE, vol.48, pp. 89-104, Jan. 1960. G. L. Turin, On Optimal Diversity Reception, IRE Trans. Inf. Theory, vol.7, no.3 pp.154-167, July 1961. G. L. Turin, On Optimal Diversity Reception, II, IRE Trans. Commun. Syst., pp. 22-31, March 1962. P. M. Hahn, Theoretical Diversity Improvement in Multiple Frequency Shift Keying, IRE Trans. Commun. Systems, vol. 10, pp. 177-184, June 1962. P. A. Bello and B. D. Nelin, The Influence of Fading Spectrum on the Binary Error Probabilities of Incoherent and Differentially Coherent Matched Filter Receivers, IRE Trans. Commun. Systems, vol. 10, pp. 160-168, June 1962. T. Kailath, Correlation Detection of Signals Perturbed by a Random Channel, IRE Trans. Inf. Theory, vol. 6, pp. 361-366, June 1960. T. Kailath, Optimum Receivers for Randomly Varying Channels, Proc. of the Fourth London Symposium on Information Theory, Butterworth Scientific Press, London, pp.109-122, 1961. T. Kailath, On Multilink and Multidimensional Channels, IRE Trans. Inf. Theory, vol. 8, pp. 260-262, April 1962. T. Kailath, Adaptive matched filters, R. Bellman ed., Mathematical optimization techniques, Berkeley, CA, University of California Press, pp. 109140, 1963. J. G. Proakis, Digital Communications, Second Edition. New York: McGrawHill, 1989. J. B. Anderson, T. Aulin and C. E. Sundberg, Digital Phase Modulation, New York: Plenum Press, 1986. T. Kailath, A General Likelihood-Ratio Formula for Random Signals in Gaussian Noise, IEEE Trans. Inf. Theory, vol. 15, pp. 350-361, May 1969. T. Kailath, Likelihood Ratios for Gaussian Processes, IEEE Trans. Inf. Theory, vol. 16, pp. 276-287, May 1970.

38

23. T. Kailath, A Further Note on a General Likelihood Formula for Random Signals in Gaussian Noise, IEEE Trans. Inf. Theory, vol. 16, pp. 393-396, July 1970. 24. T. Kailath, B.C. Levy, L. Ljung and M. Morf, Fast Time-Invariant Implementations of Gaussian Signal Detectors, IEEE Trans. Inf. Theory, vol. 24, pp. 469-477, July 1978. 25. P. Y. Kam and C. H. Teh, Reception of PSK Signals over Fading Channels via Quadrature Amplitude Estimation, IEEE Trans. Comm., vol. 31, pp. 10241027, August 1983. 26. P. Y. Kam and C. H. Teh, An Adaptive Receiver with Memory for Slowly Fading Channels, IEEE Trans. Comm., vol. 32, pp. 654-659, June 1984. 27. P. Y. Kam, Generalized Quadratic Receivers for Orthogonal Signals Over the Gaussian Channel with Unknown Phase/Fading, IEEE Trans. Comm. vol. 43, pp. 2050-2058, June 1995. 28. W. D. Lindsey, Error Probability for Rician Fading Multipath Reception of Binary and N-ary Signals, IEEE Trans. Inf. Theory, vol. 10, pp. 339-350, Oct. 1964. 29. W. D. Lindsey, Error Probability for Incoherent Diversity Reception, IEEE Trans. Inf. Theory, vol. 11, pp. 491-499, Oct. 1965. 30. W. F. Walker, The Error Performance of a Class of Binary Communications Systems in Fading and Noise, IEEE Trans. Comm. Syst, vol.12, pp. 28-45, March 1964. 31. G. D. Hingorani, Error Rates for a Class of Binary Receivers, IEEE Transactions on Communication Technology, vol. 15, pp. 209-217, April 1967. 32. J. C. Hancock and P. A. Wintz, Signal Detection Theory, New York: McGrawHill Book Company, 1966. 33. L. J. Mason, Error Probability Evaluation for Systems Employing Differential Detection in a Rician Fast Fading Environment and Gaussian Noise, IEEE Trans. Comm., vol. 35, pp. 39-46, January 1987. 34. L. J. Mason, An Error Probability Formula for M-ary DPSK in Fast Rician Fading and Gaussian Noise, IEEE Trans. Comm., vol. 35, pp. 976-978, Jan. 1987. 35. H. Salwen, Differential Phase-Shift Keying Performance under Time-Selective Multipath Fading, IEEE Trans. Comm., vol. 23, pp. 383-385, March 1975. 36. Y. Miyagaki, N. Morinaga and T. Namekawa, Error Rate Performance of M-ary DPSK Systems in Satellite/Aircraft Communications, Proc. IEEE Int. Conf. Commun.,1979, pp. 34.6.1-34.6.6. 37. A. Neul, Bit Error Rate for 4-DPSK in Fast Rician Fading and Gaussian Noise, IEEE Trans. Comm., vol. 37, pp. 1385-1387, Dec. 1989. 38. S. Elnoubi, Probability of Error Analysis of Digital Partial Response Continuous Phase Modulation with Noncoherent Detection in Mobile Radio Channels, IEEE Trans. Veh. Tech., vol. 38, pp. 19-30, February 1989.

39

39. I. Korn, Error Probability of M-ary FSK with Differential Phase Detection in Satellite Mobile Channel, IEEE Trans. Veh. Tech., vol. 38, pp. 76-85, May 1989. 40. I. Korn, GMSK with Differential Phase Detection in the Satellite Mobile Channel, IEEE Trans. Comm., vol. 38, pp. 1980-1986, Nov. 1990. 41. H. L. Van Trees, Detection, Estimation and Modulation Theory, Part III, New York: John Wiley & Sons, 1971. 42. W. C. Jakes, Microwave Mobile Communications, Piscataway, NJ: IEEE Press, 1993. 43. S. Elnoubi, Analysis of GMSK with Discriminator Detection in Mobile Radio Channels, IEEE Trans. Veh. Tech., vol. 35, pp. 71-76, May 1986. 44. S. Elnoubi, H. Badr and E.A. Youssef, BER Improvement of PRCPM in Mobile Radio Channels with Discriminator Detection Using Decision Feedback Equalization, IEEE Trans. Veh. Tech., vol. 40, pp. 694-699, Nov. 1991. 45. I. Korn, M-ary Frequency Shift Keying with Limiter-Discriminator-Integrator Detector in Satellite Mobile Channel with Narrow-Band Receiver Filter, IEEE Trans. Commun., vol. 38, pp. 1771-1778, Oct. 1990. 46. I. Korn, GMSK with Limiter Discriminator Detection in Satellite Mobile Channel, IEEE Trans. Comm., vol. 39, pp. 94-101, Jan. 1991. 47. D.K. Asano and S. Pasupathy, Improved Post-Detection Processing for LimiterDiscriminator Detection of CPM in A Rayleigh, Fast Fading Channel, IEEE Trans. Veh. Tech., vol.44, pp. 729-734, Nov. 1995. 48. G.M. Vitetta, U. Mengali and D.P. Taylor, Optimal Noncoherent Detection of FSK Signals Transmitted over Linearly Time-Selective Rayleigh Fading Channels, To appear in the IEEE Trans. Comm. 49. S. Stein, Fading Channel Issues in System Engineering, IEEE J. Select. Areas Commun., vol. 5, pp. 68-89, February 1987. 50. C. K. Rushforth, Transmitted-Reference Techniques for Random or Unknown Channels, IEEE Trans. Inf. Theory, vol. 9, pp. 39-42, January 1964. 51. G.D. Hingorani and J.C. Hancock, A Transmitted-Reference System for Communication in Random or Unknown Channel, IEEE Trans. Comm. Systems, vol. 13, pp. 293-301, September 1965. 52. W.J. Weber, Performance of Phase-Locked Loops in the presence of Fading Communication Channels, IEEE Trans. Comm., vol. 24, pp. 487-499, May 1976. 53. M. Yokoyama, BPSK System with Sounder to Combat Rayleigh Fading in Mobile Radio Communication, IEEE Trans. Veh. Tech., vol. 34, pp. 35-40, Feb. 1985. 54. M.K. Simon, Dual-Pilot Tone Technique, IEEE Trans. Veh. Tech., vol. 35, pp. 63-70, May 1986. 55. F. Davarian, Mobile Digital Communication via Tone Calibration, IEEE Trans. Veh. Tech., vol. 36, pp. 55-62, May 1987.

40

56. J.P. McGeehan and A.J. Bateman, Phase-Locked Transparent Tone-in-Band (TTIB): A New Spectrum Configuration Particularly Suited to the Transmission of Data Over SSB Mobile Radio Networks, IEEE Trans. Comm., vol. 32, pp. 81-87, Jan. 1984. 57. A. Bateman, Feedforward Transparent Tone-In-Band: Its Implementations and App.lications, IEEE Trans. Veh. Tech., vol. 39, pp. 235-243, Aug. 1990. 58. J.K. Cavers, Performance of Tone Calibration with Frequency Offset and Imperfect Pilot Filter, IEEE Trans. Veh. Tech., vol. 40, pp. 426-434, May 1991. 59. M.P. Fitz, Comments on Tone Calibration in Time-Varying Fading, IEEE Trans. Veh. Tech., vol. 41, pp. 211-213, May 1992. 60. M.P. Fitz, A Dual-Tone Reference Digital Demodulator for Mobile Digital Communications, IEEE Trans. Veh. Tech., vol. 42, pp. 156-165, May 1993. 61. J.H. Lodge, M.L. Moher and S.N. Crozier, A Comparison of Data Modulation Techniques for Land Mobile Satellite Channel, IEEE Trans. Veh. Tech., vol. 36, pp. 28-34, Feb. 1987. 62. J. Cavers and M. Liao, A Comparison of Pilot Tone and Pilot Symbol Techniques for Digital Mobile Communication, IEEE Globecom92 Rec., pp. 915-921. 63. J.K. Cavers, An Analysis of Pilot Symbol Assisted Modulation for Rayleigh Fading Channels, IEEE Trans. Veh. Tech., vol. 40, pp. 686-693, Nov.1991. 64. J.K. Cavers and J. Varaldi, Cochannel Interference and Pilot Symbol Assisted Modulation, IEEE Trans. Veh. Tech., vol. 42, pp. 407-413, Nov. 1993. 65. S. Sampei and T. Sunaga, Rayleigh Fading Compensation for QAM in Land Mobile Radio Communications, IEEE Trans. Veh. Tech., vol. 42, pp. 137-146, May 1993. 66. T. Sunaga and S. Sampei, Performance of Multi-Level QAM with PostDetection Maximal Ratio Combining Space Diversity for Digital Land-Mobile Radio Communication, IEEE Trans. Veh. Tech., Vol. 42, pp. 294-301, Aug. 1993. 67. M.L. Moher and J.H. Lodge, TCMP - A Modulation and Coding Strategy for Rician Fading Channels, IEEE J. Select. Areas Commun., vol.7, pp.1347-1355, Dec. 1989. 68. P. Ho and J.H. Kim, On Pilot Symbol Assisted Detection of CPM Schemes Operating in Fast Fading Channels, IEEE Trans. Comm., vol. 44, pp. 337-347, March 1996. 69. H.J. Scudder, Adaptive Communication Receivers, IEEE Trans. Inf. Theory, vol.11, pp.167-174, April 1965. 70. R. Haeb and H. Meyr, A Systematic Approach to Carrier Recovery and Detection of Digitally Phase Modulated Signals on Fading Channels, IEEE Trans. Comm., vol.37, pp.748-754, July 1989. 71. J.H. Lodge and M.J. Moher, Maximum Likelihood Sequence Estimation of CPM Signals Transmitted over Rayleigh Flat-Fading Channels, IEEE Trans. Comm., vol.38, pp.787-794, June 1990.

41

72. C.W. Therrien, Discrete Random Signals and Statistical Signal Processing, Englewood Cliffs, Prentice Hall, 1992. 73. S. Haykin, Adaptive Filter Theory, Englewood Cliffs, NJ: Prentice-Hall, 1986. 74. H.S. Wang and P.-C. Chang, On Verifying the First Order Markovian Assumption for a Rayleigh Fading Channel Model, IEEE Trans. Veh. Tech., vol. 45, pp. 353-357, May 1996. 75. R.E. Morley and D.L. Snyder, Maximum Likelihood Sequence Estimation for Randomly Dispersive Channels, IEEE Trans. Comm., vol. 27, pp. 833-839, June 1979. 76. F.C. Schweppe, Evaluation of Likelihood Functions for Gaussian Signals, IRE Trans. Inf. Theory, vol. 11, pp. 61-70, January 1965. 77. X. Yu and S. Pasupathy, Innovations-Based MLSE for Rayleigh Fading Channels, IEEE Trans. Comm., vol. 43, pp. 1534-1544, Feb./Mar./Apr. 1995. 78. J.H. Painter and S.C. Gupta, Recursive Ideal Observer Detection of Known Mary Signals in Multiplicative and Additive Gaussian Noise, IEEE Trans. Comm., vol. 21, pp. 948-953, August 1973. 79. J.H. Painter and L.R. Wilson, Simulation Results for the Decision-Directed MAP Receiver for M-ary Signals in Multiplicative and Additive Gaussian Noise, IEEE Trans. Comm., vol. 22, pp. 649-660, May 1974. 80. A. Aghamohammadi, H. Meyr and G. Ascheid, Adaptive Synchronization and Channel Parameter Estimation Using an Extended Kalman Filter, IEEE Trans. Comm., vol. 37, pp. 1212-1219, November 1989. 81. P.K. Varshney and A.H. Haddad, A Receiver with Memory for Fading Channels, IEEE Trans. Comm., vol. 26, pp. 278-283, Feb. 1978. 82. J.P. Seymour and M.P. Fitz, Near Optimal Symbol-by-Symbol Detection Schemes for Flat Rayleigh Fading, IEEE Trans. Comm., vol. 43, pp. 15251533, Feb./Mar./Apr. 1995. 83. J.P. Seymour and M.P. Fitz, Two-Stage Carrier Synchronization Techniques for Non-Selective Fading, IEEE Trans. Veh. Tech., vol. 44, pp. 103-110, Feb. 1995. 84. L. Bin, Decision Feedback Detection of Minimum Shift Keying, IEEE Trans. Comm., vol. 44, pp. 1073-1076, Sept. 1996. 85. R. J. Young and J. H. Lodge, Detection of CPM Signals in Fast Rayleigh FlatFading Using Adaptive Channel Estimation, IEEE Trans. Veh. Tech., vol. 44, pp. 338-347, May 1995. 86. G.M. Vitetta and D.P. Taylor, Maximum Likelihood Decoding of Uncoded and Coded PSK Signal Sequences Transmitted over Rayleigh Flat-Fading Channels, IEEE Trans. Comm., vol. 43, pp. 2750-2758, Nov. 1995. 87. R. Reggiannini, Coherent Reception of Generalized MSK over Rayleigh FlatFading Channels, IEEE Trans. Comm., vol. 42, pp. 410-416, Feb./Mar./Apr. 1994. 88. D. Divsalar and M.K. Simon, Maximum Likelihood Differential Detection of Uncoded and Trellis Coded Amplitude Phase Modulation over AWGN and

42

89.

90.

91.

92.

93.

94.

95. 96. 97.

98.

99.

100.

101.

102.

Fading Channels - Metrics and Performance, IEEE Trans. Comm., vol. 42, pp. 76-89, January 1994. M. Samiuddin and K.H. Biyari, A Comparative Study of Higher-Order Differential Phase Shift Keying Schemes over AWGN and Rayleigh Fading Channels, Int. Journ. of Wireless Inform. Networks, vol. 2, pp. 183-196, 1995. P. Ho and D. Fung, Error Performance of Multiple-Symbol Differential Detection of PSK Signals Transmitted Over Correlated Rayleigh Fading Channels, IEEE Trans. Comm., vol. 40, pp. 25-29, October 1992. K.H. Chang, W.S. Yuan and C.N. Gheorgiades, Block-by-block Channel and Sequence Estimation for ISI/Fading Channels, Proc. 7th Tyrrhenian Workshop on Digital Communications, pp. 153-170, September 1995. J.C. Han and C.N. Georghiades, Sequence Estimation in the Presence of Random Parameters Via the EM Algorithm, IEEE Trans. Comm., vol. 45, pp. 300-308, Mar.1997. J.C. Han and C.N. Georghiades, Pilot Symbol Initiated Optimal Decoder for the Land Mobile Fading Channel, Conf. Rec. Globecom95, pp. 42-47, Singapore, Nov. 1995. I. Korn, Binary PRCPM with Differential Phase Detection and Maximal Ratio Combining Diversity in Satellite Mobile, Land Mobile and Gaussian Channels, Proc. ICC96, Dallas, TX, pp. 916-920, May 1996. I. Korn, Error Floors in the Satellite and Mobile Channels, IEEE Trans. Comm., vol. 39, pp. 833-837, June 1991. T.K. Moon, The Expectation-Maximization Algorithm, IEEE Sig. Proc. Mag., vol. 13, pp. 47-63, Nov. 1996. D. Boundreau and J.H. Lodge, Adaptive Equalization of CPM Signals Transmitted over Fast Rayleigh Flat-Fading Channels, IEEE Trans. Veh. Tech., vol. 44, pp. 404-413, August 1995. G.M. Vitetta and D.P. Taylor, Maximum Likelihood Sequence Estimation of Differentially Encoded PSK Signals Transmitted over Rayleigh Frequency-Flat Fading Channels, Int. J. of Wireless Inf. Networks, Plenum Press, vol. 2, pp. 7181, April 1995. G.M. Vitetta and D.P. Taylor, Viterbi Decoding of Differentially Encoded PSK Signals Transmitted over Rayleigh Frequency-Flat Fading Channels, IEEE Trans. Comm., vol. 43, pp. 1256-1259, Feb./Mar./Apr. 1995. G.M. Vitetta, D.P. Taylor and U. Mengali, Double-Filtering Receivers for PSK Signals Transmitted over Rayleigh Frequency-Flat Fading Channels, IEEE Trans. Comm., vol. 44, pp. 686-695, June 1996 A.N. DAndrea, A. Diglio and U. Mengali, Symbol-Aided Channel Estimation with Nonselective Rayleigh Fading Channels, IEEE Trans. Veh. Tech., vol. 44, pp. 41-49, January 1995. M.V. Eyuboglu and S.U.H. Qureshi, Reduced-State Sequence Estimation with Set Partitioning and Decision Feedback, IEEE Trans. Comm., vol. 36, pp. 1320, Jan.1988.

43

103. M.V. Eyuboglu and S.U.H. Qureshi, Reduced-State Sequence Estimation for Coded Modulation on Intersymbol Interference Channels, IEEE Journ. on Sel. Areas on Comm., vol. 7, pp. 989-995, Aug. 1989. 104. A.D. Hallen and C. Heegard, Delayed Decision-Feedback Sequence Estimation, IEEE Trans. Comm., vol. 37, pp. 428-436, May 1989. 105. G.D. Forney Jr., Maximum-Likelihood Sequence Estimation of Digital Sequences in the Presence of Intersymbol Interference, IEEE Trans. Inf. Theory, vol. 18, pp. 363-378, May 1971. 106. R. Reheel, A. Polydoros and Ching-Kae Tzou, Per-Survivor Processing: a General Approach to MLSE in Uncertain Environments, IEEE Trans. Comm., vol. 43, pp. 354-364, Feb./Mar./Apr. 1995. 107. C.-E W. Sundberg and N. Seshadri, Coded Modulation for Fading Channels: An Overview, European Trans. on Telecomm. vol. 4, pp. 309-324, May-June 1993. 108. K. Yu an P. Ho, Trellis Coded Modulation with Multiple Symbol Differential Detection, Conf. Rec. ICC93, pp. 1414-1418, Geneva, Switzerland,1993. 109. G. Ungerboeck, Channel Coding with Multilevel/Phase Signals, IEEE Trans. Inf. Theory, vol. 28, pp. 55-67, January 1982. 110. E. Biglieri, D. Divsalar, P.J. McLane and M.K. Simon, Introduction to TrellisCoded Modulation with Applications, New York: MacMillan, 1991. 111. C. Schlegel and D.J. Costello, Jr., Bandwidth Efficient Coding for Fading Channels: Code Construction and Performance Analysis, IEEE J. Sel. Areas Comm., vol. 7, pp. 1356-1368, December 1989. 112. P.J. McLane, P.H. Whittke, P.K. Ho and C. Loo, PSK and DPSK Trellis Codes for Fast Fading, Shadowed Mobile Satellite Communication Channels, IEEE Trans. Comm., vol. 36, pp. 1242-1246, November 1988. 113. P. Ho and D.K.P. Fung, Error Performance of Interleaved Trellis-Coded PSK modulations in Correlated Rayleigh Fading Channels, IEEE Trans. Comm., vol. 40, pp. 1800-1809, Dec. 1992. 114. J.K. Cavers and P. Ho, Analysis of the Error Performance of Trellis-Coded Modulations in Rayleigh Fading Channels, IEEE Trans. Comm., vol. 40, pp. 74-80, Jan. 1992. 115. K. Chan and A. Bateman, The Performance of Reference-Based M-ary PSK with Trellis Coded Modulation in Rayleigh Fading, IEEE Trans. Veh. Tech., vol. 41, pp. 190-198, May 1992. 116. A. Aghamohammadi, H. Meyr and G. Ascheid, A New Method for Phase Synchronization and Automatic Gain control of Linearly Modulated Signals on Frequency-Flat Fading Channels, IEEE Trans. Comm., vol. 39, pp. 25-29, Jan. 1991. 117. G.T. Irvine and P.J. McLane, Symbol-Aided Plus Decision-Directed Reception for PSK/TCM Modulation on Shadowed Mobile Satellite Fading Channels, IEEE J. Sel. Areas Comm., vol. 10, pp. 1289-1299, Oct. 1992.

44

118. P.Y. Kam and H.M. Ching, Sequence Estimation over the Slow Nonselective Rayleigh Fading Channel with Diversity Reception and its Application to Viterbi Decoding, IEEE J. Sel. Areas Comm., vol. 10, pp. 562-570, April 1992. 119. Y. Liu and S.D. Blostein, Identification of Frequency Non-selective Fading Channels Using Decision Feedback and Adaptive Linear Prediction, IEEE Trans. Comm., vol. 43, pp. 1484-1492, Feb./Mar./Apr. 1995. 120. F. Edbauer, Performance of Interleaved Trellis-Coded Differential 8-PSK Modulation Over Fading Channels, IEEE J. Sel. Areas Comm., vol. 7, pp. 13401346, Dec. 1989. 121. R.W. Kerr and P.J. McLane, Coherent Detection of Interleaved Trellis Encoded CPFSK on Shadowed Mobile Satellite Channels, IEEE Trans. Veh. Tech., vol. 41, pp. 159-169, May 1992. 122. J.W. Modestino and S.N. Hulyalkar, Cut-Off Rate Performance of CPM Schemes Operating on the Slow-Fading Rician Channel, Proc. 5th Tyrrhenian Workshop on Digital Comm., pp. 119-130, September 1991. 123. L. Yii and G.L. Stuber, Noncoherently Detected Trellis-Coded Partial Response CPM on Mobile Radio Channels, IEEE Trans. Comm., vol. 44, pp. 967-975, Aug. 1996. 124. J. Cavers, On the Validity of the Slow and Moderate Fading Models for Matched Filter Detection of Rayleigh Fading Signals, Can. J. of Elect. & Comp. Eng., vol. 17, pp. 183-189, 1992. 125. R.S. Kennedy, Fading Dispersive Communication Channels, New York: Wiley-Interscience, 1969. 126. G.M. Vitetta and D.P. Taylor, Multi-Sampling Receivers for Uncoded and Coded PSK Signal Sequences Transmitted Over Rayleigh Frequency-Flat Fading Channels, IEEE Trans. Comm., vol. 44, pp. 130-133, February 1996. 127. W.C. Dam and D.P. Taylor, An Adaptive Maximum Likelihood Receiver for Correlated Rayleigh-Fading Channels, IEEE Trans. Comm., vol. 42, pp. 26842692, Sept. 1994. 128. L. Andriot, G. Tziritas and G. Jourdain, Discrete Realization for Receivers. Detecting Signals over Random Dispersive Channels. Part II: Doppler-Spread Channel, Signal Processing, North Holland, vol. 9, pp. 89-100, Sept. 1985. 129. G.M. Vitetta, U. Mengali and D.P. Taylor, A Blind Detection Technique for CPM Signals Transmitted over Frequency-Flat Fading Channels, Conf. Rec. Comm. Th. Mini-Conference, IEEE Globecom'96, London, November 1996. 130. G.M. Vitetta, U. Mengali and D.P. Taylor, Double-Filter Differential Detection of PSK Signals Transmitted over Linearly Time-Selective Rayleigh Fading Channels, Submitted to ICC98, Aug. 1997. 131. G.M. Vitetta, U. Mengali and D.P. Taylor, Noncoherent Detection of FSK Signals Transmitted over Linearly Time-Selective Rayleigh Fading Channels, Conf. Rec. IEEE ICC'97, Montreal, Canada, June 1997.

45

132. G.L.Turin, Introduction to Spread-Spectrum Antimultipath Techniques and their Application to Urban Digital Radio, Proc. IEEE, vol. 68, pp. 328-353, Feb. 1981. 133. M. Schwartz, W.R. Bennett, S. Stein, Communication Systems and Techniques, New York: McGraw-Hill, 1966. 134. P.A. Bello, Characterization of Randomly Time-Variant Linear Channels, IEEE Trans. Comm. Sys., vol. 9, pp. 360-393, Dec. 1963. 135. G. Ungerboeck, Adaptive Maximum-Likelihood Receiver for CarrierModulated Data-Transmission Systems, IEEE Trans. Comm., vol. 22, pp. 624636, May 1974. 136. G.E. Bottomley and S. Chennakeshu, Adaptive MLSE Equalization Forms for Wireless Communications, Fifth Virginia Tech. Symp. on Wireless Personal Comm., Blacksburg VA, May 31-June 2, 1995, pp. 183-194. 137. B.D. Hart and D.P. Taylor, Extended MLSE Receiver for the Frequency-Flat, Fast Fading Channel, IEEE Trans. Veh. Tech., vol. 46, pp. 381-389, May 1997. 138. B.D. Hart and D.P. Taylor, Extended MLSE Diversity Receiver for the Timeand Frequency-Selective Channel, IEEE Trans. Comm., vol. 45, pp. 322-333, March 1997. 139. B.D. Hart and D.P. Taylor, A Joint MLSE Receiver for Unknown, Unsynchronised, Time-Varying, Frequency-Selective, Rician Channels, To appear in IEEE Trans. Comm. 140. H.Kubo, K.Murakami and T.Fujino, An Adaptive Maximum-Likelihood Sequence Estimator for Fast Time-Varying Intersymbol Interference Channels, IEEE Trans. Comm., vol. 42, , pp. 1972-1880, Feb./Mar./Apr. 1994. 141. N.Seshadri Joint Data and Channel Estimation using Blind Trellis Search Techniques, IEEE Trans. Comm., vol. 42, Feb./Mar./Apr. 1994, pp. 1000-1011. 142. K.M. Chugg and A. Polydoros, MLSE for an Unknown Channel - Part I: Optimality Considerations, IEEE Trans. Comm., vol. 44, pp. 836-846, July 1996. 143. T.E. Duncan, Evaluation of Likelihood Functions, Information & Control, vol. 13, pp. 62-74, July 1968. 144. F.R. Magee, Jr. and J.G. Proakis, Adaptive Maximum-Likelihood Sequence Estimation for Digital Signalling in the Presence of Intersymbol Interference, IEEE Trans. Inf. Theory, vol. 19, pp. 120-124, Jan. 1973. 145. S.U.H. Qureshi and E.E. Newhall, An Adaptive Receiver for Data Transmission over Time-Dispersive Channels, IEEE Trans. Inf. Theory, vol. 19, Jul. 1973, pp. 448-457 146. K.M. Chugg and A. Polydoros, MLSE for an Unknown Channel -Part II: Tracking Performance, IEEE Trans. Comm., vol. 44, Aug. 1996, pp. 949-958. 147. D. Middleton, On the Detection of Stochastic Signals in Additive Normal Noise - Part I, IRE Trans. Inf. Theory, vol. 3, pp. 86-121, June. 1957.

46

148. Q. Dai and E. Shwedyk, Detection of Bandlimited Signals over FrequencySelective Rayleigh Fading Channels, IEEE Trans. Comm., vol. 42, pp. 941-950, 1994. 149. R. Raheli, G. Marino and P. Castoldi, Per-Survivor Processing and Tentative Decisions: What is in Between?, IEEE Trans. Comm., vol. 44, pp. 127-129, Feb. 1996. 150. A.P. Clark, S.N. Abdullah, S.G. Jaysinghe and K.H. Sun, Pseudobinary and Pseudoquaternary Detection Processes for Linearly Distorted Multilevel QAM Signals, IEEE Trans. Comm., vol. 44, pp. 127-129, Feb. 1988. 151. G. Deng, J. Cavers and P. Ho, A Reduced Dimensionality Propagation Model for Frequency-Selective Rayleigh Fading Channels, Conf. Rec. ICC95, pp. 1158-1162, June 1995. 152. R.H. Clarke, A Statistical Theory of Mobile-Radio Reception, Bell Syst. Tech. J., vol 47, July-Aug. 1968, pp. 957-1000. 153. A.B. Marcovitz, On Inverses and Quasi-Inverses of linear Time-Varying Discrete Systems, J. Franklin Inst., vol. 272, pp. 23-44. 154. M. Nakagami, The m-distribution: A General of Intensity Distribution of Rapid Fading, in Statistical Methods of Radio Wave Propagation, Pergamon Press, 1960. 155. B. Widrow, J. McCool, M. Larimore and C. Johnson, Stationary and Nonstationary Learning Characteristics of the LMS Adaptive Filter, Proc. IEEE, vol. 64, pp. 1151-1162, Aug. 1976. 156. W. Gardner, Learning Characteristics of Stochastic-Gradient-Descent Algorithms: A General Study, Analysis and Critique, Signal Processing, vol. 6, pp. 112-133, Apr. 1984. 157. L.C. Barbosa, Maximum Likelihood Sequence Estimators: A Geometric View, IEEE Trans. Inf. Theory, vol. 35, pp. 419-427, Mar. 1989. 158. W.U. Lee and F.S. Hill, , A Maximum-Likelihood Sequence Estimator with Decision-Feedback Equalisation, IEEE Trans. Comm., vol. 25, pp. 971-979, Sept. 1977. 159. D.D. Falconer and F.R. Magee, Adaptive Channel Memory Truncation for Maximum-Likelihood Sequence Estimation, Bell Syst. Tech. J., vol. 52, pp. 1541-1562, Nov. 1973 160. B.D. Hart and D.P. Taylor, On the Irreducible Error Floor in Fast Fading channels, Submitted IEEE Trans. Veh. Tech., Apr. 1997. 161. Y. Gu and T. Le-Ngoc, Adaptive Combined DFE/MLSE Techniques for ISI Channels, IEEE Trans. Comm., vol. 44, pp. 847-857, July 1996. 162. J.G. Proakis and J.H. Miller, An Adaptive Receiver for Digital Signalling Through Channels with Intersymbol Interference, IEEE Trans. Inf. Theory, vol. 15, July 1969, pp. 484-497. 163. J.-P. Linnartz, Narrowband Land Mobile Radio Networks, Artech House, Norwood, MA, 1993.

47

164. W.S. Leon and D.P. Taylor, MLSE Receiver for the Dispersive Rayleigh Fading Channel, Conf. Rec. ICC97, 8-12 June 1997, Montreal. 165. J.H. Kim, P. Ho and P.A. Goud, Performance of Binary CPM Schemes in Frequency-Selective Fading Channels with Pilot Symbol Assisted Detection, Conf. Rec. Globecom95, Nov. 1995, Singapore, pp. 2027-2032. 166. P. Ho, G. Deng and J. Cavers, A Reduced Complexity Channel Estimator for Linear Modulations Operating in Fading Dispersive Channels, Globecom95, Nov. 1995, Singapore, pp. 2043-2048. 167. S.H. Leung and B.L. Chan, A Fast Adaptive Equaliser Structure for Voiceband Communication Link, Proc. 33rd Midwest Symp. on Circuits and Systems, Calgary, Aug. 1990, pp. 5-8. 168. A. Gatherer and T.H.-Y. Meng, High Sampling Rate Adaptive Decision Feedback Equalisers, ICASSP, Apr. 1990, Albuquerque, vol. 2, pp. 909-912. 169. M.E. Rollins and S.J. Simmons, Error Performance Analysis of MLSE for Frequency-Selective Rayleigh Fading Channels with Kalman Channel Estimation, Supercom/ICC94, May 1994, New Orleans, pp. 321-325. 170. D. Divsalar and M.K. Simon, Multiple-Symbol Differential Detection of MPSK, IEEE Trans. Comm., vol. 38, pp. 300-308, Mar. 1990. 171. D. Makrakis, P.T. Mathiopoulos and D.P. Bouras, Optimal Decoding of Coded PSK and QAM Signals in Correlated Fast Fading Channels and AWGN: a Combined Envelope, Multiple Differential and Coherent Detection Approach, IEEE Trans. Comm., vol. 42, pp. 63-74, Jan. 1994. 172. D.P. Bouras, P.T. Mathiopoulos and D. Makrakis, Optimal Detection of Coded Differentially Encoded PSK and QAM Signals with Diversity in Correlated Fast Fading Channels, IEEE Trans. Veh. Tech., vol. 42, pp. 245-257, Aug. 1993. 173. C.A. Belfiore and J.H. Park, Jr., Decision Feedback Equalization, Proc. IEEE, vol. 67, pp. 1143-1156, Aug. 1979. 174. P. Monsen, Feedback Equalization for Fading Dispersive Channels, IEEE Trans. Inf. Theory, vol. 17, pp. 56-64, Jan. 1971. 175. P. Monsen, Adaptive Equalization of the Slow Fading Channel, IEEE Trans. Comm., vol. 22, pp. 1064-1075, Aug. 1974. 176. E. Eleftheriou, Adaptive Equalization Techniques for HF Channels, IEEE J. Sel. Areas. Comm., vol. 5, pp. 238-247, Feb. 1987. 177. S.H. Jamali and T. Le-Ngoc, Coded Modulation Techniques for Fading Channels, Kluwer Academic Publishers, Boston MA, 1994. 178. R.D. Gitlin and S.B. Weinstein, Fractionally-Spaced Equalization: An Improved Digital Transversal Equalizer, Bell Syst. Tech. J., vol. 60, pp. 856864, Feb. 1981. 179. E. Eleftheriou and D.D. Falconer, Tracking Properties and Steady-State Performance of RLS Adaptive Filter Algorithms, IEEE Trans. Acoust. Speech and Signal Proc., vol. 34, pp. 1097-1109, Oct. 1986.

48

180. N.W.K. Lo, D.D. Falconer and U.H. Sheikh, Adaptive Equalization and Diversity Combining for Mobile Radio using Interpolated Channel Estimates, IEEE Trans. Veh. Tech., vol. 40, pp. 636-645, Aug. 1991. 181. J.C.S. Cheung and R. Steele, Soft-Decision Feedback Equalizer for Continuous Phase Modulated Signals in Wideband Mobile Radio Channels, IEEE Trans. Comm., vol. 42, pp. 1628-1638, Feb./Mar./Apr. 1994. 182. R. DAvella, L. Moreno and M. SantAgostino, An Adaptive MLSE Receiver for TDMA Digital Mobile Radio, IEEE J. Sel. Areas Comm., vol. 7, pp. 122129, Jan. 1989. 183. T. Kailath, Measurements on Time-Variant Communication Channels, IRE Trans. Inf. Theory, vol. 8, pp. S229-S236, Sept. 1962. 184. T. Kailath, Time Variant Communication Channels, IRE Trans. Inf. Theory, vol. 9, pp. 233-237, October 1963. 185. P.A. Bello, Measurement of Random Time-Variant Channels, IEEE Trans. Inf. Theory, vol. 15, pp. 469-475, July 1969. 186. P.A. Bello and R. Esposito, Measurement Techniques for Time-Varying Dispersive Channels, Alta Frequenza, vol. XXXIX, pp. 980-996, Nov. 1970. 187. A. Baier and D.G. Heinrich, Performance of M Algorithm MLSE Equalizer in Frequency Selective Fading Mobile Radio Channels, Conf. Rec. ICC89, Boston, MA, pp. 281-285, June 1989. 188. J. Wu and A.H. Aghvami, A New Adaptive Equalizer with Channel Estimator for Mobile Radio Communications, IEEE Trans. Veh. Tech., vol. 45, pp. 467474, Aug. 1996. 189. Y.-J. Liu, M. Wallace and J.W. Ketchum, A Soft Output Bidirectional Decision Feedback Equalization Technique for TDMA Cellular Radio, IEEE J. Sel. Areas on Comm., vol. 11, pp. 1034-1045, Sept. 1993. 190. F. Ling and J.G. Proakis, Adaptive Lattice Decision-Feedback Equalizers, IEEE Trans. Comm., vol. 33, pp. 348-356, Apr. 1985. 191. M. Stojnanovic, J. Proakis and J. Catipovic, Analysis of the Impact of Channel Estimation Errors on the Performance of a Decision Feedback Equaliser in Fading Multipath Channels, IEEE Trans. Comm., vol. 43, pp. 877-885, Feb./March/Apr. 1995. 192. P. Monsen, Fading Channel Communications, IEEE Communications Magazine, pp. 16-25, Jan. 1980. 193. K. Giridar, J.J. Shynk, A.R. Iltis and A. Mahur, Adaptive MAPSD Algorithms for Symbol and Timing Recovery of Mobile Radio TDMA Signals, IEEE Trans. Comm., vol. 44, pp. 926-978, Aug. 1996. 194. J.G. Proakis, Adaptive Equalization for TDMA Digital Mobile Radio, IEEE Trans. Veh. Tech., vol. 40, pp. 333-341, May 1991. 195. S.N. Crozier, D.D. Falconer and S.A. Mahmoud, Reduced Complexity ShortBlock Data Detection Techniques for Fading Time-Dispersive Channels, IEEE Trans. Veh. Tech., vol. 41, pp. 255-265, Aug. 1992.

49

196. G.K. Kaleh, Channel Equalization for Block Transmission Systems, IEEE J. Sel. Areas Comm., vol. 13, pp. 110-121, Jan. 1995. 197. S.A. Fechtel and H. Meyr, Optimal Parametric Feedforward Estimation of Frequency-Selective Fading Radio Channels, IEEE Trans. Comm., vol. 42, pp. 1639-1650, Feb./Mar./Apr. 1994. 198. J.K. Cavers, Pilot Symbol Assisted Modulation and Differential Detection in Fading and Delay Spread, IEEE Trans. Comm., vol. 43, pp. 2206-2212, July 1995. 199. F.D. Garber and M.B. Pursley, Performance of Differentially Coherent Digital Communications over Frequency-Selective Fading Channels, IEEE Trans. Comm., vol. 36, pp. 21-31, Jan. 1988. 200. F.D. Garber and M.B. Pursley, Performance of Binary FSK Communications over Frequency-Selective Fading Channels, IEEE Trans. Comm., vol. 37, pp. 83-89, Jan 1989. 201. U. Charash, Reception through Nakagami Fading Multipath Channels with Random Delays, IEEE Trans. Comm., vol. 43, pp. 1937-1945, May 1995. 202. S. Cheng, S. McLaughlin, B. Mulgrew and P.M. Grant, Adaptive Bayesian Decision Feedback Equalizer for Dispersive Mobile Radio Channels, IEEE Trans. Comm., vol. 43, pp. 1937-1945, May 1995. 203. M.-C. Chiu and C.-C. Chao, Analysis of LMS-Adaptive MLSE Equalization on Multipath Fading Channels, IEEE Trans. Comm., vol. 44, pp. 1684-1692, Dec. 1996. 204. D. Williamson, R.A. Kennedy and G.W. Pulford, Block Decision Feedback Equalization, IEEE Trans. Comm., vol. 40, pp. 255-264, Feb. 1992. 205. S. Chen, B. Mulgrew and S. McLaughlin, Adaptive Bayesian Equalizer with Decision Feedback, IEEE Sig. Proc., vol. 41, pp. 2918-2926, Sept. 1993. 206. W.-H. Sheen and G.L. Stuber, MLSE Equalization and Decoding for Multipath Fading Channels, IEEE Trans. Comm., vol. 39, pp. 1455-1464, Oct. 1991. 207. A.P. Clark and S. Harihanan, Adaptive Channel Estimator for an HF Radio Link, IEEE Trans. Comm., vol. 37, pp. 918-926, Sept. 1989. 208. R.A. Ziegler and J.M. Cioffi, Estimation of Time-Varying Digital Radio Channels, IEEE Trans. Veh. Tech., vol. 41, pp. 134-151, May 1992. 209. E. Dahlman, New Adaptive Viterbi Detector for Fast-Fading Mobile Radio Channels, Electron. Lett., vol. 26, pp. 1572-1573, Sept. 1990. 210. R.A. Goubran, H.M. Hafez and A.U.H. Sheikh, Implementation of a Real-Time Mobile Channel Simulator using a DSP Chip, IEEE Trans. Instr. And Meas., vol. 40, pp. 709-714, Aug. 1991. 211. COST 207, Digital Land Mobile Radio Communications, Annex 1, Commission of the European Communities, 1989. 212. H. Takai, BER Performance of Anti-Multipath Modulation Scheme PSK-VP and its Optimum Phase-Waveform, IEEE Trans. Veh. Tech., vol. 42, Nov. 1993, pp. 625-640.

50

213. S. Ariyavisitakul, S. Yoshida, F. Ikegami and T. Takeuchi, A Novel AntiMultipath Modulation Technique DSK, IEEE Trans. Comm., vol. 35, Dec. 1987, pp. 1252-1264. 214. G. Ungerboeck, Fractional Tap-Spacing Equaliser and Consequences for Clock Recovery in Data Modems, IEEE Trans. Comm., vol. 24, pp. 856-864, Aug. 1976. 215. J.M. Perl, A. Shpigel and A. Reichman, Adaptive Receiver for Digital Communication over HF Channels, IEEE J. Sel. Areas Comm., vol. 5, pp. 304308, Feb. 1987. 216. A. Wautier, J.-C. Dany, C. Mourot and V. Kumar, A New Method for Predicting the Channel Estimate Influence on Performance of TDMA Mobile Radio Systems, IEEE Trans. Veh. Tech., vol. 44, pp. 594-602, Aug. 1995. 217. G. DAria, R. Piermarini and V. Zingarelli, Fast Adaptive Equalisers for Narrow-Band TDMA Mobile Radio, IEEE Trans. Veh. Tech., vol. 40, pp. 392404, May 1991. 218. G. Carayannis, D.G. Manolakis and N. Kalouptsidis, A Fast Sequential Algorithm for Least-Squares Filtering and Prediction, IEEE Ac., Sp., & Sig. Proc., vol. 31, pp. 1394-1402, Dec. 1983. 219. D.D. Falconer and L. Ljung, Application of Fast Kalman Estimation to Adaptive Equalisation, IEEE Trans. Comm., vol. 26, pp. 1439-1446, Oct. 1978,. 220. E.H. Satorius and J.D. Pack, Application of Least Squares Lattice Algorithms to Adaptive Equalisation, IEEE Trans. Comm., vol. 136, pp. 136-142, Feb. 1981. 221. J.A. Bingham, Improved Methods of Accelerating the Convergence of Adaptive Equalisers for Partial-Response Signals, IEEE Trans. Comm., vol. 35, pp. 277-260, Mar. 1987. 222. B. Toplis and S. Pasupathy, Tracking Improvements in Fast RLS Algorithms using a Variable Forgetting Factor, IEEE Ac., Sp., & Sig. Proc., vol. 36, pp. 206-227, Feb. 1988. 223. R.A. Valenzuela, Performance of Adaptive Equalisation for Indoor Radio Communications, IEEE Trans. Comm., vol. 37, pp. 291-293, Mar. 1989. 224. B. Baccetti, S. Bellini, G. Filiberti and G. Tartara, Full Digital Adaptive Equalisation in 64-QAM Radio Systems, IEEE J. Sel. Areas Comm., vol. 5, pp. 466- 475, Apr. 1987. 225. F. Ling and S.U.H. Qureshi, Convergence and Steady State Behaviour of a Phase-Splitting Fractionally Spaced Equaliser, IEEE Trans. Comm., vol. 38, pp. 418-425, Apr. 1990. 226. M. Barton and D.W. Tufts, A Suboptimum Linear Receiver Based on a Parametric Channel Model, IEEE Trans. Comm., vol. 39, pp. 1328-1334, Sep. 1991. 227. S.B. Gelfand, C.S. Ravishankar and E.J. Delp, Tree-Structured Piecewise Linear Adaptive Equalisation, IEEE Trans. Comm., vol. 41, pp. 70-82, Jan. 1993.

51

228. R. Agusti and F. Cassadevall, Performance of Fractioned and Nonfractioned Equalisers with High-Level QAM, IEEE J. Sel. Areas Comm., vol. 5, pp. 476483, Apr. 1987. 229. O. Andrisano, G. Bianconi and L. Calandrino, Adaptive Equalisation of High Capacity M-QAM Radio Systems on Multipath Fading Channels, IEEE J. Sel. Areas Comm., vol. 5, pp. 457-465, Apr. 1987. 230. G.J. Foschini, Performance Bound for Maximum Likelihood Sequence Estimation of Digital Sequence in the Presence of Intersymbol Interference, IEEE Trans. Inf. Theory, vol. 21, pp. 47-50, Jan. 1975. 231. S.Verdu, Maximum Likelihood Sequence Detection for Intersymbol Interference Channels: a New Upper Bound on Error Probability, IEEE Trans. Inf. Theory, vol. 33, pp. 62-68, Jan. 1987. 232. J.E. Mazo, Exact Matched Filter Bound for Two-Beam Rayleigh Fading, IEEE Trans. Comm., vol. 39, pp. 1027-1030, Jul. 1991. 233. E.J. Baghdady, Novel Techniques for Counteracting Multipath Interference Effects in Receiving Systems, IEEE J. Sel. Areas Comm., vol. 5, pp. 274-285, Feb. 1987. 234. R. Price, P.E. Green, A Communication Technique for Multipath Channels, Proc. IRE, vol. 46, pp. 555-570, Mar. 1958. 235. M. Alles, S. Pasupathy, Channel Knowledge and Optimal Performance for Two-Wave Rayleigh Fading Channels, IEEE Trans. Veh. Tech., vol. 43, pp. 820, Feb. 1994. 236. M.-C. Chiu and C.-c. Chao, Performance of Joint Equalisation and TrellisCoded Modulation on Multipath Fading Channels, IEEE Trans. Comm., vol. 43, no. 2/3/4, pp.1230-1234, Feb./Mar./Apr. 1995. 237. A. Wautier, J.-C. Dany, C. Mourot and V. Kumar, A New Method for Predicting the Channel Estimate Influence on Performance of TDMA Mobile Radio Systems, IEEE Trans. Veh. Tech., vol. 44, pp. 594-602, Aug. 1995. 238. D.-L. Liu and K. Feher, Pilot-Symbol Aided Coherent M-ary PSK in Frequency-Selective Fast Rayleigh Fading Channel, IEEE Trans. Comm., vol. 42, pp. 54-62, Jan. 1994. 239. T. Eyceoz and A. Duel-Hallen, Simplified Block Adaptive Diversity Equaliser for Cellular Mobile Radio, IEEE Comm Lett., vol. 1, pp15-19, Jan. 1987. 240. K. Pahlavan and J.W. Matthews, Performance of Adaptive Matched Filter Receivers over Fading Multipath Channels, IEEE Trans. Comm., vol. 38, pp. 2106-2113, Dec. 1990. 241. G. Castellini, E.D. Re and L. Perucci, A Continuously Adaptive MLSE Receiver for Mobile Communications: Algorithm and Performance, IEEE Trans. Comm., vol. 45, pp. 80-89, Jan. 1997. 242. G. DAria, Simulation and Performance of the Pan-European Land Mobile Radio System, IEEE Trans. Veh. Tech., vol. 41, pp. 177-189, May 1992.

52

243. E. Katz and G.L. Stuber, Sequential Sequence Estimation for Trellis-Coded Modulation on Multipath Fading ISI Channels, IEEE Trans. Comm., vol. 43, pp. 2883-2885, Dec. 1995. 244. O. Macchi, Optimisation of Adaptive Identification for Time-Varying Filters, IEEE Trans. Auto. Contr., vol. 31, pp. 283-287, Mar. 1996. 245. J.W.M. Bergmans, Tracking Capabilities of the LMS Adaptive Filter in the Presence of Gain Variation, IEEE Ac., Sp., & Sig. Proc., vol. 38, pp. 712-714, Apr. 1990. 246. M. Niedzwiecki, First-Order Tracking Properties of Weighted Least Squares Estimation, IEEE Trans. Auto. Contr., vol. 33, pp. 94-98, Jan. 1988. 247. F. Ling and J.G. Proakis, Nonstationary Learning Characteristics of Least Squares Adaptive Estimation Algorithms, Proc. IEEE ICASSP84, San Diego, Mar. 1984. 248. J. Lin, J.G. Proakis, F. Ling and H. Lev-ari, Optimal Tracking of Time-Varying Channels: A Frequency Domain Approach for Known and New Algorithms, IEEE J. Sel. Areas Comm., vol. 13, pp.141-154, Jan. 1995. 249. B. Toplis and S. Pasupathy, Tracking Improvements in Fast RLS Algorithms using a Variable Forgetting Factor, IEEE Ac., Sp., & Sig. Proc., vol. 36, pp. 206-227, Feb. 1988. 250. S. Haykin, A.H. Sayed, J.R. Zeidler, P. Yee and P.C. Wei, Adaptive Tracking of Linear Time-Variant Systems by Extended RLS Algorithms, IEEE Sig. Proc., vol. 45, pp.1118-1127, May 1997. 251. K.A. Hamied and G.L. Stuber, Performance of Trellis-Coded Modulation for Equalised Multipath Fading ISI Channels, IEEE Trans. Veh. Tech., vol. 44, pp. 50-58, Feb. 1995. 252. D.D. Falconer, A.U.H. Sheikh, E. Eleftheriou and M. Tobis, Comparison of DFE and MLSE Receiver Performance on HF Channels, IEEE Trans. Comm., vol. 33, pp. 484-486, May 1985. 253. R.A. Iltis, A Bayesian Maximum-Likelihood Sequence Estimation Algorithm for a priori Unknown Channels and Symbol Timing, IEEE J. Sel. Areas Comm., vol. 10, pp. 579-588, Apr. 1992. 254. M. Erkurt and J.G. Proakis, Joint Data Detection and Channel Estimation for Rapidly Fading Channels, Proc. IEEE Globecom Conf., Orlando, Dec. 1992. 255. W.P. Chou and P.J. McLane, 16-State Nonlinear Equaliser for IS-54 Digital Cellular Channels, IEEE Trans. Veh. Tech., vol. 45, pp.13-25, Feb. 1996. 256. G. Benelli, A. Garzelli, F. Salvi, Simplified Viterbi Processors for the GSM Pan-European Cellular Communication System, IEEE Trans. Veh. Tech., vol. 43, pp. 870-878, Nov. 1994. 257. P.J. McLane, A Residual Intersymbol Interference Error Bound for TruncatedState Viterbi Detectors, IEEE Trans. Inf. Theory, vol. 26, pp. 549-553, Sep. 1980.

53

258. H.C. Guren and N. Holte, Decision Feedback Sequence Estimation for Continuous Phase Modulation on a Linear Multipath Channel, IEEE Trans. Comm., vol. 41, pp. 280-284, Feb. 1993. 259. S. Ariyavisitakul and L.J. Greenstein, Reduced-Complexity Equalisation Techniques for Broadband Wireless Channels, IEEE J. Sel. Areas Comm., vol. 15, pp. 5-15, Jan. 1997. 260. R.E. Kamel and Y. Bar-Ness, Reduced-Complexity Sequence Estimation Using State Partitioning, IEEE Trans. Comm., vol. 44, pp. 1057-1063, Sept. 1996. 261. W.-H. Sheen and G.L. Stuber, Error Probability for Maximum Likelihood Sequence Estimation of Trellis-Coded Modulation on ISI Channels, IEEE Trans. Comm., vol. 42, pp. 1427-1430, Feb./Mar./Apr. 1994. 262. C.T. Beare, The Choice of the Desired Impulse Response in Combined LinearViterbi Algorithm Equalisers, IEEE Trans. Comm., vol. 26, pp. 1301-1307, Aug. 1978. 263. K. Wesolowski, An Efficient DFE & ML Suboptimum Receiver for Data Transmission over Dispersive Channels using Two-Dimensional Signal Constellation, IEEE Trans. Comm., vol. 35, pp. 337-339, Mar. 1987. 264. J.B. Anderson, Sequential Coding Algorithms: A Survey and Cost Analysis, IEEE Trans. Comm., vol. 32, pp. 169-176, Feb. 1984. 265. F. Xiong, A. Zerik and E. Shwedyk, Sequential Sequence Estimation for Channels with Intersymbol Interference of Finite or Infinite Length, IEEE Trans. Comm., vol. 38, pp. 795-804, Jun. 1990. 266. S.J. Simmons, Breadth-First Trellis Decoding with Adaptive Effort, IEEE Trans. Comm., vol. 38, pp. 3-12, Jan. 1990. 267. P. Jung, Performance Evaluation of a Novel M-Detector for Coherent Receiver Antenna Diversity in a GSM-Type Mobile Radio System, IEEE J. Sel. Areas Comm., vol. 13, pp. 80-88, Jan. 1995. 268. R. Mehlan and H. Meyr, Soft Output M-Algorithm Equaliser and Trellis-Coded Modulation for Mobile Radio Channels, Proc. IEEE 42nd Veh. Tech. Conf., VTC-92, Denver, 1992, pp.582-591. 269. E. Katz and G.L. Stuber, Sequential Sequence Estimation for Trellis-Coded Modulation on Multipath Fading ISI Channels, IEEE Trans. Comm., vol. 43, pp. 2883-2885, Dec. 1995. 270. N. Benvenuto and R. Marchesani, The Viterbi Algorithm for Sparse Channels, IEEE Trans. Comm., vol. 44, pp. 287-289, Mar. 1996. 271. K. Abend, T.J. Hartley, B.D. Fritchman and C. Gumacos, On Optimum Receivers for Channels having Memory, IEEE Trans. Inf. Theory, vol. 14, pp. 152-157, Nov. 1968. 272. G.D. Forney, The Viterbi Algorithm, Proc. IEEE, vol. 61, pp. 268-278, Mar. 1973.

54

273. C. Berrou and A. Glavieux, Near Optimum Error Correcting Coding and Decoding: Turbo-Codes, IEEE Trans. Comm., vol. 44, pp. 1261-1271, Oct. 1996. 274. J. Erfanian, S. Pasupathy and G. Gulak, Reduced Complexity Symbol Detectors with Parallel Structures for ISI Channels, IEEE Trans. Comm., vol. 42, pp. 1661-1671, Feb./Mar./Apr. 1994. 275. K. Abend and B.D. Fritchman, Statistical Detection for Communication Channels with Intersymbol Interference, Proc. IEEE, vol. 58, pp. 779-785, May 1970. 276. J.F. Hayes, T.M. Cover and J.B. Riera, Optimal Sequence Detection and Optimal Symbol-by-Symbol Detection: Similar Algorithms, IEEE Trans. Comm., vol. 30, pp. 152-157, Jan. 1982. 277. R.E. Lawrence and H. Kaufman, The Kalman Filter for the Equalisation of a Digital Communications Channel, IEEE Trans. Comm. Tech., vol. 19, pp. 1127-1141, Dec. 1971. 278. J. Erfanian, S. Pasupathy and G. Gulak, Reduced Complexity Symbol Detectors with Parallel Structures for ISI Channels, IEEE Trans. Comm., vol. 42, pp. 5462, Feb./Mar./Apr. 1994. 279. R.A. Iltis, J.J. Shynk and K. Giridhar, Bayesian Algorithms for Blind Equalisation using Parallel Adaptive Filtering, IEEE Trans. Comm., vol. 42, pp. 1017-1032, Feb./Mar./Apr. 1994. 280. G.-K. Lee, S.B. Gelfand and M.P. Fitz, Bayesian Decision Feedback Techniques for Deconvolution, IEEE J. Sel. Areas Comm., vol. 13, pp. 155166, Jan. 1995. 281. G.J. Gibson, S. Siu and C.F.N. Cowan, The Application of Nonlinear Structures to the Reconstruction of Binary Signals, IEEE Trans. Sig. Proc., vol. 39, pp. 1877-1884, Aug. 1991. 282. Z.-J. Xiang and G.-G. Bi, A New Lattice Polynomial Perceptron and its Applications to Frequency-Selective Fading Channel Equalisation and ACI Suppression, IEEE Trans. Comm., vol. 44, pp. 761-767, Jul. 1996. 283. Z. Xiang, G. Bi and T. Le-Ngoc, Polynomial Perceptrons and their Applications to Fading Channel Equalisation and Co-channel Interference Suppression, IEEE Sig. Proc., vol. 32, pp. 2470-2480, Sep. 1994. 284. P.-R. Chang and B.-C. Wang, Adaptive Decision Feedback Equalisation for Digital Satellite Channels using Multilayer Neural Networks, IEEE J. Sel. Areas Comm., vol. 13, pp. 316-324, Feb. 1995. 285. I. Cha and S.A. Kassam, Channel Equalisation using Adaptive Complex Radial Basis Function Networks, IEEE J. Sel. Areas Comm., vol. 13, pp.122-131, Jan. 1995. 286. S. Chen and B. Mulgrew, A Clustering Technique for Digital Communications Channel Equalisation using Radial Basis Function Networks, IEEE Trans. Neural Net., vol. 4, pp. 570-579, Jul. 1993.

55

287. R.A. Kennedy, B.D.O. Anderson and R.R. Bitmead, Channels Leading to Rapid Error Recovery for Decision Feedback Equalisers, IEEE Trans. Comm., vol. 37, pp. 1126-1135, Nov. 1989. 288. R.A. Kennedy and B.D.O. Anderson, Error Recovery of Decision Feedback Equalisers on Exponential Impulse Response Channels, IEEE Trans. Comm., vol. 35, pp. 846-848, Aug. 1987. 289. R.A. Kennedy and B.D.O. Anderson, Recovery Times of Decision Feedback Equalisers on Noiseless Channels, IEEE Trans. Comm., vol. 35, pp. 1012-1021, Aug. 1987. 290. D.L. Duttweiler, J.E. Mazo and D.G. Messerschmitt, An Upper Bound on the Error Probability in Decision-Feedback Equalisation, IEEE Trans. Inf. Theory, vol. 20, pp. 490-497, Jul. 1974. 291. R.A. Kennedy and B.D.O. Anderson, Tight Bounds on the Error Probabilities of Decision Feedback Equalisers, IEEE Trans. Comm., vol. 35, pp. 1022-1028, Oct. 1987. 292. J.C. Cartledge, Outage Performance of QAM Digital Radio Using Adaptive Equalisation and Switched Spaced Diversity Reception, IEEE Trans. Comm., vol. 35, pp. 166-171, Feb. 1987. 293. P. Balaban and J. Salz, Optimum Diversity Combining and Equalisation in Digital Data transmission with Applications to Cellular Mobile Radio - Part I: Theoretical Considerations, IEEE Trans. Comm., vol. 40, pp. 885-894, May 1992. 294. P. Balaban and J. Salz, Optimum Diversity Combining and Equalisation in Digital Data transmission with Applications to Cellular Mobile Radio - Part II: Numerical Results, IEEE Trans. Comm., vol. 40, pp. 895-907, May 1992. 295. J.E. Smee and N.C. Beaulieu, On the Equivalence of the Simultaneous and Separate MMSE Optimisations of a DFE FFF and FBF, IEEE Trans. Comm., vol. 45, pp. 156-159, Feb. 1997. 296. B. Farhang-Boroujeny, Channel Estimation via Channel Identification: Algorithms and Simulation Results for Rapidly Fading HF Channels, IEEE Trans. Comm., vol. 44, pp. 1409-1412, Nov. 1996. 297. T. Nagayasu, S. Sampei and Y. Kamio, Complexity Reduction and Performance Improvement of a Decision Feedback Equaliser for 16QAM in Land Mobile Communications, IEEE Trans. Veh. Tech., vol. 44, pp. 570-578, Aug. 1995. 298. T. NaGayasu, S. Sampei and Y. Kamio, Complexity Reduction and Performance Improvement of a Decision Feedback Equaliser for 16QAM in Land Mobile Communications, IEEE Trans. Veh. Tech., vol. 44, pp. 570-578, Aug. 1995. 299. S. Ariyavisitakul, A Decision Feedback Equaliser with Time-Reversal Structure, IEEE J. Sel. Areas Comm., vol. 10, pp.599-613, Apr. 1992.

56

300. M.V. Eyuboglu, Detection of Coded Modulation Signals on Linear, Severely Distorted Channels using Decision-Feedback Noise Prediction with Interleaving, IEEE Trans. Comm., vol. 36, pp. 401-409, Apr. 1988. 301. K. Zhou, J.G. Proakis and F. Ling, Decision-Feedback Equalisation of TimeDispersive Channels with Coded Modulation, IEEE Trans. Comm., vol. 38, pp. 18-24, Jan. 1990. 302. A. Masoomzadeh-Fard and S. Pasupathy, Nonlinear Equalisation of Multipath Fading Channels with Noncoherent Demodulation, IEEE J. Sel. Areas Comm., vol. 14, Apr. 1996, pp. 512-520. 303. G.J. Foschini, Equalising without Altering or Detecting Data, Bell Syst. Tech. J., vol. 64, pp. 1885-1912, Oct. 1985. 304. A. Benveniste and M. Goursat, Blind Equalisers, IEEE Trans. Comm., vol. 32, pp. 871-883, Aug. 1984. 305. Y. Sato, A Method of Self-recovering Equalisation for Multilevel Amplitude Modulation Systems, IEEE Trans. Comm., vol. 23, pp. 679-682, Jun. 1975. 306. D.N. Godard, Self-Recovering Equalisation and Carrier Tracking in TwoDimensional Data Communication Systems, IEEE Trans. Comm., vol. 28, pp. 1867-1875, Nov. 1980. 307. S. Bellini, Bussgang Techniques for Blind Equalisation, Proc. Globecom86, Dec. 1986, pp. 1634-1640. 308. A. Benveniste, M. Goursat and G. Ruget, Robust Identification of a Nonminimum Phase System: Blind Adjustment of a Linear Equaliser in Data Communications, IEEE Trans. Auto. Contr., vol. 25, pp.385-399, Jun. 1980. 309. F.B. Ueng and Y.T. Su, Adaptive Blind Equalisation using Second- and Higher Order Statistics, IEEE J. Sel. Areas Comm., vol. 13, pp. 132-140, Jan. 1995. 310. J.K. Tugnait, Blind Equalisation of Digital Communication Channel Impulse Response, IEEE Trans. Comm., vol. 44, pp. 607-1616, Feb./Mar./Apr. 1994. 311. J.K. Tugnait, Blind Equalisation and Estimation of Digital Communication FIR Channels Using Cumulant Matching, IEEE Trans. Comm., vol. 43, pp. 12401245, Feb./Mar./Apr. 1995. 312. J.L.Valenzuela, A.Valdovinos and F.J. Casadevall, Performance of Blind Equalisation with Higher Order Statistics in Indoor Radio Environments, IEEE Trans. Veh. Tech., vol. 46, pp. 369-374, May 1997. 313. B. Porat and B. Friedlander, Blind Equalisation of Digital Communication Channels using High-Order Moments, IEEE Trans. Sig. Proc., vol. 39, pp. 522526, Feb. 1991. 314. F.-C. Zheng, S. McLaughlin and B. Mulgrew, Blind Equalisation of Nonminimum Phase Channels: Higher Order Cumulant Based Algorithm, IEEE Sig. Proc., vol. 41, pp. 681-691, Feb. 1993. 315. O. Shalvi and E. Weinstein, New Criteria for Blind Deconvolution of Nonminimum Phase Systems (Channels) IEEE Trans. Inf. Theory, vol. 36, pp. 312-321, Mar. 1990.

57

316. Y. Sato, A Blind Sequence Detection and its Application to Digital Mobile Communications, IEEE J. Sel. Areas Comm., vol. 13, pp. 49-58, Jan. 1995. 317. L. Tong, Blind Sequence Estimation, IEEE Trans. Comm., vol. 43, pp. 29862994, Dec. 1995. 318. M.K. Tsatsanis and G.B. Giannakis, Equalisation of Rapidly Fading Channels: Self-Recovering Methods, IEEE Trans. Comm., vol. 44, pp. 619-630, May 1996. 319. J.K. Tugnait, Blind Equalisation and Estimation of FIR Communications Channels using Fractional Sampling, IEEE Trans. Comm., vol. 44, pp. 324-336, Mar. 1996. 320. L. Tong, G. Xu and T. Kailath, Blind Identification and Equalisation based on Second-Order Statistics: a Time Domain Approach, IEEE Trans. Inf. Theory, vol. 40, pp. 340-349, Mar. 1994. 321. J.K. Tugnait, On Blind Identifiability of Multipath Channels using Fractional Sampling and Second-Order Cyclostationary Statistics, IEEE Trans. Inf. Theory, vol. 41, pp. 308-311, Jan. 1995. 322. J.K. Tugnait, Blind Equalisation and Estimation of FIR Communications Channels using Fractional Sampling, IEEE Trans. Comm., vol. 44, pp. 324-336, Mar. 1996. 323. S.J. Nowlan and G.E. Hinton, A Soft Decision-Directed LMS Algorithm for Blind Equalisation, IEEE Trans. Comm., vol. 41, pp. 275-279, Feb. 1993. 324. Y. Lou, Channel Estimation Standard and Adaptive Blind Equalisation, IEEE Trans. Comm., vol. 43, pp. 182-186, Feb./Mar./Apr. 1995. 325. G.-K. Lee, S.B. Gelfand and M.P. Fitz, Bayesian Techniques for Blind Deconvolution, IEEE Trans. Comm., vol. 44, pp. 826-835, Jul. 1996. 326. O. Shalvi and E. Weinstein, Super-Exponential Methods for Blind Deconvolution, IEEE Trans. Inf. Theory, vol. 39, pp. 504-519, Mar. 1993. 327. Y. Li and K.J. Liu, Blind Adaptive Spatial-Temporal Equalisation Algorithms for Wireless Communications Using Antenna Arrays, IEEE Comm. Lett., vol. 1, pp. 504-519, Jan. 1997. 328. H. Liu and G. Xu, Smart Antennas in Wireless Systems: Uplink Multiuser Blind Channel and Sequence Detection, IEEE Trans. Comm., vol. 45, pp. 187199, Feb. 1997. 329. M.K. Tsatsanis, G.B. Giannakis and G. Zhou, Estimation and Equalisation of Fading Channels with Random Coefficients, Sig. Proc., vol. 53, pp. 211-229, Sep.1996. 330. L.M. Davis, I.B. Collings and R.J. Evans, Identification of Time-Varying Linear Channels, Int. Conf. Ac. Sp. & Sig. Proc., ICAASP97 Munich, Apr. 1997. 331. X. Yu and S. Pasupathy, Error Performance of Innovations-Based MLSE for Rayleigh Fading Channels, IEEE Trans. Veh. Tech., vol. 45, pp. 631-642, Nov. 1996.

58

332. D.W. Matolak and S.G. Wilson, Detection for a Statistically Known, TimeVarying Dispersive Channel, IEEE Trans. Comm., vol. 44, pp. 1673-1683, Dec. 1996. 333. M.J. Gertsman and J.H. Lodge, Symbol-by-Symbol MAP Demodulation of CPM and PSK Signals on Rayleigh Flat-Fading Channels, IEEE Trans. Comm., vol. 45, pp. 788-799, July 1997. 334. S.M. Kay, Fundamentals of Statistical Signal Processing: Estimation Theory, Prentice-Hall, Englewood Cliffs, NJ, 1993.

Figure 1: Diagram of the communications system.

dt
0

h (t,u)
0

C0 Select the largest

r(t)

Data Decision

h1 (t,u)
T

C1

dt
0

Fig. 2: Kailaths estimator-correlator receiver for binary signalling (time-continuous case). The time-varying filters { h i (t , u ), i = 0 ,1} represent MMSE estimators, whereas C i , i = 0, 1} are proper bias terms. {

59

Oscillator Attenuator RF Output Modulator

Data Source

Manchester Encoding

(a) S(f) Pilot Data Spectrum

(b)

Pilot Band-Pass Filter Input

Calibration Subsystem Data Decision

Band-Pass Filter

Matched Filter

(c)

Fig. 3: (a) TCT transmitter scheme; (b) Baseband spectrum of the transmitted signal; (c) TCT receiver scheme.

60

Pilot Symbol Source

Data Source

RF Output Modulator Switch (a) (K-1) Data Symbols

Pilot Symbols (b) Pilot Samples Matched Filter Output Switch Data Samples

Channel Estimator Data Out

Delay Line

Channel Compensation

Data Decision

(c)

Fig. 4: (a) PSAM transmitter; (b) PSAM frame format (one pilot symbol is transmitted after (K 1) information symbols); (c) PSAM receiver.

61

Second Stage

Improved Data Decision

Matched Filter Output Samples

Fading Compensation

Data Decision

Symbol Removal

One Symbol Delay

One-Step Predictor First Stage

Fig. 5: General scheme of a symbol-by-symbol receiver with decision feedback; a second detection stage can be added to improve the reliability of the data decisions produced by the first stage.

Survivor path of the trellis state k

k+1

~ ck

k-2

k-1

k+1

Fig. 6: State-trellis of a VA-based algorithm. 62

Figure 7: Diagram of an inverting equaliser.

ri

VITERBI ALGORITHM CHANNEL IMPULSE RESPONSE CHANNEL ESTIMATOR

k c

Figure 8: Diagram of an adaptive MLSE equaliser.

ri

VITERBI ALGORITHM
CHANNEL IMPULSE RESPONSES

k c

SURVIVOR SEQUENCES

CHANNEL ESTIMATORS

Figure 9: Diagram of a Per-Survivor-Processing MLSE equaliser.

63

Channel Estimator Matched Filter Output

CSI Deinterleaver Data Out

Delay Line

Deinterleaver

Channel Decoder

Fig. 10: Two-stage receiver for an interleaved coded transmission.

64