267

DOWNLINK INTERFERENCE AND POWER

ALLOCATION ANALYSIS DURING SOFT HANDOFF OF
WCDMA CELLULAR NETWORKS


Alok Aggarwal Rajendra Prasad Mohan Lal Sunita Devi
Deptt. of CSE, Department of EE, Institute Computer Center Deptt. of WRD & M
DIT, Dehradun IIT Roorkee, Roorkee IIT Roorkee, Roorkee IIT Roorkee, Roorkee
Email Id: Email Id: Email Id: Email Id:
alok289@yahoo.com rpdeefee@iitr.ernet.in mluccfcc@iitr.ernet.in


Abstract
There has been a tremendous growth in wireless
communication technology over the past decade.
The significant increase in subscribers and traffic,
new bandwidth consuming applications such as
gaming, music downloading and video streaming are
placing new demands on capacity. The answer to the
capacity demand is the provision of new spectrum
and the development of a new technology
Wideband Code Division Multiple Access
(WCDMA). WCDMA is an interference limited
system. This paper analyzes the interference and
power allocation on the downlink direction of
WCDMA during soft handoff.

I Introduction
Wideband Code Division Multiple Access
(WCDMA) has been selected as the radio-
interference technology for 3
rd
generation mobile
communication systems. The availability of high bit
rate services of 3
rd
generation systems will offer the
possibility for the transfer of high bit rate packet
data needed for wireless Internet access as well as
circuit switched video services [1]. A significant
feature of the next generation mobile communication
system is the capacity of providing communication
services for high speed Internet access with which
web browsing will be more popular for mobile
terminal users [2][3]. In the web browsing, mobile
stations send short request messages in the uplink
and receive texts, images and video in the downlink
and thus the amount of traffic is larger in the
downlink than in uplink. Many theoretical studies
have shown that a WCDMA system can support a
greater number of subscribers than a TDMA or
FDMA system [4][5]. But the CDMA system is
interference limited. An important feature of CDMA
system is soft handoff (SHO). In soft handoff, the
mobile station starts communication with a new base
station without disconnecting from the old base
station. The voice quality is improved due to
diversity provided by extra channel path at the cell
edge if compared with hard handoff. From the
network point of view, however more resources are
required to support this procedure. As a crucial
technology for seemless communication, SHO has
been investigated in literature since the emergence
of CDMA.

The objective of this paper is to analyze the
interference and power allocation on the downlink
direction of WCDMA during soft handoff.
Performance of soft handoff is estimated on the basis
of soft handoff effects on power allocation for
downlink dedicated channels.
Paper is analyzed as follows: In the Section II,
system model along with propagation model is
given. In section III, downlink interference analysis
is given. In section IV, intra-cell and inter-cell
interference are given. In section V soft handoff
effects on downlink interference are given. In
section VI, power allocation in soft handoff
condition (2-way soft handoff and 3-way soft
handoff) and with soft handoff condition are given,
along with simulation results. In section VII,
conclusions based on the results of Section VI are
given.

II System and Propagation Model
The propagation attenuation is generally modeled as
the product of
th
power of distance and a log-
normal component representing shadowing loses [9].
XXXII NATIONAL SYSTEMS CONFERENCE, NSC 2008, December 17-19, 2008
268
We consider a system with ideal 19 co-channel
macro cells [9].

III Downlink Interference Analysis
CDMA systems are interference-limited and so
interference evaluation is one of the fundamental
procedures for analyzing the CDMA systems. The
total interference experienced by a mobile is
composed of two parts: intra-cell and inter-cell
interference. In the uplink, to a certain mobile, the
intra-cell interference comes from all the other
mobiles served by the same BS, the inter-cell
interference is composed of all the signals received
from all the mobiles in other cells other than the
mobiles serving cell, as shown in figure 1.
Therefore, in the uplink, the interference
experienced by a certain mobile is related to the load
distribution within the network but not related to the
mobiles own location.


Figure 1: Uplink interference

In the downlink, as shown in figure 2, the intra-cell
interference to a certain mobile comes from its
serving base station. This interference is caused by
the partial loss of orthogonality among the users due
to multipath effect. WCDMA employs the
orthogonal codes in the downlink to separate users
and without any multi-path propagation the
orthogonality remains and there is no intra-cell
interference. Typically, the orthogonality is between
0.4 and 0.9 [6] (1 corresponds to perfect
orthogonality). The intra-cell interference includes
part of the power for common control channels and
the power for the downlink traffic channels for the
other users in the same cell. The inter-cell
interference is the power received by the mobile
from all the other base stations except for its own
serving base station. Because in WCDMA FDD
mode, base stations are not synchronized, the inter-
cell interference does not get benefits from the
orthognality as happens with the intra-cell
interference. Because the interference sources are
fixed, in the downlink direction to a certain mobile
the downlink interference is inevitably linked to the
mobiles location. The following analysis shows this
feature and reveals the soft handoff effects in the
downlink interference.


Figure 2: Downlink interference


IV Intra-cell and inter-cell interference
To a mobile located at (r
1
,
1
) in figure 2, assuming
BS
1
is its serving base station, the downlink intra-
cell interference, I
intra_cell
, which is received from
BS
1
, can be expressed as
10 /
1
-
i 1 T cell ra int
r ) a 1 ( P I

10 =

(1)
where P
T1
is the total transmit power of BS
1
; a is the
downlink orthogonal factor with 1 for perfect
orthogonality and 0 for non-orthogonality. Because
the intra-cell interference and the desired signal are
both transmitted from the same source hence they
experience the same attenuation. Thus there is need
to use power control in a single cell system.

The inter-cell interference, I
inter_cell
, can be expressed
as

10 =
M
2 i
-
i Ti cell er int
10 /
i
r . P I

(2)
where P
Ti
is the total transmit power of BS
i;
r
i
is the
distance from the mobile to BS
i;
M is the index of
BSs which are taken into account for the inter-cell
interference. Theoretically, the inter-cell interference
comes from all the BSs around other than the
serving BS. Here, the BSs in the first and second
tiers around BS
1
are considered because the power
received from the BSs outside the second tier is
negligible. It can be seen that the intra-cell
interference relies on the distance from the mobile to
269
its serving BS, r
1
,

but it is independent of
1
.

However, the inter-cell interference depends not
only on r
1
but also on
1
because the distance from
the mobile to other base station, e.g. BS
i
, r
i
, is a
function of r
1
and
1
.

19 ... 13 11, 9, i for

6 / ) 8 i ( [ Cos . R . 3 . r . 2 ) R . 3 ( r r
18 ... 12 , 10 , 8 i for
6 / ) 8 i ( [ Cos . R . 3 . 2 . r . 2 ) R 3 . 2 ( r r
7 i 2 for
3 / ) 2 i ( [ Cos . R . 3 . r . 2 ) R 3 ( r r
1 1
2 2
1 i
1 1
2 2
1 i
1 1
2 2
1 i
=
+ + =
=
+ + =

+ + =
(3)

It is clear that to a certain mobile the downlink
interference is related closely to the mobiles
location. This is because the interference sources are
fixed base stations in the downlink direction.
Assuming that the load is uniformly distributed
within the system and the total transmit power of
each base station is the same, denoted by P
T
, can be
written as

10 =

=

. P r P I
T
M
2 i
-
i T cell er int

10 /
i (4)
T
cell er int
P
I

=
(5)
where the factor provides a measure of the inter-
cell interference to the base station total transmit
power. is related to the mobiles location. Here the
cell radius is normalized to 1.

V Soft handoff effects on downlink
interference
To guarantee the QoS, the base station needs to
allocate the proper amount of power to each mobile
to compensate for the interference. If the mobile is
not in soft handoff status, as mobile 1 shown in
figure 3(a), only one downlink channel is set up
between the mobile and its serving base station, BS
1
,
according to the service requirement and the total
downlink interference received by mobile 1, which
is denoted by I
0
, power P is allocated to the
downlink channel between the mobile and BS
1
. This
channel acts as intra-cell interference and inter-cell
interference to mobile 2 and mobile 3 respectively.

If mobile 1 is in soft handoff status, it communicates
with BS
1
and BS
2
simultaneously. Two downlink
dedicated channels are set up to support the soft
handoff as shown in figure 3(b). Let P
1
and P
2

represent the power allocation to channels from BS
1
and BS
2
separately. P
1
acts as intra-cell interference
to mobile 2 and inter-cell interference to mobile 3
and P
2
acts as inter-cell interference to mobile 2 and
intra-cell interference to mobile 3.




Figure 3: Soft handoff effects on the downlink
interference


Comparing the two cases, soft handoff effects on
downlink direction are quite complicated. Without
soft handoff mobile 1 contributes power P to the
total downlink interference. With soft handoff the
total contribution is the sum of P
1
and P
2
. The
increment of the interference due to mobile 1 has
influence on all the other active mobiles within the
system, all these mobiles need to adjust their channel
power to meet the change in the interference. This,
in return, changes the total interference received by
mobile 1, resulting the alteration of P or P
1
and P
2

(SHO case). This circulation repeats until the system
reaches a new balance. In CDMA systems, power
control is the functionality responsible for this
adjustment. Because of the interference limited
feature, experiencing less interference is always the
main principle of the radio resource allocation in
CDMA systems. Therefore, whether soft handoff
leads to lower interference than the conventional
270
hard handoff or not, depends intimately on the value
of P, P
1
and P
2
. These powers are related to certain
facts, such as the location of the mobile, the radio
attenuation and the power division strategy
employed during the soft handoff.


VI Power Allocation
Without Soft Handoff
Without soft handoff mobile only communicates
with one base station. Let us consider that this base
station is BS
1
. Then the received E
b
/I
0
would be

vR
W
0
I
b
E
=
(
(
(
(
(
(

=

+

19
2 i
-
i
r
Ti
P
-
1
a).r - .(1
T1
P
-
1
.r
s
P
(6)

Here v is the activity factor, R is the service bit rate
and W is the chip rate.

(6) can be rewritten as
(
(
(

|
|
.
|

\
|
+
|
|
.
|

\
|
=

=
19
2 i
1
i
Ti 1 T
t
0
b
r
r
P ) a 1 .( P
I
E
W
R . v
Ps
(7)

The target value of (E
b
/I
0
) is E
b
/I
0,
which is decided
by RNC.

If load is distributed evenly then
(
(
(

|
|
.
|

\
|
+
|
|
.
|

\
|
=

=
19
2 i
1
i
T
t
0
b
r
r
) a 1 ( P .
I
E

W
R . v
Ps
(8)
= A
1.
P
T
(9)

Here A
1
is the relative strength of the required power
for the mobile located at (r
1
,
1
) without soft handoff.
For other parameters, we have used the typical
values as R=12.2 kbit/s, a=0.6, W=3840 kchip/s,
(E
b
/I
0
)
t
=5dB.

With Soft Handoff
Two Way Soft Handoff Case
In this case all the base stations which are in active
set need to allocate proper power for the DL
channels linked to this mobile. Let BS
1
and BS
2
be
the serving base stations of the mobile, then the
received E
b
/I
0,
based on maximal ratio combining
scheme of the signal coming from both base stations,
would be


2 0
I
b
E
1
0
I
b
E
0
I
b
E

(
(

(
(

+ =
(10)
(
(
(
(
(

=
19
2 j
1 j
-
j
r
Tj
P
-
2
a).r - .(1
T2
P
-
2
.r
b
P
19
2 i
-
i
r
Ti
P
-
1
a).r - .(1
T1
P
-
1
.r
a
P

vR
W



. (11)

Where P
a
and P
b
are the transmit power of BS
1
and
BS
2
respectively and r
1
and r
2
are the distances
between mobile and BS
1
and BS
2
respectively.

During macro-diversity, in inner closed loop power
control there is a need of creating balance of
downlink power among active cells, for which
adjustment loop is employed. Due to this stability
problem and transmission power is increased but on
the other hand power drifting can be avoided, which
is due to activation of two power control loops as we
are dealing with 2-way soft handover case. In perfect
situation

P
a
= P
b
(12)

Hence comparing (11) and (12)

=

=

|
|
.
|

\
|
+
+
|
|
.
|

\
|
+
|
|
.
|

\
|
=
19
) 2 j (
1 j 2
j
19
2 i 1
i
T
t
0
b
a
r
r
a 1
1
r
r
a 1
1
P .
I
E
R . v
P

(13)
271
Same expression would hold for P
b
. Thus total
power to support this mobile would be the addition
of P
a
and P
b
, which would be

a
waySHO 2
P . 2 P =
(14)


r
r
a 1
1
r
r
a 1
1
P .
I
E
R . v . 2
P

19
) 2 j (
1 j
2
j
19
2 i
1
i
T
t
0
b
waySHO 2

=

=

|
|
.
|

\
|
+
+
|
|
.
|

\
|
+
|
|
.
|

\
|
=

(15)

= A
2.
P
T
(16)


Here the factor A
2
is the relative strength of the total
power required for the mobile under 2-way Soft
handover.

With Shadowing Effects:

=


=


|
|
.
|

\
|
+
+
|
.
|

\
|
+
|
.
|

\
|
=
19
2 j
1 j
10 / ) (
2
j
19
2 i
10 / ) (
1
i
T
t
0
b
waySHO 2
2 j 1 i
10
r
r
) a 1 (
1
10
r
r
) a 1 (
1
P .
I
E

W
R . v
2.
P
(17)

Three Way Soft Handoff Case
Let BS
1,
BS
2
and BS
3
be the serving base stations of
the mobile, then the received E
b
/I
0,
based on
maximal ratio combining scheme of the signal
coming from both base stations, would be


3
0
I
b
E
2
0
I
b
E
1
0
I
b
E
0
I
b
E
(
(

(
(

(
(

+ + =
(18)

Similar to the last case, we can get


19
) 3 k (
1 k
3
r
k
r
a 1
1
19
) 2 j (
1 j
2
r
j
r
a 1
1
19
2 i
1
r
i
r
a 1
1
T
P .
t
0
I
b
E
R . v . 3
waySHO 3
P

=

+
+

=

+
+

=

+
=
|
.
|

\
|
|
|
.
|

\
|
|
.
|

\
|
|
.
|

\
|
(19)

= A
3.
P
T
(20)

With Shadowing Effects:

=

|
|
.
|

\
|
+
+

=

|
|
.
|

\
|
+
+

=

|
|
.
|

\
|
+
|
|
.
|

\
|
=
19
) 3 k (
1 k
3
r
k
r
a 1
1
19
) 2 j (
1 j
2
r
j
r
a 1
1
19
2 i
1
r
i
r
a 1
1
T
P .
t
0
I
b
E
R . v . 3
waySHO 3
P
(21)

VII Conclusion
In the downlink direction the inter-cell interference
is related closely to the location of the mobile user.
To the users at the cell boundaries, the inter-cell
interference is the main part of the total interference,
especially with higher orthogonality. Soft handoff
effects on the downlink direction are quite
complicated, depending on such factors as the
location of the mobile, the radio attenuation and the
power division strategy employed. Soft handoff
requires more power than the case of without soft
handoff If there are more base stations involved,
more power consumption would be resulted.
Further, at the cell boundaries, soft handoff
decreases the total average power needed by mobiles
and inter-cell interference is the more contributory to
the total interference. In downlink direction, the
inter-cell interference is closely related to the
location of the mobile. Fade margin of the individual
connections is reduced due to soft handoff.


VIII References
1. E. Dahlman, P. Beming, J. Knutsson, F. Overjo,
M. Person and C. Roobol, WCDMA the radio
interface for future mobile multimedia
communication, IEEE Tran. on Veh. Tech., vol.
47, no. 4, pp. 1105-1118, 1998.
2. F. Adachi, M. Sawahashi and H. Suda,
Wideband DS-CDMA for next generation
272
mobile communication systems, IEEE
Communication Magazine, pp. 56-69, Sep.
1998.
3. E. Dahlman, B. Gudmundson, M. Nilsson, and J.
Skold, UMTS/IMT 2000 Baed on Wideband
CDMA, IEEE Communication Magazine, pp.
70-80, Sep. 1998.
4. K.S. Gilhousen, On the capacity of a cellular
CDMA systems, IEEE Tran. on VT, Vol. 40,
pp. 303-312, 1999.
5. K.S. Gilhousen, Increased capacity using
CDMA for mobile satellite communication,
IEEE Journal on Selected Areas in
Communication, Vol. 8, pp. 503-514, 1990.
6. J.P. Castro, The UMTS Network and Radio
Access Technology: Air Interface Techniques
for Future Mobile Systems, Wiley, John &
Sons, Incorporated: UK, may 2001.
7. J. Cho and D. Hong, Statistical model of
downlink interference for the performance
evaluation of CDMA systems, IEEE
Communication Letters, Vol. 6, Issue 11, pp.
494-496, Nov. 2002.
8. A. Chheda, A performance comparison of the
CDMA IS-95B and IS-95A Soft Handoff
Algorithm, VTC99, vol.2, pp. 1407-1412,
1999.
9. Alok Aggarwal, Rajendra Prasad, Mohan Lal,
and Sunita Devi, Downlink Soft Handoff
Effects in WCDMA Networks under Different
Power Control Conditions, Journal of
Information Technology, Vol. 3, No.3, pp. 31-
38, Sep. 2007.


0
0.01
0.02
0.03
0.04
0.05
0.4 0.6 0.8 1
r1/R
p
o
w
e
r

r
a
t
i
o
=0 =30 =45 =60

Figure 4: Power allocation versus r
1
/R for 2-way
SHO case when shadowing effects are considered
0
0.01
0.02
0.03
0.04
0.05
0.4 0.6 0.8 1
r1/R
p
o
w
e
r

r
a
t
i
o
=0 =30 =45 =60
Figure 5: Power allocation versus r
1
/R for 2-
way SHO case when shadowing effects are not
considered

0
0.01
0.02
0.03
0.04
0.05
0.4 0.6 0.8 1
r1/R
p
o
w
e
r

r
a
t
i
o
=0 =30 =45 =60

Figure 6: Power allocation versus r
1
/R for 3-way
SHO case when shadowing effects are considered

0
0.01
0.02
0.03
0.04
0.05
0.4 0.6 0.8 1
r1/R
p
o
w
e
r

r
a
t
i
o
=0 =30 =45 =60

Figure 7: Power allocation versus r
1
/R for 3-way
SHO case when shadowing effects are not
considered