Beruflich Dokumente
Kultur Dokumente
www.elsevier.com/locate/isatrans
Received 21 February 2007; received in revised form 14 May 2007; accepted 31 May 2007
Available online 30 July 2007
Abstract
In this paper, a mathematical model and control theoretical framework for designing AQM controllers in networks supporting TCP Vegas
is introduced. We have emphasized on a modified TCP Vegas algorithm that can respond to congestion signals through explicit congestion
notification (ECN). The overall nonlinear delayed differential equations of the dynamics model of closed loop system have been derived based
on TCP Vegas model. The model is then linearized to derive a transfer function representation between the packet marking probability and the
bottleneck router queue length as the input and output of the modified TCP Vegas/AQM system. The model properties have been then examined
especially for the case of single bottleneck homogeneous network which is closely investigated. Finally an AQM controller based on Coefficient
Diagram Method (CDM) has been designed for the system and its performance has been compared with some other AQM controllers. CDM
is a new indirect pole placement method that considers the speed, stability and robustness of the closed loop system in terms of time domain
specifications. In order for synthesizing the simulation scenarios, our campus router traffic has been studied experimentally for a sample period of
one hour and the corresponding parameters has been extracted. The simulation results are representative of good performance of developed TCP
Vegas/AQM structure for different simulated scenarios.
c 2007, ISA. Published by Elsevier Ltd. All rights reserved.
Keywords: Active queue management; Coefficient diagram method; Computer networks; Congestion control; Internet measurement; Network modelling; TCP Vegas
feedback from the network. This property, however, leads to marking probabilities and queue lengths in bottleneck routers
some types of performance degradation in cases that the round as system inputs and outputs, respectively. The derived model
trip propagation delay can not be calculated accurately. Some is then used to analyze system behaviour and to design AQM
examples are when there is backlogged data in the queue or controller. In special case of a single bottleneck shared by
in case of rerouting [5,6]. There are some improved Vegas some homogeneous flows, the developed model is similar to
algorithms proposed in the literature that attempt to treat TCP one derived in [9] for TCP Reno/AQM networks by Misra
Vegas drawbacks [7,8]. In [7], a modified Vegas called Vegas- et al. As part of our work, we use this model to investigate
Like AMP-TCP has been introduced which employs model the homogeneous single bottleneck system properties more
predictive control strategy in the end-hosts to regulate their closely in terms of topology and traffic parameters. Afterwards,
number of in-flight packets such that each flow derives its design of an AQM controller is performed based on Coefficient
desired round trip time value. Another proposal called Vegas- Diagram Method (CDM). The design procedure of CDM as
A has been introduced in [8] to address the limitations of TCP an AQM controller for TCP Reno/AQM networks have been
Vegas such as its problem in case of rerouting. In this method, discussed by the authors in series of papers [14–16] and its
the desired number of in-flight packets is adaptively determined out-performance with respect to common AQM controllers
based on actual throughput of TCP for each flow. However, has been shown there. This good background motivated the
none of these methods use explicit/implicit feedback signals authors to design the CDM as AQM controller for networks
from congestion status of the underlying networks, while it has supporting TCP Vegas. In this paper, the performance of
been shown that by cooperation of end-users and intermediate the TCP Vegas/CDM networks is evaluated through different
nodes (routers) in the network, the overall behavior of the simulation scenarios in comparison with other AQM methods
networks can be noticeably improved and congestion can be such as RED and PI. In order for synthesizing our simulation
prevented or treated more adequately [6]. scenario, the properties of the traffic passing through our
Active Queue Management (AQM) methods are those campus router have been measured via some experiments.
techniques which attempt to prevent the congestion and Packet length and RTT, traffic composition, and flow size
regulate the queue length by sending congestion signals through and number of active flows are the considered features of the
marking/dropping packets, which is assumed to eventually measured data. The data have been gathered for one hour on
cause the senders reduce their sending rates. They try to November 20 2006, 9–10 AM.
solve the congestion control problem by regulating their queue The remainder of this paper is organized as follows. In
length around a desired value [9,10]. There are different AQM Section 2, the principals of TCP Vegas operation is described.
methods introduced in the literature. They are designed such The modified Vegas and its relationship with AQM control
that they can cooperate with the TCP variants that can react signal, the linearized model and special case of homogenous
to congestion notification signals [6]. Many of these AQM
networks are presented in Section 3. Section 4 deals with CDM
methods are designed based on control theory methods and use
design as an AQM controller. Simulation results are brought in
the fluid flow models of different TCP variants especially for
Section 5. The paper is concluded in Section 6.
the TCP Reno model of [11]. For example in [12], proportional
plus integral (PI) controller and in [13] an adaptive stable PI
has been designed based on this model. Another example is the 2. TCP vegas principals
coefficient diagram method controller designed by the authors
in [14–16]. In the same way, by modifying Vegas algorithm so In this section, the principals of the operation of a general
that it reacts to the congestion signals coming from the routers, network supporting TCP Vegas are presented [5]. Consider a
one can expect more precise reaction as well. In [6], a modified network consisting of a set of L links with finite capacities
Vegas is proposed, in which the information of packet marking cl , l ∈ L. Assume the links are shared by a set of N sources
probability in the queues in addition to Round Trip Time (RTT) indexed by r . Each source r attains a utility Ur (xr ) when
is used to estimate the propagation round-trip delay and to it transmits its data packets with a rate of xr (in packets/s)
determine the data sending rate, when random early marking through a set of links defined by the L × N routing matrix
(REM) algorithm [9] is employed in routers as AQM method.
In this paper, the proposed idea in [6] has been extended 1 if source r uses link l
Rlr =
to derive a framework for designing AQM methods for network 0 otherwise.
supporting TCP Vegas. We have emphasized on a modified TCP
Vegas algorithm that can respond to congestion signals through Associated with each link l is a congestion measure pl (t),
explicit congestion notification (ECN). If networks support called ‘price’, which is the scaled queuing delay at link l. Let us
the ECN, AQM schemes mark the ECN bits as congestion denote the total round trip propagation delay of source r by dr ,
notification [17] instead of dropping packets. We first formulate the forward delay from source r to link l by τrl , the backward
the model of the closed loop TCP Vegas/AQM network in its delay from link l to source r by τlr , and the round trip time by
general form for an arbitrary number of bottlenecks. Then this Tr . Then at time t, the aggregate price in the path of source r
model is linearized to obtain a transfer function representation will be:
of the small signal behavior of TCP Vegas/AQM networks.
X
qr (t) := Rlr pl (t − τlr ), (1)
The transfer functions relate small variations in the packet l
N. Bigdeli, M. Haeri / ISA Transactions 47 (2008) 143–155 145
and the link l observes the aggregate source rate of 3.1. Modified vegas/AQM networks modelling
X
yl (t) := Rlr xr (t − τrl ), (2) Consider the modified Vegas in [6], in which the variation in
r window size is a function of dropping/marking probability of
packets in the intermediate routers, that is
where, Tr = τrl + τlr , ∀l ∈ Lis the round trip time.
According to an interpretation of Vegas, a source monitors 1 qr (t) + εr
wr (t) + if xr (t) < αr
the difference between its expected rate and its actual rate, and (t)
T dr
r
increases or decreases its window size by one in the next RTT
wr (t + 1) = 1 qr (t) + εr (6)
according to weather the difference is less than or greater than wr (t) − if xr (t) > αr
T (t) d
r r
a parameter αr [5]. Here, αr is a Vegas protocol parameter in
wr (t) otherwise,
terms of KB/s and has significant effect on the fairness of Vegas
protocol [5]. Let wr (t) be the window size of source r at time t, where εr is a function of the aggregate marking/dropping
then the change in window size can be modeled by one packet probability, m r , along the dedicated path for source r , that is
per RTT according to
εr = µr m r , (7)
1 wr (t) wr (t)
wr (t) + if − < αr where µr is a constant that should be chosen. m r can be
(t)
Tr dr Tr approximated as
wr (t + 1) = 1 wr (t) wr (t) (3a)
w (t) > α
Y
r −
Tr (t)
if
dr
−
Tr
r m r (t) = 1 − Rlr (1 − pd,l (t − τlr ))
wr (t) else . l
X
∼
= Rlr pd,l (t − τlr ), (8)
In [2], wr (t)/dr is referred to as the expected rate, l
wr (t)/Tr (t) as the actual rate, and the difference wr (t)/dr −
w where pd,l is the marking/dropping probability for the queue in
Pr (t)/Tr (t) = xr (t)q Pr (t)/dr as ‘DIFF’; where qr (t) = link l. By employing such algorithm, the TCP Vegas reacts to
l Rlr pl (t − τlr ) = l Rlr bl (t − τlr )/cl is the end-to-end
queuing delay (backlog) in path of source r and bl (t) is the the congestion notification signals fed back by the router AQM
queue length of link l [5]. But, the dynamics of queue length controllers.
As our complementary point, we argue that many effective
variation in link l can be considered as the difference between
AQM methods such as RED [10] use the exponentially
its input data rate and the rate of servicing data. If the data
weighted moving average (EWMA) of the queue length.
processing is not the bottleneck in the corresponding router, the
In [18], the necessity and usefulness of such assumption is
queue length dynamic can be modelled as
discussed in details. The average queue length bavg,l (t) for
ḃl (t) = yl (t) − cl . (3b) the queue l as an exponentially weighted moving average of
instantaneous queue length bl (t) with a weight factor ρl is [10]:
From (3) and replacing the sending rate xr (t) = Z t
wr (t)/Tr (t), the Vegas algorithm can be described by the bavg,l = ρl es−t bl (s)ds. (9)
following couple of differential equations: −∞
Replacing (6) into (5), and from (3b) and (9), the overall
(yl (t) − cl )/cl if pl (t) > 0
ṗl (t) = (4) nonlinear model of the interaction between TCP Vegas sources
(yl (t) − cl )+ /cl if pl (t) = 0, and queue length dynamic in intermediate routers can be
ẋr (t) = sgn(1 − xr (t)qr (t)/αr dr )/Tr2 (t), (5) derived as three sets of nonlinear differential equations as
qr (t) + µr m r (t)
where, (z)+ = max(0, z) and sgn(z) is the sign function [5]. ẋr = sgn 1 − xr (t) /Tr2 (t)
αr dr
(10)
3. Modified vegas and AQM design framework ḃl (t) = (yl (t) − cl )+
ḃavg,l (t) = −ρl bavg,l (t) + ρl bl (t),
The performance of Vegas algorithm would be improved if
where pl (t), the average queuing delay, is now defined as
a source can make use of marking information of the routers
(those employ AQM) to estimate the queue occupancy of pl (t) = bavg,l (t)/cl , (11)
the bottleneck [6]. On the other hand, by employing AQM b (t−τ )
and Tr (t) = qr (t) + dr = l Rlr avg,l cl rl + dr . The other
P
methods in routers, one can regulate the queuing delay and
therefore, the QoS requirements is satisfied by controlling definitions remain unchanged.
factors such as average queuing delay and its jitter or supporting
3.2. Model linearization and the transfer function representa-
differentiated services. In [6], the first aspect, i.e. improving
tion
Vegas performance has been dealt with. Here, we use the same
context to achieve the second goal, i.e. deriving a framework for In this section, the derived model is analysed in order to
designing AQM methods in order to satisfy QoS requirements. design the AQM congestion controller. To do so, we first derive
146 N. Bigdeli, M. Haeri / ISA Transactions 47 (2008) 143–155
Fig. 3. Block diagram of the small signal interactions in the linearized model
of Eq. (19).
Fig. 8. (a) The empirical observed packet length; and (b) the packet length
distribution. Fig. 10. The composition of IP traffic for different protocols in terms of (a)
number of packets and (b) volume of bytes.
Fig. 11. Experimental flow size in terms of volume of byte versus number of
packets.
Fig. 9. (a) The observed RTT; and (b) its empirical cumulative distribution
function.
Fig. 13. CDM controller performance for in comparison with PI and RED for Fig. 14. CDM controller performance for in comparison with PI and RED for
nominal parameters: CDM (solid), PI (dash), and RED (dot). unknown parameters: CDM (solid), PI (dash), and RED (dot).
To design the CDM controller, we should choose proper 5.2.1. Exp. 1: Unknown parameters
values for τ and γi as stated in Section 4.2. Selecting τ = The performance of the designed CDM controller along
4T p∗ = 3 s and γi = [∞, 7.5, 6, 6, 6, 6, ∞], the closed loop with those of RED and PI controllers are investigated for the
case that the network parameters are assumed unknown. It is
characteristic polynomial becomes P(s) = 3.134e−6s 6 +
assumed that the real values of the plant parameters are: N p =
0.0102s 5 + 5.4874s 4 + 493.86s 3 + 7.407e3s 2 + 1.9e4s + 65, d p = 0.35 s, and C p = 80 packets/s and all other network
6.173e3. Therefore, the controller polynomials are: and controller parameters remain the same as considered for
the nominal network. The simulation result for regulating the
A(s) = 4.18e−6s 3 + 0.0133s 2 + 6.56s queue length around B p = 50 has been shown in Fig. 14. It
(36)
B(s) = 0.0346s 3 + 1.1849s 2 + 2.978s + 1. is seen that the queue regulation objective has been obtained
for CDM, while for PI some small oscillations are observed
For comparison, two well-known AQM controllers i.e. in this case. For RED, the equilibrium queue length has been
RED [6,10] and proportional plus integral (PI) controllers doubled, which is due to the RED algorithm which does not
are also simulated. The RED parameters are set the same explicitly consider the desired queue length for calculating
as those of [6]. The parameters maximum drop probability, packet marking/dropping probability [28]. Although the packet
minimum queue threshold and maximum queue threshold marking probability of CDM has been increased due to the
(i.e. pmax , minth and maxth ) of RED are set to 0.1, 10, and difference in the operating point in effect of the network
150, respectively. The PI controller has been designed based on parameter mismatches, it is still less than that PI controller.
Ziegler–Nichols method [28]. The resulting parameters become However, RED controller has the smallest packet marking
k p = 0.0572 and ki = 0.0475 as the proportional and integral probability among all.
gains, respectively. The responses of the designed controllers Seeking for the limits of applicability, the behaviour of the
for the nominal network parameters are shown in Fig. 13. designed controllers for various load/network parameters has
been investigated in the above described network. For this
From the graphs in this figure, it is observed that both queue
purpose, ±50% variations have been applied to the parameters
and rate regulation has been best done by the CDM method.
C p , N p and d p ; and the controllers’ performance have been
PI performance is also good in this case, but, the transient
examined. The underlying results have been summarized in
behaviour of CDM is better than PI. Damping oscillations in
Table 2. From this table, the simulation results show that
response and a large settling time as well as a small steady-state the CDM controller keeps its performance for all over the
error of the queue length are the properties of RED performance variation ranges. The steady state queue length does never
with respect to CDM and PI AQM controllers. show deviation more than 12% from its desired value and the
In continue we investigate the performance of the response settles down in a reasonable time with respect to the
designed controllers under different unknown and time-varying nominal response. Moreover, such performance is observed
circumstances. for larger ranges of parameter variations as well (up to 100%
N. Bigdeli, M. Haeri / ISA Transactions 47 (2008) 143–155 153
Table 2
The performance summary of the designed controllers for ±50% variations in load/network parameters
Parameter Stabilty region (%) Steady state queue vaiations in stable region Setting time variations
CDM PI RED CDM PI RED CDM PI RED
Np ±50% [−5%, 40%] [−20%, 50%] ±6% ±24% [−20%, 70%] ±15% [−100%, 200%] ±40%
Asc. Desc. Asc. Desc. Desc. Desc.
Cp ±50% [−50%, 5%] [−50%, 30%] ±5% ±25% [−20%, 10%] −25%, 10% [+15%, 40%] ±50%
Asc. Asc. Asc. Desc. Asc. Desc.
dp ±50% [−10%, 50%] [−30%, 50%] [−15%, +4%] ±50% ±25%] [−6%, 14%] [−20%, 8%] ±35%
Asc. Desc. Asc. Asc. Desc. Asc.
Here ‘Asc.’ abbreviates ‘Ascending’ and ‘Desc.’ abbreviates ‘Descending’.
Fig. 15. CDM controller performance for in comparison with PI and RED for time-varying parameters: (a) CDM (left column graphs), (b) PI (middle graphs), and
(c) RED (right column graphs).
increase). On the other hand, PI and RED become unstable for in a linear ascending/descending manner as mentioned in
parameter variations outside the specified ranges in Table 2. Table 2. Besides, for RED and PI, the settling time varies
Besides, the ability of queue length regulation and the transient significantly in effect of parameter variations which is not
response of PI and RED degrade dramatically in comparison desirable.
with CDM. For the network with PI as AQM controller,
deviation from nominal parameters leads to small oscillations 5.2.2. Exp. 2: Time-varying parameters
around the queue length equilibrium point. The amplitude In Fig. 15, the performance of the CDM controller is
of such oscillations increases as the deviation increases until compared with the PI and RED AQM controllers in time-
it reaches instability; while the margins of stability are not varying situation. For this purpose, step changes in the
symmetrical with respect to the nominal parameters. Queue parameters C p , N p and d p have been applied to the network
length equilibrium point in a network employing RED, changes according to the last row of Fig. 15. As seen, while both PI and
154 N. Bigdeli, M. Haeri / ISA Transactions 47 (2008) 143–155
Fig. 16. CDM robustness index ratios γi /γi∗ (left column graphs) and CDM instability indices γi γi+1 (right column graphs) for various network parameter ratios;
(a) and (b) for N p /N p0 , (c) and (d) for C p /C p0 , (e) and (f) for d p /d p0 .
RED show oscillating behaviours in response to such variations, of cooperating with AQM algorithms employed in routers to
by employing CDM controller both data sending rate and queue support the QoS requirements. In this paper, at first the overall
length have been stabilized around their desired values with a nonlinear delayed differential equations of the dynamics model
relatively low marking/dropping probability. of closed loop system have been derived based on TCP Vegas
model. The model is then linearized to derive a transfer function
5.2.3. Exp. 3: Robustness analysis representation between the packet marking probability and the
As discussed in Section 4, robustness with respect to bottleneck router queue length as the input and output of the
network parameter variations is one of the major goals in modified TCP Vegas/AQM system. For the case of a network
AQM design. For CDM as an AQM method, the stability with homogeneous data sources sharing a single bottleneck,
and instability indices have been introduced in (30a) and an AQM controller has been designed based on Coefficient
(30b). Besides, robustness condition has been given in (30c). Diagram Method (CDM). The controller has been previously
Regarding this relationships, the robustness and stability employed by the authors as an AQM method for networks
properties can be investigated through investigating the index supporting TCP Reno and its out-performance with respect to
ratios of γi /γi∗ for i = 2, 3, 4 and instability indices of γi γi+1 the other controllers such as RED and PI has been shown. This
for i = 2, 3, 4. The graphs in Fig. 16 show the variations of good background motivated the authors to design the CDM
these index ratios and instability indices for different values of as AQM controller for networks supporting TCP Vegas. In
network parameters i.e. N p , C p , and d p . From these figures, this paper, the performance of the TCP Vegas/CDM networks
increase in N p and d p increases the robustness and stability has been evaluated through different simulation scenarios
of the system while increase in C p alleviates both of them. and its out-performance with respect to other AQM methods
However, the index ratios and the instability indices remain i.e. RED and PI (that are simulated for comparison) has been
more than 5 and 10 for all the investigated range for C p which is shown. It should be noted that designing any model-based
representative of very good margins of stability and robustness. AQM controller, and especially CDM, is strongly correlated
These values are pretty more for N p and d p as observed in the with TCP variant and its model. For example, the degree
Fig. 16, which guarantee good performance of the controller in of CDM polynomials and the selected CDM stability indices
a dynamically varying network. Note however that these results for a network with the same load, bandwidth, and delay
are in agreement with the simulation results in Exp. 1. parameters are not the same for networks supporting TCP
Vegas and networks supporting TCP Reno. However, the good
6. Conclusion performance of CDM with both types of TCPs is representative
of the good potential of CDM to be employed as an AQM
In this paper, a mathematical model and control theoretical controller.
framework for designing AQM controllers in networks In order to synthesize our simulation scenario, the properties
supporting TCP Vegas is introduced. We have emphasized on a of the traffic passing our campus router have been measured
modified TCP Vegas algorithm that can respond to congestion through some experiments. Packet length and round trip time
signals through ECN. By this property, Vegas achieves ability (RTT), traffic composition, and flow size and number of active
N. Bigdeli, M. Haeri / ISA Transactions 47 (2008) 143–155 155
flows are the considered features of the measured data. We have based congestion control. Computer Communications 2006;29(11):
analysed the trace files to derive three types of data: (1) the 1963–78.
[8] Srijith KN, Jacob L, Ananda AL. TCP Vegas-A: Improving the
average packet length to determine the bandwidth capacity in
performance of TCP Vegas. Computer Communications 2005;28(4):
terms of packets/s; (2) the experienced RTT to determine a 429–40.
reasonable value for the round trip propagation delay and round [9] Athuraliya S, Li VH, Low SH, Yin Q. REM: Active queue management.
trip time values; (3) the traffic mixture and flow size distribution IEEE Network 2001;15(3):48–53.
to determine an approximate number of active long-lived flows [10] Floyd S, Jacobson V. Random early detection gateways for congestion
and to validate our assumption about long-lived flows. Our avoidance. IEEE/ACM Transactions on Networking 1993;1:397–413.
[11] Misra V, Gong WB, Towsley D. Fluid-based analysis of a network of
analysis results show that more than 90% of passing IP traffic AQM routers supporting TCP flows with an application to RED. In: Proc.
comes from TCP sources while a low percent of TCP flows ACM/SIGCOMM 2000. p. 151–61.
really carry data. However, these flows called long-live flows [12] Hollot CV, Misra V, Towsley D, Gong WB. Analysis and design of
are dominant in flow size both in number of packets and volume controllers for AQM routers supporting TCP flows. IEEE Transactions
on Automatic Control 2002;47:945–59.
of bytes.
[13] Chang X, Muppala JK. A stable queue-based adaptive controller
Simulation results indicate that the CDM-AQM controller for improving AQM performance. Computer Networks 2006;50(13):
performs very well for network supporting TCP Vegas even in 2204–24.
the case of unknown network parameters or parameter varia- [14] Bigdeli N, Haeri M. Design of a robust AQM strategy for dynamic
tions. Simulation results illustrate capability of the controller to TCP/AQM networks based on CDM. In: Proc. of CCA 2005. p. 716–21.
[15] Bigdeli N, Haeri M. Adaptive-CDM: A new AQM controller to cope with
cope with the network variations. As another point, the stability
TCP/AQM networks dynamics. In: Proc. of ICICS 2005. p. 1520–4.
and robustness properties of the designed controller have been [16] Bigdeli N, Haeri M. Performance analysis of CDM as an AQM congestion
discussed in this paper that can help the designer to have a judg- controller in dynamic networks. In: Proc. of EuroCon 2005. p. 680–83.
ment about the performance margins of the designed controller. [17] Ramakrishnan K, Floyd S. A proposal to add explicit congestion
notification. Tech. rep. RFC, 2481. 1998.
Acknowledgements [18] Floyd S, Kohler E. Internet research needs better models. In: Proc. of first
workshop on hot topics in networks (HotNets-I) 2002.
[19] Choe H, Low SH. Stabilized vegas. In: Proc. of IEEE INFOCOM 2003.
Authors gratefully acknowledge the financial support of p. 2290–300.
the Iranian Telecommunication Research Centre (ITRC) under [20] Manabe S. The Coefficient diagram method. In: Proc. of the 14th IFAC
Grant T500-8210. The authors also deeply appreciate the symposium on automatic control in aerospace 1998. p. 199–210.
Information Technology Services centre of Sharif University [21] Manabe S. Coefficient diagram method as applied to attitude control
of controlled-bias-satellite. In: Proc. of the 13th IFAC symposium on
of Technology for supporting them with the measurement
automatic control in aerospace 1994. p. 322–7.
facilities. [22] Manabe S. Brief tutorial and survey of coefficient diagram method. In:
Proc. of the 4th Asian control conference proceedings 2002. p. 1161–6.
References [23] Hamamci SE. Simple polynomial controller design by the coefficient
diagram method. WSEAS Transactions on Circuits and Systems 2004;
[1] Moronski J. Freeze TCP: A true end-to-end TCP enhancement mechanism 3(4):951–6.
for mobile environment. In: Proc. of IEEE INFOCOM 2000. p. 1537–45. [24] Manabe S. Controller design of two-mass resonant system by coefficient
[2] Brakmo LS, Peterson LL. TCP Vegas: end to end congestion avoidance diagram method. The Transactions of The Institute of Electrical Engineers
on a global internet. IEEE Journal on Selected Areas on Communications of Japan 1998;118-D: 1(1):58–66.
1995;13(8):1465–80. [25] Lipatov A, Sokhlov N. Some sufficient conditions for stability and
[3] Bonald T. Comparison of TCP Reno and TCP Vegas via fluid instability of continuous linear stationary systems. Automatic Remote
approximation. Tech. rep. RR 3563. 1998. Control 1978;39:1285–91.
[4] Mo J, La RJ, Anantharam RJ, Walrand JC. Analysis and comparison of [26] Hamamci SE, Ucar A. A robust model-based control for uncertain
TCP Reno and Vegas. In: Proc. of IEEE INFOCOM 1999. p. 1556–63. systems. Transaction Institute of Measurement and Control 2002;24(4):
[5] Low SH, Peterson LL, Wang L. Understanding TCP Vegas: A duality 331–45.
model. Journal of ACM 2002;49(2):207–35. [27] Hamamci SE. A robust polynomial-based control for stable processes with
[6] Kim K, Choi CH. Queue delay estimation and its application to TCP time delay. Electrical Engineering 2005;87(3):163–72.
Vegas. Computer Networks 2003;43(5):619–31. [28] Goodwin CG, Graebe SF, Salgado ME. Control System Design. New
[7] Haeri M, Mohsenian Rad AH. Adaptive model predictive TCP delay- Jersey (USA): Prentice-Hall, Inc.; 2001.