Proceedings of the 42nd IEEE
Conferenceon Decision and Control
M w i , Hawaii USA, December 2003
Robust Stability and Disturbance Attenuation Analysis of a Class of Networked Control Systems
Hai Lin Guisheng Zhai
Department of Electrical Engineering Department of Opto-Mechanronics
University of Notre Dame Wakayama University
Notre Dame, IN 46556, USA Sakaedani, Wakayama 640-8510, Japan
Abstract-In this paper, stability and disturbance attenua-
tion issues for a class of Networked Control Systems (NCSs)
under uncertain access delay and packet dropout elTects are
considered. Our aim is to find conditions on the delay and
packet dropout rate, under which the system stability and
71" disturbance attenuation properties are preserved to a
desired level. The basic idea in this paper is to formulate such
Networked Control System as a discrete-time switched system.
Then the NCSs' stability and performance problems can be
reduced to corresponding problems for the switched systems,
which have been studied for decades and for which a number
of results are available in the literature. The techniques in this
paper are based on recent progress in the discrete-time switched
systems and piecewise Lyapunov functions.
I. INTRODUCTION
By Networked Control Systems (NCSs), we mean feed-
hack control systems where networks, typically digital band-
limited serial communication channels, are used for the
connections between spatially distributed system components
like sensors and actuators to controllers. These channels may
be shared by other feedback control loops. In traditional feed-
hack control systems these connections are through point-to-
point cables. Compared with the point-to-point cables, the
introduction of serial communication networks has many
advantages, such as high system testability and resource
utilization, as well as low weight, space, power and wiring
requirements [9], 1141. These advantages make the networks
connecting sensorslactuators to controllers more and more
popular in many applications, including traffic control, satel-
lite clusters, mobile robotics etc. Recently modeling, analysis
and control of networked control systems with limited com-
munication capability has emerged as a topic of significant
interest to control community, see for example [4], [3], [6],
~ 4 1 [I].
. t91.
Time delay typically has negative effects on the Networked
Control Systems' stability and performance. There are sev-
eral situations where time delay may arise. First, transmission
delay is caused by the limited bit rate of the communication
channels. Secondly, the channel in NCSs is usually shared
by multiple sources of data, and the channel is usually
multiplexed by time-division method. Therefore, there are
delays caused by a node waiting to send out a message
through a busy channel, which is usually called accessing
delay and serves as the main source of delays in NCSs. There
are also some delays caused by processing and propagation,
which are usually negligible for NCSs. Another interesting
0-7803-7924-1103/$17.0002003 IEEE 1182
TuP06-2
Panos J. Antsaklis
Department of Electrical Engineering
University of Notre Dame
Notre Dame, IN 46556, USA
problem in NCSs is the packet dropout issue. Because of
the uncertainties and noise in the communication channels,
there may exist unavoidable errors in the transmitted packet
or even losst. If this happens, the conuptedpacket is dropped
and the receiver (controller or actuator) uses the packet
that it received most recently. In addition, packet dropout
may occur when one packet, say sampled values from the
sensor, reaches the destination later than its successors. In
such situation, the old packet is dropped, and its successive
packet is used instead. There is another important issue in
NCSs, that is the quantization effect. With finite hit-rate
constraints, quantization has to be taken into consideration
in NCSs. Therefore, quantization and limited bit rate issues
have attracted many researchers' attention, see for example
[3], [6], 141. It has been known [3], [6] that an exponential
data representation scheme is most desirable under certain
conditions.
In this paper, we consider uncertain time delay and packet
dropout issues of NCSs in the framework of switched sys-
tems. The strength of this approach comes from the solid
theoretic results existing in the literature for stability, robust
performance etc. for switched systems. By a switched system,
we mean a hybrid dynamical system consisting of a finite
number of subsystems described by differential or difference
equations and a logical rule that orchestrates switching he-
tween these subsystems. Properties of this type of model have
been studied for the past fifty years t o consider engineering
systems that contain relays and/or hysteresis. Recently, there
has been increasing interest in the stability analysis and
switching control design of switched systems (see, e.g., [IO],
[SI, [ l l ] , [12] and the references cited therein). Notice that
hybridhwitched system research has provided useful results
and promising techniques to deal with NCSs with delay and
packet dropout effects. For example, in [14], the multiple
Lyapunov function method [2] was employed to analyze the
stability of NCSs with network-induced delay. In addition,
the packet dropout effects in NCSs are closely related to
the controller failures studied in the fault tolerance literature.
There are some recent work on the controller failure analysis
based on switched system techniques 11 I], [13]. By using a
piecewise Lyapunov function, the author in [I31 showed that
if the controller failures did not occur too frequently or last
'Error control coding andlor Automatic Repeal ~CQUCSI
(ARQ) mecha-
nism may be employed, but the possibility of e m r occurring still exisrs.
too long, then global exponential stability and disturbance
attenuation property of the system were guaranteed to be
preserved.
In this paper, we investigate the robust stability analysis
and disturbance attenuation problem for a class of Networked
Control Systems (NCSs) under uncertain access delay and
packet dropout effects. Our aim is to find conditions con-
cerning the delay and packet dropout rate, under which the
system stability and 31" disturbance attenuation properties
are preserved to a desired level. We first analyze the nature
of the uncertain access delay and packet dropout effects on
NCSs in Section 11. Then in Section 111, we model the NCS
as a discrete-time switched system. Therefore the NCSs'
robust stability and performance problems can be boiled
down to the stability analysis and disturbance attenuation
problems of switched systems. In Section IV, the robust
stability problem for such NCSs with uncertain access de-
lay and packet dropout effects is studied, and disturbance
attenuation properties for such NCSs are studied in Section
V. The techniques employed in this paper are based on recent
progress in the continuous-time and discrete-time switched
systems [ 121, [ll]. Finally, concluding remarks are presented.
11. ACCESS DELAY AND PACKET DROPOUT
We assume that the communication scheme has the fol-
lowing properties. For the network link layer, the delays
caused by processing and propagation are ignored, and we
only consider the access delay which serves as the main
source of delays in NCSs. Dependent on data traffic, the
communication bus is either busy or idle (available). If
available, communication between sender and receiver is
instantaneous. Errors may occur during the communication
and destroy the packet, and this is considered as a packet
dropout.
The model of the NCS discussed in this paper is shown
in Figure 1 . For simplicity, but without loss of generality,
we may combine all the time delay and packet dropout
effects into the sensor to controller path and assume that
the controller-actuator communicates ideally.
We assume that the plant can be modeled as a continuous-
time linear time-invariant system described by
= Aez(t) + BCu(t)+ E c d ( t )
{ 5z ((tt)) = CCz(t)
where z ( t )E W" is the state variable, u ( t ) E W"' is control
input, and z ( t ) E WP is the controlled output. The disturbance
input d ( t ) is contained in D c R'. AC E WnX",Be E RnXm
and EC E RnX' are constant matrices related to the system
state, and C e E RPxn is the output matrix.
For the above NCS, it is assumed that the plant output
node (sensor) is clock-driven. In other words, after each clock
cycle (sampling time Ts),the output node attempts to send Fig. 2. The illustration of uncertain time delay and packet dropout of
Networked Control Systems.
a packet containing the most recent state (output) samples.
If the communication bus is idle, then the packet will be
1183
I
j
i--
Fig. 1. The Networked Control Systems' model.
transmitted to the controller. Else if the bus is busy, then the
output node will wait sometime, say w < T,, and try again.
After several attempts or when newer sampled data becomes
available, if the transmission still can not be completed, then
the packet is discarded, which is also considered as a packet
dropout. On the other hand, the controller and actuator are
event driven and work in a simpler way. The controller, as
a receiver, has a receiving buffer which contains the most
recently received data packet from the sensors (the overflow
of the buffer may be dealt with as packet dropouts). The
controller reads the buffer periodically at a higher frequency
than the sampling frequency, say every % for some integer
N large enough. Whenever there is new data in the buffer,
then the controller will calculate the new control signal and
transmit to the actuator. According to the assumption, the
controller-actuator communicates without delay or packet
dropout. Upon the new control signal arrival, the actuator
will update the output of the Zero-Order-Hold (ZOH) to the
new value.
Based on the above assumptions and discussion, the time
delay and packet dropout pattern can be shown in Figure 2. In
this figure, The small bullet, 0 , stands for the packet being
transmitted successfully from the sensor to the controller's
receiving buffer, maybe with some delay, and being read
by the controller, at some time t = kT, +~ 5 , and the
new control signal is updated in the actuator instantly. The
actuator will hold this new value until next update control
signal comes. The symbsl, 0 , denotes the packet being
dropped, due to error, bus busy, conflict or buffer overflow
etc.
0 Ts "TS
 111. MODELS FOR NETWORKED CONTROL Then, we may plug in the u[k]= K z [ k ] and get
SYSTEMS
+
z [ k 11
In this section, we will consider the sampled-data model of
the plant. Because we d o not assume the synchronization be-
+
= As[k] K B K z [- + +
~ 11 ( N - t c ) B K z [ k ] Ed[k]
tween the sampler and the digital controller, the control signal + +
= [ A ( N - n ) B K ] z [ k ] ~ t B K z [- +
k 11 Ed[k]
is no longer of constant value within a sampling period.
Therefore the control signal within a sampling period has to
If we let 4 k j = [ z[til1, then the above equa-
be divided into subintervals corresponding to the controller's tions can be written as:
reading buffer period, T = 3.
Within each subinterval.
the control signal is constant under the assumptions of the
previous section. Hence the continuous-time plant may be
discretized into the following sampled-data systems:

z[k+ 11 = Az[k] + [BB-1 ... B]

r

1
u l ~ k1~
"'.~ I
1
+ E d [ k ] (I)
where n = 0 , 1 , 2 , . . . ,D,,,. Note that K = 0 implies
r = 0, which corresponds to the previous "no delay"
case. And the controlled output z [ k ] is given by
where A = eAPT*, B = &%eAC"BCdq and E =
J,'. eA'qECdq. Note that for linear time-invariant plant and z [ k ]= [ 0 c ] ?[k]
constant-periodic sampling, the matrices A, B and E are where C = Cc.
constant. 3) If packet dropout happens, due to corrupted packet or
During each sampling period, several different cases may delay T > Dmo,x %, then the actuator will implement
arise, which can be listed as follows. the previous control signal, i.e. u'[k] = ~ ' [ k=] . . . =
u N [ k ]= u [ k - 11, then the state transition equation (1)
1) No delay, r = 0, u'[k] = u 2 [ k ]= ' . . = u N [ k ]= u [ k ] , for this case can be written as follows.
then the state transition equation ( I ) for this case can
be written as follows.
r ulkl 1
1 r'lk- 11 1
= Az[k] + N . Bu[k - 11t Ed[k]
Using the same variable transformation as in the above
+ Ed[k] case, we get
= Azjk] + N . Bulk] + Ed[k]
2) Delay 7 = K x %,
where n = 1 , 2 . . . . ,D,,,. For
this case u'[k] = u'[k] = . . . = un[k] = ~ [ -k 11,
uK+'[k]= u"+'[k] = . .. = u N [ k ] = u[k],and
Equation (1) can be written as:
The controlled output z [ k ] is given by

z [ k ]= [0 c ] ?[k]
z[k + 11 = A i [ k ]+ [ B B . .. B ] I "[:[kjll I where C = Cc.
In the next section, we will formulate the above NCSs as a
class of discrete-time switched systems.
IV. STABILITY ANALYSIS
+ Ed[k] Motivated by the above analysis of NCSs, we introduce
a family of discrete-time linear systems described by the
= Az[k] + K . B u [ k - 11 +(N- K) +
B u [ k ] Ed[k] following difference equations.
Let us assume that the controller uses just the time-
invariant linear feedback control law, u [ k ] = K z [ k ] ,
+
z [ k 11 = A , z [ k ] E,d[k], k E Z++ (2)
which may he obtained as the solution of a LQR prob- where z [ k ] E E%" is the state variable, and the disturbance
lem without considering the network induced effects. input d [ k ] is contained in D ' C .'%E A, E RnX"and E, E

1184
R””‘ are constant matrices indexed by q E Q, where the
finite set Q = { q l . 42,. . . ,qn} is called the set of modes.
Combine the family of discrete-time uncertain linear sys-
-
tems (2) with a class of piecewise constant functions of time
U : Z+ Q. Then we can define the following linear time-
varying system as a discrete-time switched linear system
+ +
~ [ kl ] = A , j k ] ~ [ k ] E,lqd[k], k E Z+ (3)
The signal u [ k ] is called a switching signal.
Associated with the switched system (3), a controlled
output r [ k ] is considered.
4kl = C,[klz[kl
where Calk] E RPYnand t j k ] E RP.
For the NCS we considered in this paper, we may formu-
late it as a switched system with D,, +
2 different modes,
which can he expressed as follows.
t [ k + I] = Ar?[kj + E,d[k]
{ z [ k ] = C,?[k]
(4)
[0 I
w h e r e A K = KBK . A ~ + ( N - K ) B K ] ’ ~ ~ =
and C, = I 0 C 1 for K = 0 . 1.. 2., . ’ ’ D,,,. . N . And the
[ L]
set of moles Q is hiven by Q = { 0 , 1 , 2 , .. D,,,, N}.
Note that K = 0 implies T = 0, which corresponds to the
“no delay” case, while K. = N corresponds to the “packet
dropout” case.
It is reasonable to assume that, for the cases of no delay
(n = 0) or small delay ( R 5 no), the corresponding
state matrix As’s are Schur stable, while, for the cases of
large delay ( K > K O ) or packet dropout ( K = N ) , the
Ax’s are not Schur stable. Therefore, in this paper it is
assumed that the first r, corresponding to K O , of all the
+
DmaZ 2 matrices in { A , } are Schur stable, while the
rest matrices are not Schur stable, where T 5 D,,, + 2
and K E Q = { 0 , 1 , 2 , .. . ,D,,, N}. In the sequel, for
simplicity of nutation, we will index the switched NCS model
with i, for i E Q = {O, 1 , 2 , ’ . . , D,,, N}. In this section,
we set d[k]= 0 in (4) for the purpose to study its stability.
It is known that for Schur stable systems z [ k + l ] = A , z [ k ] ,
there always exist positive scalars XI < 1 and hi’s, i 5 T
such that llAtll 5 h;X; for any k 2 1’. Note that for any
+
Schur unstable system z [ k 11 = A i z [ k ] (i > T ) , there
always exist a constant 0 < U < 1 making the system
+
z [ k 11 = u A i z [ k ] Schur stable. Hence we may assume
that there exist positive scalars Xz 2 1 and hi’s, i > T such
that llA:11 5 h;X$ for any k 2 1. Therefore, we get
Following [12], we introduce the notations as below. Denote
h = maxi{hi}.For any switching signal u ( k ) and any kZ >
’The vector/mauir norm considered here is the 1’ norm and its induced
matrix norm.
1185
kl > 0, let N o ( k l ,kz) denote the number of switchings of
u ( k ) on the interval [kl,kz). Let K , ( k l .kz) denote the total
period that the i-th subsystem is activated during [ k l , k z ) .
Define K - ( k I , kZ) = xis?, i E Q K,(kl, kz), which stands
for the total activation period of the Schur stable subsystems.
On the other hand, K + ( k l , k?) = xi>v, iGQ Ki(k1, kz)
denotes the total activation period of the Schur unstable
+
subsystems. We have K - ( k l , k2) K + ( k l l k2) = kz - k 1.
For given NO 2 0, T ~ let . Sa(ra) denote the set of all
switching signals satisfying
k
No(0. k ) 5 No +-
Ta
(6)
where the constant ra is called the average dwell time and
NO the chatter bound. The idea is that there may exist
consecutive switching separated by less than T,, hut the
average time interval between consecutive switcbings is not
less than ra. Note that the concept of average dwell time
between subsystems was originally proposed for continuous-
time switched systems in [8]. With these assumptions and
notations, we may apply the techniques and results developed
in [12] to the NCSs and get the following theorem for
globally exponential stability. The proof of the theorem is
not difficult by using the technique of Theorem 3 in [121,
and thus is omitted here.
Theorem I: For any given X E AI,^), the NCS (4) is
globally exponentially stable with stability degree X if there
exists a finite constant r,‘ and A’ E (X1,X) such that the
K+(O,k ) and “(0, k ) satisfy the following two conditions
a
I ) infk>o K O k 2 1”A ,nA3-l,AL
-1nA. holds for some scalar
A’ E ( A I , A);
2) The. average dwell time is not smaller than T;, i.e.
N,(O, k ) 5 NO + 6,where 7,’ = *, and NO
may he specified argitrarily.
Remark I : The first condition implies that if we expect the
entire system to have decay rate A, we should restrict the total
number of lost packets and large delay packets in the sense
that on average-K+(O, k) has &I upper-bound, K+(O,k ) 5
InX*-InAlk
lnA?-l”A, ‘
Remark 2: The main point of the second condition can
he described as follows. Although the first condition may
he satisfied, which means that on average the packet lost
is limited and the total number of large delayed packet is
hounded, in the worst case the packet dropout and large
access delay happen in a burst fashion. For such worst case,
the NCSs may fail to achieve the decay rate. The second
condition restricts the frequency of the packet dropout and
large delayed packet, and to make sure the above worst case
can not happen.
Remark 3: The above theorem says that the NCSs’ sta-
bility, with most of the packets arriving in a timely fashion,
does not degenerate seriously, which is reasonable.
 V. DISTURBANCE ATTENUATION ANALYSIS where r ( j ) = z T [ j ] z [ j-
] $ d T [ j ] d [ j ] .Since V ( k i ) 5
In this section, we will study the disturbance attenuation pV(kty) holds on every switching point k;, we obtain by
induction that
property for the NCSs (4). Note that the Cz gain property
of discrete-time switched systems was studied in [I21 under "(k) 5 -
wNm(O,k)AZK-(O.L) A+2 K + ( O , k ) V (0)
the assumption that all subsystems were Schur stable. In this k--1

section, we will extend the CZ gain property of discrete-time - +

-Z K ~ ( j , k - - 1 ) X 2 K + ( j . k - - l ) ~
PN~(j,k-1)~ (i)
j=0
switched system to the case that not all subsystems are Schur
k--1
stable. The techniques used in this section are similar to those
in [l I ] for continuous-time switched systems.
- - + r(i)
P N ~ ( j , k - l ) ~-Z K - ( j , t - l ) ~ Z K + ( j . k - - l )
j=O
Following the assumptions in [12], the initial state is
assumed to be the origin, z[O] = 0. And we assume that The last equality is because of the zero initial state assump-
the Schur stable subsystems achieve an Cz gain smaller than tion z[O] = 0.
^io.It is known that there exist a positive scalar A- < 1 and We assume that on any interval ( k l , k2) the total activation
a set of positive definite matrices Pi, for i 5 r and i E Q, periods of the unstable subsystems satisfies K + ( k l , k z ) 5
such that 1" X--In A-
,nA+-lnX- (kz - h),or equivalently
ArP,Ai-XzP;+CrC,+ArP,E,(,r-E~PiE,)-'E'P,Ai <0
holds [7]. Observing that for Schur unstable subsystems,
there always exist a constant 0 < U < 1, such that the
subsystems (U&, Ei>uC,) can achieve the C2 gain level -(D.
Therefore, we assume that for Schur unstable subsystems
holds for some scalar A' E (A-, 1) and Vk2 > kl 2 0. Then
there exist a positive scalar A+ 2 1 and a set of positive we get
definite matrices Pi, for i > T and i E Q, such that A2K- ( j , k - I ) A Z K + ( i . k - l )
ArP;Ai-X:P,+CrC,+ArP;E;(~~r-E~PiEj)-'E~P<
, . A0~
- +
* ZK-(j,k-l)+ZK+(jlk-l) = (r)Z(k-l-j)
5 (A)
Using the solution P;'s,we define the following piecewise
Ljapwiov function candidate Therefore, we get
k-1
V ( k )= V m [ k ] (=
z ) zT[kIPr[k,"[kl (7) V ( k )5 -CpNd%--l) (A*)z(k-I-j)r(j) (11)
for the switched system, where is switched among j=O
the solution Pi's in accordance with the piecewise constant When .p = 1, we get from V ( k )2 0 and (1 I ) that
switching signal u[k].It can be shown as in [I21 that there
always exist constant scalars alra2z 0, for example,
011 = infiGQA,,,(P,), az = supiGQA>,(P,), such that
~ p N ~ ( j , k - l ) ( X * ~ z ( k - l - ~ ) 0
r ( ~ ~ I (12)
j=O

aillzllz I I o ~ z I I ~VZ
I I ~E, W", V i E Q (8) We sum (12) from k = 1 to k = +m to obtain
Here A,,(P,) and A,(P,) denotes the largest and smallest
eigenvalue of Pi respectively. There exist a constant scalar
/L 2 1 such that

V,(z)5 pv,(z): vz ER", v2,j E Q (9)

A conservative choice is p = supk,lGq
m
w.
1)
Following the steps in [121, for each V,(z) = z T [ k ] P , z [ k ]
along the solutions of the corresponding subsystem, we may
obtain that
which means
V , ( z [ k+ 11) - V , ( z [ k l )

4 - ( 1 - X Z ) V , ( z [ k ]-
-(1 - A 3 V , ( o [ k ] -
) z ~ [ k l z [ +rodT[kld[k]
k]
) r T [ k ] z [ k+?odT[k]d[k]
For a piecewise constant switching signal u[k]and any
]
Therefore, Cz gain 70is achieved for the switched system,
given integer k > 0, we let k1 < . . . < ki (i 2 1) denote the namely the NCS (4).
switching points of u [ k ]over the interval [0, k). Then, using For the case fi > 1, we multiply both sides of ( 1 1) by
the above difference inequalities, we obtain p-No(OJ-l) to get
~ k - 1 -iVe(o,j)(A*)Z(k-l-j) T '
OP 1314j1 (14)
7;X,=o w
@1 -N,(O.j)(A*)Z(k-l--j)dT[.
I 3ldijl

1186
Now, we choose a positive scalar X larger than 1 to consider
the following average dwell time condition: for any positive
integer j > 0,
Therefore f i - N ” ( o J ) > X-23 holds for any j > 0, where
X = P (zTc)-l. Then, from (14) we obtain
Similarly, we sum both sides of the above inequality from
k = 1 to k = +m to get
and thus
3 4 3=1
holds for any d[k] E &[O, +m). Following the notation in
[IZ],we say that a weighted LZ gain ’yo is achieved. In
summary, we prove the following theorem.
77ieorern 2: The NCS (4) achieves a weighted Lz gain
‘yo if the K + ( k l , k z ) satisfies (IO) and N,(O,k) satisfy the
condition of (15).
Remark 4; Similarly, the condition (10) resuicts the num-
ber of the packet dropout and large delayed packet, while
the condition ( I 5 ) restricts the happening frequency of them.
Both of the conditions are given in the sense of average over
time.
VI. CONCLUDING REMARKS
In this paper, we modeled a class of NCSs under uncertain
access delay and packet dropout as discrete-time linear
switched systems. The stability and disturbance attenuation
issues for such NCSs were studied in the framework of
switched systems. The strength of this approach comes
from the solid theoretic results existing in the literature of
switchedhybrid systems. It was shown that the exponential
stability and disturbance attenuation level might be preserved
for the NCSs under certain bounds on the amount and rate
of the dropped and large delayed packets. Although we only
consider state feedback control law here, the techniques and
results developed here can be easily extended to the case of
static output feedback control law. It should be pointed out
that the conditions concerning the delay and packet dropout
rate for the preservation of the NCSs’ stability and ‘H”
disturbance attenuation properties were based on Lyapunov
theory. Therefore, the conditions are sufficient only and
maybe conservative for some cases.
1187
VII. ACKNOWLEDGMENTS
The partial support of the National Science Foundation
(NSF ECS99-12458 & CCROI-13131) is gratefully acknowl-
edged. The first author appreciates the support from Center
of Applied Mathematics Fellowship (2003-04), University of
Notre Dame.
VIII. REFERENCES
[I] P. Bauer, C. Lorand, M. Sichitiu, and K. Premaratne,
“Total delay compensation in LAN control systems and
implications for scheduling,” in Pmceedings of 2001
American Control Conference, pp. 4300-4305, Arlington,
VA, 2001.
121 M. S. Branicky, “Multiple Lyapunov functions and other
analysis tools for switched and hydrid systems,” IEEE
Trnnsacrions on Auronmric Conrrol, vol. 43, no. 4, pp.
475-482, 1998.
[3] R. W. Bmkett, and D. Liberzon, “Quantized feedback
stabilization of linear systems,” IEEE Transacrions on
Aurornaric Conrml, vol. 45, no. 7, pp. 1279-1289, 2000.
141 D. E Delchamps, “Stabilizing a linear system with quan-
tized state feedback:’ IEEE Transactions on Autosuric
Control, vol. 35, no. 8, pp. 916-924, 1990.
[5] R. A. Decarlo, M. S. Branicky, S. Pettenson, and B.
Lennmson. “Penoectives and results on the stabilitv
and stabilizability bf hybrid systems:’ Proccedings of &
IEEE, vol. 88, no. 7. !m. 1069-1082. 2000.
[6] N.Elia. and S. K. Miiter, ”Stabilization of linear systems
with limited information,” IEEE Transactions on Aulo-
malic Conrml, vol. 46, no. 9, pp. 1384-1400, 2001.
[7] P. Gahinet, and P. Apkarian, “A linear matrix inequality
approach to H , control,” Intemarional Journal of Robust
ond Nonlinear Conrrol, vol. 4, pp. 421-448, 1994.
[8] I. P. Hespanha, and A. S. Morse, “Stability of Switched
Systems with Average Dwell-time,’’ in Proceedings of {he
38th Conference on Decision and Control, pp. 2655-2660,
Dec. 1999.
[9] H. Ishii, and B. A. Francis, Limited Dura Rare in Control
Sjsrerns wirh Network, Lecture Notes in Control and
Information Sciences. vol. 275, Springer, Berlin, 2002.
[IO] D. Liberzon, and A. S. Morse, “Basic problems in stability
and design of switched systems,” IEEE Conrml Sysrerns
Magazine, vol. 19, no. 15, pp. 59-70. 1999.
[Ill G. Zhai, B. Hu, K. Yasuda, and A. N. Michel, “Distur-
bance attenuation properties of time-controlled switched
systems:’ Journal of rhe Franklin Insrirure, vol. 338, pp.
765-779, 2001.
[I21 G. Zhai, B. Hu, K. Yasuda, and A. N.Michel, “Qualitative
analvsis of discrete-time switched svstems:’ in Pmcecd-
ings of 2W2 American Conrml Confemmce, vol. 3, pp.
1880-1885. 2002.
1131 G. Zhai, ”Stability and H , Disturbance Attenuation
Analysis for LTI Control Systems with Controller Fail-
ures,” in Pmceedings of rhe 41st Conference on Decision
and Conrml. pp. 3869-3874, 2002.
[I41 W. Zhang, M. S. Branicky, and S. M. Phillips, “Stability
of networked control systems:’ IEEE Conrml Sysrems
Magazine, vol. 21, no. I , pp. 84-99, 2001.