A Distributed Routing Algorithm for Supporting Connection-Oriented Service in Wireless Networks with Time-Varying Connectivity

brrrclergymanNetworking and Communications

Jul 18, 2012 (5 years and 4 months ago)

304 views

A Distributed Routing Algorithmfor Supporting Connection-Oriented
Service in Wireless Networks with Time-Varying Connectivity
￿
Anastassios Michail
Department of Electrical Engineering and
Institute for Systems Research
University of Maryland at College Park
College Park,MD20742,USA
tassos@eng.umd.edu
Anthony Ephremides
Department of Electrical Engineering and
Institute for Systems Research
University of Maryland at College Park
College Park,MD 20742,USA
tony@eng.umd.edu
Abstract
We develop and simulate a distributed dynamic routing
algorithm,capable of identifying paths for establishing and
maintaining connection-oriented sessions in wireless com-
munication networks which are characterized by frequent
and unpredictable changes in connectivity.Our approach is
a new protocol which runs atop a protocol for connection-
less datagramservice and establishes circuit routes for ini-
tial connection based on a mechanism of short packets ex-
change and on distributed information about availability of
network resources.We explore the idea of predictive re-
routing in that the algorithm takes advantage of the possi-
bility to convert a connectivity change into a “soft” failure
to maintain and re-route on-going sessions.The algorithm
is simulated in Opnet and results show that the “soften-
ing” of link failures can improve performance as captured
in terms of new call blocking probability and probability of
forced termination of on-going sessions.
1.Introduction
We consider the problem of dynamic routing in mobile
wireless networks which are characterized by frequent and
unpredictable changes in connectivity.We examine ad-hoc
networks consisting only of mobile wireless communica-
tion nodes which move arbitrarily and do not pertain to any
form of fixed network architecture,such as that of cellular
networks.The existence of communication links between
nodes depends on nodes’ location and distance,and on vari-
ous physical layer factors such as transmission power levels,
￿
Prepared through collaborative participation in the Advanced
Telecommunications/Information Distribution Research Program(ATIRP)
Consortium sponsored by the U.S.Army Research Laboratory under Co-
operative Agreement DAAL01-96-2-0002.
channel interference,antenna patterns and multipath propa-
gation effects.
Prior work [1,2,4,5] has focused on connection-
less datagram service networks.The shift to connection-
oriented type of service dictates the use of routing algo-
rithms that can satisfy three main objectives:(1) discover
routes as fast as possible,(2) establish connections between
pairs of nodes through the discovered paths by reserving
network resources and (3) react to connectivity fluctuations
and maintain ongoing sessions,to the extent this can be fea-
sible.Under these circumstances solutions to the routing
problemshould be provided by algorithms that execute in a
distributed fashion without any need for global connectivity
information.
To these ends we have developed a fully distributed algo-
rithm for identifying and maintaining paths between com-
municating pairs of nodes,based on a mechanism of short
packets exchange.The starting point of our approach is
a distributed routing algorithm for connection-less data-
gram service that was presented in [1] and to which we
will refer further on as CE algorithm.This algorithm as-
sumes adequate bandwidth and an underlying access pro-
tocol,which is interference free.It relies on the execution
by all nodes and separately for each potential destination
of a two-phase procedure.The first phase (“query-reply”
message exchange) results in the establishment of a directed
graph on a subset of the network that is “rooted” at the des-
tination node and thus provides an initial set of routes.The
second phase involves a similar structured exchange of con-
trol messages to react to the “failure” of an element of the
previously established route and is intended to discover a
new“by-pass” route.The algorithmadapts to large amounts
of topological changes with no need for global topologi-
cal knowledge,by building routes only as needed,instead
of maintaining routes from all nodes to the potential des-
tination nodes.Additionally,the extra routes built during
the “query-reply” phase increase protocol reliability.The
most important properties of the CE algorithm encompass
discovery of loop-free routes,deadlock-free operation and
capability of detecting catastrophic network partitions.
Shifting our focus to session-oriented service and,as is
natural in this case,to the core of limited available band-
width,we propose a new algorithm that establishes circuit
routes for initial connection in the same manner of the CE
data algorithmbut that reacts differently,as it should,in the
case of a link failure or connectivity change.The algorithm
maintains the main properties of the CE routing algorithm
but is modified to include in the message exchange addi-
tional information on the availability of network resources.
In addition,in order to maintain on-going sessions in the
presence of topological changes the algorithmtakes advan-
tage of the possibility to convert a wireless connectivity
change to a “soft” failure.
In the next section we present algorithmoperation,start-
ing with an overview and continueing with a definition of
the network model,the router structure and a detailed de-
scription of the algorithm execution rules.In section 3 we
discuss performance evaluation and simulation results of
the algorithmand we conclude in section 4 with a short dis-
cussion of the main results and plans for future work.
2.Algorithmdescription
2.1.Overview
We consider multiple independently executing versions
of the distributed routing CE algorithm,each one running
for a specific destination node.The new algorithm for
connection-oriented service runs atop the CE protocol and
utilizes route information to establish newconnections.The
existing or new routes are explored in a hop-by-hop packet
transmission mechanismin search of those paths which will
guarantee admission of the request.As the CE protocol re-
acts to connectivity changes by reorganizing its routes,the
overlaid protocol updates its information to be used either
for attempts to accommodate newconnection requests or to
dynamically re-route on going sessions which experience
quality degradation as a result of the changes in connectiv-
ity.
2.2.Network model
We model the network with a graph
 ￿ ￿  ￿  ￿
,where

represents the finite set of nodes and

the set of commu-
nication links.Each node
 ￿ 
has a unique node identifier
(ID) and each link
￿ ￿  ￿ ￿ 
can be used either for one-way
or two-way communication between nodes

and

,depend-
ing on the circumstances.All nodes are mobile and connec-
tivity changes continuously,resulting in a time-varying set
of communication links

.Each active link
￿ ￿  ￿ ￿ 
can
either be undirected or directed.If the latter is true,then if
the link is directed from

to

,node

is characterized as
downstream(DN) neighbor of

.Similarly,if the link is di-
rected from

to

node

is an upstream (UP) neighbor of

.
An underlying link-level protocol is assumed which as-
sures distributed knowledge of the changes in connectiv-
ity,in the sense that each node

is aware of all its adja-
cent nodes at all times,which are referred to as its neigh-
bors.The set of the neighbors of node

varies also with
time.We assume for simplicity that transmitted packets are
received correctly and simultaneous two-way transmission
over a link that would cause interference does not occur.
The detailed mechanism of this link-level protocol opera-
tion is beyond the scope of this paper and will not be ad-
dressed at the present time.
2.3.Router structure
Every node has a fixed number of transmitter-receiver
pairs (transceivers) used to set up communication links with
other nodes.The number of transceivers that are in an
IDLE state,varies dynamically with time depending on the
traffic load and the average session duration.A necessary
condition for a new connection to be established is that at
least one transceiver is available at every node in the path
to the destination.New connection requests are blocked
when one or more nodes along the path do not have any of
their transceivers idle.A pure first-come first-serve policy
is assumed without considering any preemptive policies in
which a high priority session can preempt an ongoing ses-
sion of lower priority.For simplicity no priority is given to
hand-off requests which have to compete against new con-
nection requests in search of communication paths.Amod-
ified version of our algorithmin which hand-off requests are
given priority over newcall arrivals was presented in [3].
Sessions are distinguished by a unique ID,a numeric
triple consisting of the source and destination IDs and a
counter which is incremented by one for every newconnec-
tion between the same source–destination pair.Each node
maintains a “Connectivity Table” with information on all
sessions for which it serves as a source,destination or in-
termediate relay node.The connectivity table keeps a sep-
arate entry for each transceiver with the transceiver ID,the
session ID,the incoming and outgoing link indices and the
status of the communication transceiver.
2.4.Algorithmexecution
Algorithmexecution can be viewed as occurring in three
logical phases,the “Construction phase”,the “Maintenance
phase” and the “Termination Phase”,which execute simul-
taneously in a dynamic topology.
A.Construction phase
During the construction phase mobile nodes desiring com-
munication with other mobile users in the network,place
connection requests which “travel” along paths provided by
the underlying CE algorithm that terminate at the destina-
tion node.A request may be admitted and the session will
be established if the chosen path can provide the sufficient
network resources required for end-to-end flowof informa-
tion.
Without loss of generality we assume that at some point
in time a node not adjacent to the destination node,DEST,
desires a connection.We also assume for simplicity that the
QRY-RPY process (construction phase of CE) has already
occurred and the part of the network under discussion can
be represented by a directed acyclic graph
 ￿ 
rooted
at the DEST (figure 1(a)).Hence any node in this graph
will always have at least one DNneighbor and by properties
of the CE protocol there definitely exists at least one route
initiating at any source node and terminating at the DEST.
A source node which desires a connection to the DEST
transmits a “Connection-Request” (CR) packet along one of
the existing DN links.If multiple DN links exist a decision
over which link to transmit is made either upon informa-
tion on the resources available along the existing paths or
randomly,if no such information has been obtained.In par-
ticular the parameter for the selection of the DN link is the
available number of transceivers along the outgoing paths
and such information is collected during the algorithmcon-
struction phase by messages piggybacked in the transmitted
acknowledgments.
Unless it is the DEST node,any other node receiving a
CR temporarily reserves a transceiver (if at least one idle
transceiver is available) and retransmits the CR to a DN
node (see example in figure 1(b)(c)).If no idle transceiver
is available at the time the CR is received,a “Negative Ac-
knowledgment” control packet (NAK) is generated and sent
back to the UP neighbor to indicate temporary lack of re-
sources,and the particular link is blocked for future trans-
missions of the same CR.The NAKis a control packet gen-
erated by a node to indicate lack of resources to accommo-
date the request for a connection.Any node receiving a
NAK attempts to retransmit the rejected CR to a different
DN neighbor (if such a neighbor exists).To avoid multiple
unnecessary attempts of transmitting over the same link,a
link blocking rule is considered according to which recep-
tion of a NAKover a DNlink automatically marks this link
as “Blocked” for the specific CR.
If a CR reaches the DEST and the request is admitted,
the destination node updates the corresponding entry in its
connectivity table and transmits backwards to the source an
“Acknowledgment” (ACK) control message (node DEST in
the example of figure 1(d)).Otherwise if the request cannot
be serviced by the DEST,a NAKis transmitted back to the
1
3
4
2
DEST
ACK
(a) Initiate CR transmission
1
2
3
4
1
2
3
4
1
2
3
4
CR
DEST
CR
DEST DEST
(b) CR propagation
CR
DEST reached

(c) CR propagation )
1
2
3
4
DEST
1
2
3
4
DEST
1
2
3
4
DEST
1
2
3
4
DEST
1
2
3
4
DEST
(f) ACK propagation
ACK
ACK ACK
1
3
4
2
DEST
ACK
CH
CH
(e) ACK propagation in
reverse direction (cont)
(session established)
(g) Link 2-3 fails
Connection Handoff Request
(d) ACK propagation
(h) CH reaches node 3
(already in the path) reverse direction (cont)
(j) Session handed-off
to new path
(i) ACK propagation in
1
3
4
2
1
3
4
2
DESTDEST
TRM
TRM
TRM

(k) TRM tears down session (l) TRM tears down session
Figure 1.Example of algorithmexecution
link over which the CR was received.
ACK messages are generated and transmitted by desti-
nation nodes,are destined to the source node of the CR and
must follow the same path of the CR in the reverse direc-
tion (see example in Figure 1(d-f)),updating the connec-
tivity tables of each node in the path.Link failures may
destroy a path before the CR-ACK phase is completed and
this may result in reception of an ACK over a DN link but
for a session that has already been interrupted (before even
acknowledged).
A node receiving an ACK updates the connectivity ta-
ble by confirming the reservation and forwards the ACK to
the upstream node.A more enhanced version of the algo-
rithm is when every ACK is broadcast by a node to all its
upstream neighbors.The advantage of this mechanism is
based on a slight modification of the packet format to also
carry the maximumnumber of available transceivers among
all possible paths to the destination.This value is compared
to the receiving node’s idle transceivers and the minimum
value replaces the entry in the packet field.Nodes do not
anymore select randomly over which DN link to transmit a
CR,but make their choice based on collected side informa-
tion.
This mechanism provides all nodes with the maximum
amount of information but at the cost of “outdated” infor-
mation in cases of large networks with high rate of topolog-
ical changes,since by the moment the ACK is received the
information may already be inaccurate.It also results in a
high number of control packets transmissions which could
slowdown execution of the algorithm.
B.Maintenance phase
During the maintenance phase the algorithm reacts to con-
nectivity changes that affect on-going sessions.In order
to maintain connectivity for an additional amount of time,
the algorithm takes advantage of the possibility to convert
a wireless connectivity change to a soft failure,and search
for a by–pass route to handoff the on–going session.
We employ the notion of a soft failure (opposed to a hard
failure which is a complete loss of connectivity) to charac-
terize degradation in the link quality.A link which expe-
riences a soft failure may still be used for transmission at
some cost,for instance at a lower information rate,but the
algorithm has the chance to locate and reserve a by–pass
route to be used by the failing session.The physical layer
mechanisms during a soft failure could involve adaptive de-
modulation and efficient bit allocation algorithms which are
beyond the scope of this paper.
A node reacts to a soft–failure by generating and trans-
mitting a “Connection-Handoff” (CH) control message in
search of by-pass routes.In general CHpropagationfollows
the same rules as the CR with main difference that an ACK
may be generated either by the DEST or also by any other
node that is already in the path but is not affected by the link
failure (see example in figure 1(g-j)).Accordingly ACK or
NAKpropagationdiffers fromthe construction phase in that
it ceases when the node which requested the hand–off has
been reached.Note that a node already in the path knows
that a received CHis being served by comparing the session
IDto in its connectivity table entries.
Of course,when a hard link failure takes place connec-
tivity is completely lost and the source node has to re–
establish the connection to the destination through the con-
struction phase procedure.
C.Termination phase
During the termination phase,nodes clear entries in their
connectivity tables that are no longer needed either because
a connection was terminated or because it was interrupted
due to a link failure or was handed–off to a newpath.When
a session is completed the source node which initiated the
call tears it down by generating a TRM control message
and broadcasting it along the session path to the DEST (see
example in figure 1(k-l)).Another situation where a TRM
packet is needed is when a link experiences a hard failure.
In that case the two edges of the link lose communication
with each other and therefore must notify the rest of the
nodes located in the two resulting segments of the path to
tear down the connection and clear the relevant entries in
their tables.
3.Performance analysis
We have simulated the system in Opnet Modeler,a
discrete–event simulator with the required features in mod-
eling a distributed algorithm.In this section we highlight
the main properties of the simulation model and present
some initial simulation results.
3.1.Mobile call model
New call arrivals to the network are assumed to be Pois-
son with rate
￿ ￿ ￿
request/min.The call holding time
is exponentially distributed with mean
￿ ￿ ￿￿
sec.Each
new call arrival is equally likely to arrive at any node as its
source,and any of the remaining nodes is equally likely to
be its destination.
3.2.Connectivity model
To capture node mobility and physical layer impairment,
we employ a model of fixed-node topology but with dy-
namic link status.In particular,the source and destination
of each call remain fixed as long as the call is in progress.
Instead,the status of each link is assumed to be subject to
dynamic (perhaps random) changes throughout the call du-
ration.
Asimple three–state probabilistic link status model is as-
sumed.In particular,the possible states of the link status are
“FULL”,“HALF”,or “ZERO”,corresponding to the status
of the link being able to support transmissions at the full
rate,half rate,or link out-of-service,respectively.Specif-
ically,the transition from FULL to HALF state models a
soft link failure,which parallels a link quality degradation
situation,whereas the FULL to ZERO state transition cor-
responds to a hard link failure implying complete loss of
connectivity.This model allows a unified treatment of node
mobility and physical impairment in a flexible fashion.The
difference fromother commonly used two-state models [1]
is that we can model both a soft and a hard failure.
We run simulations for different values of average hold-
ing times for each state and steady state distribution.The
steady state distribution gives an indication of how much
time on average each link spends at each case,while the
holding times determine the average rate of topological
changes,which gets higher when the holding times are de-
creasing.
3.3.Simulation results
Performance is measured end–to–end in terms of new
call blocking probability and forced termination probability
and compared under the assumptions of hard link–failures
only,against soft link–failures.We consider a baseline
topology of 10 nodes but examine two different connectiv-
ity senarios,“topology1” (with a maximumof 17 links) and
“topology2” (with a maximum of 24 links).In all cases
links are on average 95%of the time FULL and in the case
of the 3 state model 3%of the time under a soft failure sit-
uation whereas in the case of the 2 state model 5% of the
time under a hard failure.
0
1
2
3
4
5
6
7
8
9
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Average rate of link status change (changes/min)
Probability
Topology 1
Pf − 2−state model
Pf − 3−state model
Pb − 2−state model
Pb − 3−state model
Figure 2.Results for Topology 1
Figures 2 and 3 show the end–to–end probabilities of
forced termination


and of new call blocking


as func-
tions of the average rate of link status change
￿

(in num-
ber of changes per link per minute) for the cases of topo1
and topo2 respectively.Note that


increases as
￿

in-
creases since more sessions are interrupted.Clearly the use
of the three–state probabilistic model for the link status re-
sults in some improvement in


.We also observe that


decreases when
￿

increases.In other words,as the link
failures become more frequent,the new call requests have
a better chance of being admitted end-to-end.The reason
is that more capacity is made available at all nodes when
more calls in progress get force–terminated,so it is more
likely for a newcall to find capacity end–to–end.This indi-
cates that a combination of blocking and forced-termination
probabilities,rather than blocking alone,has to be consid-
ered in the end-to-end designs for such systems.
Comparison of the results for the two topologies shows
that the additional route diversity of topo2 results in a sig-
nificant drop of


whereas the improvement in


is not
that dramatic,mainly because of a limitation of the CE pro-
tocol which temporarily blocks some DN links in order to
avoid formation of loops.Hence the temporary link block-
ing rule prevents our algorithm from taking full advantage
of route redundancy.
4.Conclusion and future work
We have presented an algorithm capable of support-
ing connection-oriented service in wireless all-mobile net-
0
1
2
3
4
5
6
7
8
9
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Average rate of link status change (changes/min)
Probability
Topology 2
Pf − 2−state model
Pf − 3−state model
Pb − 2−state model
Pb − 3−state model
Figure 3.Results for Topology 2
works.We have based our approach on a protocol for
connectionless type of service to discover routes between
source and destination pairs.The algorithm features a pre-
dictive re-routing scheme,where by the appropriate mecha-
nisms wireless connectivity changes can be modeled as soft
link failures and can maintain on-going sessions until by-
pass routes have been discovered.Simulation results show
that performance improves when the three state probabilis-
tic model is used.Our future work will focus on investigat-
ing the joint problemof routing and call admission control
in order to achieve better improvement in performance as
far as forced termination is concerned.We are also going
to look into physical layer issues and power saving related
procedures.
￿
References
[1] M.Corson and A.Ephremides.A distributed routing algo-
rithmfor mobile wireless networks.Wireless Networks,1:61–
81,1995.
[2] J.J.Garcia-Luna-Aceves and S.Murthy.A path-finding al-
gorithm for loop-free routing.IEEE/ACM Transactions on
Networking,5:148–160,February 1997.
[3] A.Michail,W.Chen,and A.Ephremides.Distributed rout-
ing and resource allocation for connection-oriented traffic in
ad-hoc wireless networks.In Conference on Information Sci-
ences and Systems,March 1998.
[4] V.Park and M.S.Corson.Ahighly adaptive distributed rout-
ing algorithm for mobile wireless networks.In Proc.IEEE
INFOCOM’97,Japan,April 1997.
[5] C.Perkins and P.Bhagwat.Highly dynamic destination-
sequenced distance vector routing (dsdv) for mobile comput-
ers.In ACMSIGCOMM,October 1994.
￿
The views and conclusions contained in this document are those of the
authors and should not be interpreted as representing the official policies,
either expressed or implied,of the Army Research Laboratory or the U.S.
Government.