Int. J. of Computers, Communications & Control, ISSN 1841-9836, E-ISSN 1841-9844
Vol. IV (2009), No. 4, pp. 427-438

Fuzzy Controller Based QoS Routing Algorithm with a Multiclass Scheme
for MANET

B. Sun, C. Gui, Q. Zhang, H. Chen

Baolin Sun, Chao Gui, Qifei Zhang
School of Computing
Hubei University of Economics
Wuhan 430205, China
E-mail: blsun@163.com, guichao@126.com, cheffly@gmail.com

Hua Chen
College of Science
Wuhan University of Science and Engineering
Wuhan 430073, China
E-mail: Qiuchen_1022@163.com

Abstract: A mobile ad hoc network (MANET) consists of a set of mobile hosts
that can communicate with each other without the assistance of base stations. Due
to the dynamic nature of the network topology and restricted resources, quality of
service (QoS) and multicast routing in MANET are challenging tasks which attract
the interests of many people. In this paper, we present a fuzzy controller based QoS
routing algorithm with a multiclass scheme (FQRA) in mobile ad hoc networks. The
performance of this scheduler is studied using NS2 (Network Simulator version 2)
and evaluated in terms of quantitative measures such as packet delivery ratio, path
success ratio and average end-to-end delay. Simulations show that the approach is
efficient, promising and applicable in ad hoc networks.
Keywords: Mobile ad hoc networks, Fuzzy controllers, Routing algorithm, QoS

1 Introduction

Mobile ad hoc networks (MANET) consist of mobile nodes that autonomously establish connectivity
via multihop wireless communications. There is no use of a static network infrastructure such as base
station or any centralized administration in MANET. In ad hoc network, if two nodes are not within radio
range, all message communication between them must pass through one or more intermediate nodes. All
the nodes are free to move around randomly, thus changing the network topology dynamically [1-10].
These type of networks have many advantages, such as self reconfiguration and adaptability to highly
variable mobile characteristics like the transmission conditions, propagation channel distribution charac-
teristics and power level. They are useful in many situations such as military applications, conferences,
lectures, emergency search, and rescue operations. However, such benefits come with new challenges
which mainly resides in the unpredictability of the network topology due to mobility of nodes and the
limited available bandwidth due to the wireless channel. These characteristics demand a new way of
designing and operating these type of networks. For such networks, an effective routing protocol is crit-
ical for adapting to node mobility as well as possible channel error to provide a feasible path for data
transmission [1-10].

AODV is an on-demand distance vector routing protocol [2]. The protocol is well known for the use
in ad hoc networks. The use of multicasting with the network has many benefits. Multicasting reduces
the communication cost for applications that sending the same data to many recipients [3, 5-8]. Instead
of sending via multiple unicast, multicast reduces the channel bandwidth, sender and router processing
and delivery delay. In addition, multicast gives robust communication whereby the receiver address is
unknown or modifiable without the knowledge of the source within the wireless environment.

Copyright c© 2006-2009 by CCC Publications



428 B. Sun, C. Gui, Q. Zhang, H. Chen

Quality of service (QoS) support for multimedia applications is closely related to resource allocation,
the objective of which is to decide how to reserve resources such that QoS requirements of all the appli-
cations can be satisfied [3, 6-11, 16, 21]. The goals of QoS routing are twofold: finding a suitable route
through the network between the source and the destination that will have the necessary resources avail-
able to meet the QoS constraints; and achieving global efficiency in the utilization of resources. How-
ever, it is a significant technical challenge to provide reliable high-speed end-to-end communications
in these networks, due to their dynamic topology, distributed management, and multi-hop connections.
The provision of QoS requirements is of utmost importance for the development of future networks. For
supporting QoS-aware applications, QoS based routing algorithms such as ticket base routing (TBR) [3]
and core extraction dynamic source routing (CEDAR) [4] are proposed. In CEDAR, each host compares
degree with its neighbors. The host with the largest degrees among its neighbors is selected as the core
node. The core node is responsible for recording all information for its members. When a source host
requests to establish a QoS routing path to some destination host, its core node must construct a QoS
routing path. The advantage of core-based management is that it reduces the number of control packets
during path construction. Chang proposes a two-level management approach for efficiently constructing
and maintaining a QoS routing path in ad hoc wireless networks, significantly reducing the quantity of
control packets [5]. In the first phase, the mobile hosts are partitioned into a number of complete graphs,
each represented by a supernode managed by an agent. In the second phase, some agents of a larger
degree than neighboring agents are selected as core nodes. Lorenz and Orda demonstrate in the litera-
ture [16] that this uncertainty places additional constraints on QoS provisioning. Xiao et al proposed a
Dynamic Backup Routes Routing Protocol (DBR2P), a backup node mechanism for quick reconnection
during link failures [20]. DBR2P is an on-demand routing protocol and it can set up many routes to reach
a destination node in a given period. These algorithms determine a path that satisfies the required QoS.

Fuzzy logic exploits the pervasive imprecision, uncertainty and partial truth of the real world using
simple linguistic statements and thereby achieves tractability, robustness, and low solution cost [11, 13-
19, 23]. Fuzzy logic based decision algorithm influences caching decisions of multiple paths uncovered
during route discovery and avoids low quality paths [11, 14-19]. Differentiated resource allocation con-
sidering message type and network queue status is evaluated using fuzzy logic scheme [11]. Fernandez,
Hu, kazemian, Zhang et al propose the use of fuzzy logic controllers for the dynamic reconfiguration of
edge and core routers [13-15, 23]. This reconfiguration allows for adjusting the network provisioning
according to the incoming traffic and the QoS level achieved. Hu and Peter, utilize a self tuning fuzzy
controller to apply an end-to-end rate-based feedback flow control algorithm for the available bit rate
(ABR) service in ATM [14]. Sheng, et al propose an adaptive routing algorithm in which the link cost
was dynamically assigned using a fuzzy system [18]. Their results show the traffic in the networks is
rerouted to have less congestion or spare capacity. Sheu and Chen propose a fuzzy bandwidth alloca-
tion controller (FBAC) to support services including restricted time-bounded service such as voice and
video in wireless networks [19]. Their results show the traffic in the network is rerouted to have less
congestion or spare capacity. In those studies, the parameters and rules were further calibrated to obtain
a more efficient evaluation. Because the ad hoc network traffics have self-similarity, a common statistical
method has been widely used to verify self-similarity of time-series, so that the characteristics of the ad
hoc wireless networks can be grasped.

In this paper, we present a Fuzzy controller based QoS Routing Algorithm with a multiclass scheme
(FQRA) in mobile ad hoc networks. FQRA proposes to deal with route table management for keeping
the active routes’ lifetime. FQRA applies a fuzzy logic system to dynamically evaluate the route expiry
time. The fuzzy logic is chosen because there are uncertainties associated with node mobility and the
estimation of link crash; moreover, there is a mathematical model capable of estimating the node mobil-
ity. In addition, FQRA is able to take some controlling factors into consideration. Therefore, FQRA is a
multiclass scheme fuzzy evaluation for QoS routing protocol. The performance of the FQRA is studied
using NS2 [24] and evaluated in terms of quantitative measures such as improved path success ratio,



Fuzzy Controller Based QoS Routing Algorithm with a Multiclass Scheme for MANET 429

reduced average end-to-end delay and increased packet delivery ratio.
The rest of the paper is organized as follows. In Section 2, we introduces the ad hoc network model

and route issues. In Section 3, we presents the fuzzy QoS controller. In Section 4, we evaluate the per-
formance of the algorithm and present the simulation results. Finally, Section 5 concludes the summary
of the work and future challenges.

2 Network Model and Routing Issues

A network is usually represented as a weighted digraph G = (N, E), where N denotes the set of
nodes and E denotes the set of communication links connecting the nodes. |N| and |E| denote the number
of nodes and links in the network respectively [3-11, 16]. In G(N, E), considering a QoS constrained
multicast routing problem from a source node to multi-destination nodes, namely given a non-empty
set M={s, u, u, . . . , um}, M ⊆ N, s is source node, U ={u, u, . . . , um} is a set of destination nodes.
Multicast tree T = (NT , ET ), whereNT ⊆ N, ET ⊆ E, if C(T ) is the cost of T , PT (s, u) is the path from
source node s to destination u ∈ U in T , DT (s, u) and BT (s, u) are the delay and usable bandwidth of
PT (s, u).

Definition 1. The cost of multicast tree T is:

C(T ) =
∑

e∈ET C(e), e ∈ ET .

Definition 2. The bandwidth, and delay route of multicast tree T is the value of link bandwidth, and
delay in the path from source node s to each destination node u ∈ U . i.e.

BT (s, u)= min(B(e), e ∈ ET ).
DT (s, u)= max(

∑
e∈PT (s, u) D(e), u ∈ U ).

Definition 3. Assume the minimum bandwidth constraint of multicast tree is B and the maximum delay
constraint is D, given a multicast demand R; then, the problem of bandwidth and delay constrained
multicast routing is to find a multicast tree T , satisfying:

(1) Bandwidth constraint: BT (s, u) ≥ B, u ∈ U .
(2) Delay constraint: DT (s, u) ≤ D, u ∈ U .
Suppose S(R) is the set, and S(R) satisfies the conditions above; then, the multicast tree T which we

find is:

C(T ) = min (C(Ts), Ts ∈ S(R))

3 Fuzzy QoS Controller

3.1 Fuzzy Logic Controller

The fuzzy logic was introduced by Zadeh as a generalization of the boolean logic [22]. The difference
between these logics is that fuzzy set theory provides a form to represent uncertainties; that is, it accepts
conditions partially true or partially false. Fuzzy logic is a good logic to treat random uncertainty, i.e.,
when the prediction of a sequence of events is not possible.

Fuzzy logic control system is rule-based system in which a set of so-called fuzzy rules represents a
control decision mechanism to adjust the effects of certain causes that come from the system. The aim
of the fuzzy control system is normally to substitute for or replace a skilled human operator with a fuzzy
rule-based system. Specifically, based on the current state of a system, an inference engine equipped



430 B. Sun, C. Gui, Q. Zhang, H. Chen

with a fuzzy rule base determines an on-line decision to adjust the system behavior in order to guarantee
that it is optimal in some certain senses.

There are generally two kinds of fuzzy logic controllers. One is the feedback controller, which is not
suitable for the high performance communication networks. Another one, which is used in this paper,
is shown in Figure 1. The output of the fuzzy logic controller in Figure 1 is used to tune the controlled
system’s parameters based on the state of the system. This control mechanism is different from the
conventional feedback control and considered as an adaptive control [11, 14, 19, 23].

 

 

 

 

Figure 1: The fuzzy logic controller

The specific features of the fuzzy controller depend on the model under control and performance
measurement. However, in principle, in the fuzzy controller we explore the implicit and explicit relation-
ships within the system and subsequently develop the optimal fuzzy control rules as well as a knowledge
base.

In the route discovery process, new routes are created. The creation of new routes makes use of
forward packet (Table 1) and backward packet (Table 2). A forward packet is broadcasted by the sender
and will be relayed by the neighbors of the sender.

When the forward packet reaches the destination node, it extracts the information collected by the
forward packet and destroys it and subsequently creates a backward packet which follows the track of
the forward packet, but in the reverse direction.

Table 1: Forward packet.

Source address Destination address Sequence number

Number of hops QoS Metric Intermediate node

Table 2: Backward packet.

Source address Destination address Number of hops

QoS Metric Intermediate node

The packets flow through router is presented in Figure 2. The aim was to decrease TCP and UDP
traffic settling, rise and fall times; decrease overshoots and undershoots; and to stabilize throughput in
the router’s output connections to let the model adapt for load and link capacity variations.



Fuzzy Controller Based QoS Routing Algorithm with a Multiclass Scheme for MANET 431
 

 

 

 

 

Figure 2: Schematic presentation of packets flow through router

3.2 Scheduler Controller

The packet scheduler used in our architecture is WRR (Weighted Round Robin). In this scheduler,
queues are served according to a configurable weight that can be changed during network operation. This
allows having control of the bandwidth assigned to each service class. The packet delay and discard rate
for each queue (class) can be controlled by changing this weight. An example of membership function
of schedule controller is showed in Figure 3. The fuzzy scheduler proposed here calculates the priority
index of each packet. Here we consider all the inputs which decide the priority associated with the
packet, unlike the previous scheduling schemes. Other membership functions are: packet delay in the
expedited forwarding queue and discard rate due to queue overflow in the besteffort class. The output
membership functions are also defined as trapezoid functions by the same previous reasons. We make
use of the center of gravity defuzzification method, since it gives better results. The output membership
function gives the weights assigned to each class in the WRR scheduler.

 

 

 

 

Figure 3: Scheduler membership functions

3.3 Fuzzy Rule Base (Fuzzy Routing)

Fuzzy systems reason with multi-valued fuzzy sets instead of crisp sets. The Fuzzy Logic Controller
(FLC) (Figure 1) has two inputs: Residual Bandwidth and Traffic Class and one output: Fuzzy Routing
Decision [12, 22].

Mamdani fuzzy-rule based systems constitute of a linguistic description in both the antecedent parts
and the consequent parts. Each rule (Table 3) is a description of a condition-action statement that may
be clearly interpreted by the users. Rule base is an IF-THEN rule group with fuzzy sets that represents
the desired behavior of a fuzzy system. It can be defined in agreement with the administrative policy.

IF x is Ai and . . . and xn is Ain THEN y is Ci, i = 1, 2, . . . , L

where L is the number of fuzzy rules, x j ∈ U j, j=1, 2, . . . , n, are the input variables, y is the output
variable, Ai j are the fuzzy sets of the input linguistic variable x j and Ci is called the set of the output
linguistic variable y. Ai j and Ci are characterized by both membership functions.

Inputs are of the form: x is A′, x is A
′
, . . . , xn is A

′
n where A

′
, A

′
, . . . , A

′
n are fuzzy subsets of U,

U, . . . Un, which are the universe of discourse of inputs.



432 B. Sun, C. Gui, Q. Zhang, H. Chen

Table 3: Fuzzy rule base - QoS classes and application type.

QoS Class Application Type

1 Video conference
2 SDTV-quality voice
3 CD-quality audio
4 High-quality voice

4 Simulation

4.1 Random Graph Generation

In generating random graphs, we have adopted the method used in Kazemian, where vertices are
placed randomly in a rectangular coordinate grid by generating uniformly distributed values for their x
and y coordinates [15]. The remaining edges of the graph are chosen by examining each possible edge
(u,v) and generating a random number 0 ≤ r <1. If r is less than a probability function P(u, v) based on
the edge distance between u and v, then the edge is included in the graph. The distance for each edge is
the Euclidean distance denoted as d(u, v) between the nodes that form the end-points of the edge. We
use the probability

P(u, v) = β exp[− d(u,v)α L ].

where d(u, v) is geometric distance from node u to node v and L is maximum distance between two nodes.
The parameters α and β are in the range (0, 1) and can be used to obtain certain desirable characteristics
in the topology; parameter α can be used to control short edge and long edge of the random graph, and
parameter β can be used to control the value of average degree of the random graph.

4.2 Simulation Model

To conduct the simulation studies, we used randomly generated networks on which the algorithms
were executed. This ensures that the simulation results are independent of the characteristics of any
particular network topology. Using randomly generated network topologies also provides the necessary
flexibility to tune various network parameters such as average degree, number of nodes, and number of
edges, and to study the effect of these parameters on the performance of the algorithms. The platform
used was the NS2 (Network Simulator version 2) [24]. NS2 is a discrete event simulator targeted at
networking research. NS2 provides substantial support for simulation of TCP, routing, and multicast
protocols over wired and wireless networks.

Fuzzy Routing Decision was implemented using the Fuzzy Logic Toolbox in MATLAB [25]. The
simulator ran with various input configuration settings and the statistics collected were analyzed in com-
parison with other well-known on demand routing protocol AODV [2].

Our simulation modeled a network of mobile nodes placed randomly within 1000 × 1000 meter
area. Each node had a radio propagation range of 250 meters and channel capacity of 10 Mbps. Two-ray
propagation model was used. The IEEE 802.11 distributed coordination function was used as the medium
access control protocol. A random waypoint mobility model was used: each node randomly selected a
position and moves toward that location with a speed ranging from just above 0 m/s to 10 m/s. When the
node reached that position, it became stationary for a programmable pause time; then it selected another
position and repeated the process. The simulation was repeated with different seed values. A traffic



Fuzzy Controller Based QoS Routing Algorithm with a Multiclass Scheme for MANET 433

generator was developed to simulate CBR (Constant Bit Rate) sources. The size of the data payload was
512 bytes. Data sessions with randomly selected sources and destinations were simulated. Each source
transmitted data packets at a minimum rate of 4 packets/sec. and maximum rate of 10 packets/sec. Traffic
classes were randomly assigned and simulation was carried out with different bandwidth requirements.
There were no network partitions throughout the simulation. Each simulation was executed for 600
seconds of simulation time. Multiple runs with different seed values were conducted for each scenario
and collected data was averaged over those runs. Table 4 lists the simulation parameters which are used
as default values unless otherwise specified.

Table 4: Simulation parameters

Number of nodes 100
Terrain range 1000m × 1000m
Transmission range 250 m
Average node degree 3-5
Node’s mobility speed 0-10 m/s
Mobility model Random way point
Propagation model Free space
Channel bandwidth 5 Mbps
Links delay 20-200 ms
Traffic type CBR
Data payload 512 bytes/packet
Node pause time 0-10 seconds

4.3 Performance Measures

The following measures are used in computing the scheduler performance. These measures were
derived from one suggested by the MANET working group for routing protocol evaluation.

Packet delivery ratio: Packet delivery ratio is the ratio of the number of data packets actually
delivered to the destinations to the number of data packets supposed to be received. This estimate gives
us an idea about how successful the protocol is in delivering packets to the application layer. A high
value of packet delivery ratio indicates that most of the packets are being delivered to the higher layers
and is a good indicator of the algorithm performance.

Path success ratio: The ratio of the number of connection request discover to the destinations to the
number of routed connection requests. The successful routing request is defined as the path computed
and established by the algorithm satisfies the delay and bandwidth constraints. This number presents the
effectiveness of the algorithm.

Average end-to-end delay: This indicates the end-to-end delay experienced by packets from source
to destination. The average end-to-end packet delay is computed as the ratio of total end-to-end delays to
the aggregate number of packets successfully delivered to the destination nodes during a simulation run.
A higher value of end-to-end delay means that the network is congested and hence the routing algorithm
does not perform well.



434 B. Sun, C. Gui, Q. Zhang, H. Chen

4.4 Simulation Results

In this performance evaluation the following performance measures were evaluated: packet delivery
ratio, percentile of path success ratio and edge-to-edge delay. For each evaluation, we used CBR. All
simulations started with initial scheduler configuration with 60% of the bandwidth for each class. To
eliminate simulation results with an empty network, we started collecting results 30 seconds after the
beginning of the simulation.

After optimization procedure was executed, we could verify the result comparing packet delivery
ratio, path success ratio, and average end-to-end delay.

Figure 4 shows the performance analysis of the packet delivery ratio vs. network size for the FQRA,
AODV, and Non-QoS (not quality of service metric constraints) in ad hoc network. Using Non-QoS
constraints algorithm as a base line for comparison, the result shows that both FQRA and AODV are
much better than the Non-QoS constraints algorithm, FQRA is considerably better than AODV. This
is because that AODV needs to rediscover the route to retransmit data packets that are lost due to the
node’s mobility or unreal route paths during the communication. The advantage of FQRA is resulted
from choosing the right routing path or updating the unreal route paths just in time by the virtue of the
suitable route lifetime estimation.

 

 

 

Figure 4: Packet delivery ratio vs. Networks size

Figure 5 shows the comparison of the three algorithms with respect to the average end-to-end delay
vs. number of nodes. It shows that the end-to-end delay increases usually with the increasing node.
From the Figure 5, we can see that when the nodes increase, FQRA algorithm average end-to-end delay
is lower than that of AODV and Non-QoS algorithm. This is because the fuzzy scheduler controller
has stability routes and gives more precedence to the packets. Both the AODV and Non-QoS algorithm
need more time and more control overhead than the FQRA does to recover unreal paths (broken paths)
and to discover new paths. As the nodes increase, more packets are received and thus end-to-end delay
increases. The end-to-end delay is important as many real time applications require a small latency to
deliver usable data within stipulated period of time.

Figure 6 depicts a comparison path success rate to find the path through FQRA, AODV and Non-QoS
in ad hoc network. With the relaxation of bandwidth constraints, the path success rate becomes higher
for Non-QoS. The success rate is still higher than that of Non-QoS, which means more suitable for the
routing choosing under timely data transmission application and dynamic network structure.

The average and-to-end delay performance as shown in the Figure 7, proves that the end-to-end
delay improves when scheduler is included. As the mobility varies from 0-10 m/s, the fuzzy controllers
scheduler provides an end-to-end delay reduced by around 0.01 seconds to 0.05 seconds. It shows that
the end-to-end delay increases usually with the increasing speed. It also shows that FQRA is much better
than the other two. Both the Non-QoS and the AODV need more time and more control overhead than
the FQRA does to recover unreal paths and to discover new paths.



Fuzzy Controller Based QoS Routing Algorithm with a Multiclass Scheme for MANET 435

 

Figure 5: Average end-to-end delay vs. Networks size

 

Figure 6: Path success ratio vs. Bandwidth

 

Figure 7: Average end-to-end delay vs. Node’s mobility speed



436 B. Sun, C. Gui, Q. Zhang, H. Chen

5 Conclusion and Future Work

Our QoS routing algorithm has produced significant improvements in throughput, average end-to-
end delay and path success ratio. The fuzzy controllers scheduler algorithm attaches a QoS class to each
packet in the queue of the node. Unlike the normal sorting procedure for scheduling packet, the crisp
QoS class is calculated by the fuzzy scheduler based on the above inputs which are derived from the
network. The membership functions and rule bases of the fuzzy scheduler are carefully designed. The
use of fuzzy logic improves the handling of inaccuracy and uncertainties of the ingress traffic into the
domain.

In this paper, we present a fuzzy controller based QoS routing algorithm with a multiclass scheme in
mobile ad hoc networks. The performance of this scheduler is studied using NS2 and evaluated in terms
of quantitative measures such as path success ratio, average end-to-end delay and throughput. Simulation
shows that the approach is efficient, promising and applicable in ad hoc networks.

Future work includes comparison with "crisp" versions of the fuzzy algorithm to isolate the contri-
butions of fuzzy logic, as well as applications of fuzzy control to power consumption and directional
antennas in MANETs. We also intend to compare FQRA with other QoS routing algorithm.

Acknowledgement

The authors thank the editors and the anonymous reviewers for their valuable comments that helped
to improve the paper. This work is supported by China Postdoctoral Science Foundation of China (No.
20070410955), The Young and Middle-aged Elitists’ Scientific and Technological Innovation Team
Project of the Institutions of Higher Education in Hubei Province (No. T200902), and Key Scientific
Research Project of Hubei Education Department (No. D20081904, Q20091903, B20091904).

Bibliography

[1] M. A. Rajan, M. G. Chandra, L. C. Reddy and P. Hiremath, Concepts of Graph Theory Relevant
to Ad-hoc Networks. International Journal of Computers, Communications & Control, Vol. 3, No.
Suppl, pp. 465-469, 2008.

[2] C. Perkins, E. Belding-Royer and S. Das, Ad Hoc On-Demand Distance Vector (AODV) Routing.
IETF RFC 3561, July 2003.

[3] S. Chen and K. Nahrstedt, Distributed Quality of Service routing in Ad hoc networks. IEEE Journal
on Selected Areas in Communication, Vol. 17, No. 8, pp. 1488-1504, 1999.

[4] R. Sivakumar, P. Sinha and V. Bharghavan, CEDAR: Core Extraction Distributed Ad hoc Routing.
IEEE Journal on Selected Areas in Communication, Vol. 17, No. 8, pp. 1454-1465, 1999.

[5] C. Y. Chang, C. T. Chang, T. S. Chen and H. R. Chang, Hierarchical management protocol for
constructing a QoS communication path in wireless Ad Hoc networks. Information Sciences, Vol.
177, No. 13, pp. 2621-2641, 2007.

[6] B. L. Sun and L. Y. Li, A QoS Multicast Routing Optimization Algorithms Based on Genetic Algo-
rithm, Journal of Communications and Networks, Vol. 8, No. 1, pp. 116-122, 2006.

[7] B. L. Sun, L. Y. Li, Q. Yang and Y. Xiang, An Entropy-Based Stability QoS Multicast Routing
Protocol in Ad Hoc Network, Lecture Notes in Computer Science, Vol. 3947, Springer-Verlag, pp.
217-226, 2006.



Fuzzy Controller Based QoS Routing Algorithm with a Multiclass Scheme for MANET 437

[8] B. L. Sun, S. C. Pi, C. Gui, et al, Multiple Constraints QoS Multicast Routing Optimization Algo-
rithm in MANET based on GA. Progress in Natural Science, Vol. 18, No. 3, pp. 331-336, 2008.

[9] B. L. Sun, H. LI, Y. Zeng and Q. Q. QIN, Fuzzy QoS Controllers in Mobile Ad Hoc Networks using
Genetic Algorithms. Journal of Computational Information Systems, Vol. 3, No. 6, pp. 2255-2260,
2007.

[10] B. L. Sun and L. Y. Li, A Distributed QoS Multicast Routing Protocol in Ad Hoc Networks. Journal
of Systems Engineering and Electronics, Vol. 17, No. 3, pp. 692-698, 2006.

[11] A. Thomas, C. Chellappan and C. Jayakumar, ANTHOC - QoS: Quality of Service Routing in
Mobile Ad Hoc Networks using Swarm Intelligence. in: The Second Asia Pacific Conference on
Mobile Technology, Applications and Systems, Guangzhou, China, 15- 17 November, 2005.

[12] Y. H. Wang and C. F. Chao, Dynamic backup routes routing protocol for mobile ad hoc networks,
Information Sciences, Vol. 176, No. 2, pp. 161-185, 2006.

[13] M. P. Fernandez, A. de Castro, P. Pedroza and J. F. de Rezende, QoS provisioning across a diffserv
domain using policy-based management. in: Globecom 2001, San Antonio, USA, Nov. 2001.

[14] R. Q. Hu and D. Peter, A predictive self-tuning fuzzy logic feedback rate controller. IEEE/ACM
Transactions on Networking, Vol. 8, No. 6, pp. 697-709, 2000.

[15] H. B. Kazemian and L. Meng, A fuzzy control scheme for video transmission in Bluetooth wireless.
Information Sciences, Vol. 176, No. 9, pp. 1266-1289, 2006.

[16] D. H. Lorenz and A. Orda, Qos routing in networks with uncertain parameters. IEEE/ACM Trans-
actions on Networking, Vol. 6, No. 6, pp. 768-778, 1998.

[17] S. Rea and D. Pesch, Multi-metric routing decisions for ad hoc networks using fuzzy logic. in:
Proc. of 1st Intl. Symposium on Wireless Communication Systems, Mauritius, 20-22 September, pp.
403-407, 2004.

[18] H. M. Sheng, J. C. Wang, H. H. Huang and D. C. Yen, Fuzzy measure on vehicle routing problem
of hospital materials. Expert Systems with Applications, Vol. 30, No. 2, pp. 367-377, 2006.

[19] S. T. Sheu and M. H. Chen, A fuzzy bandwidth allocation controller to support real-time traffic
over wireless network. in: Proc. of IEEE Wireless Communications and Networking Conference,
New Orleans, LA, Vol. 1, pp. 1348-1352, 1999.

[20] Y. Xiao, C. L. P. Chen and Y. Wang, Optimal admission control for multi-class of wireless adaptive
multimedia services, IEICE Transaction on Communications, special issue on Mobile Multimedia
Communications, Vol. E84-B, No. 4, pp. 795-804, 2001.

[21] B. L. Ye, M. Y. Guo, D. X Chen and S. L. Lu, A degree-constrained QoS-aware routing algorithm
for application layer multicast, Information Sciences, Vol. 177, No. 17, pp. 3613-3626, 2007.

[22] L. A. Zadeh, Fuzzy sets. Information and Control, Vol. 8, No. 3, pp. 338-353, 1965.

[23] R. Zhang and Y. Phillis, Fuzzy Control of Queueing System with Heterogeneius Servers. IEEE
Trans. Fuzzy Systems, Vol. 7, No. 1, pp. 17-26, 1999.

[24] The Network Simulator - ns-2, http://www.isi.edu/nsnam/ns/.

[25] Fuzzy Logic Tool Box (For use with MATLAB) Users’ Guide Version 2, The MathWorks Inc.,
http://www.mathtools.net/MATLAB/Fuzzy_Logic/.



438 B. Sun, C. Gui, Q. Zhang, H. Chen

Baolin Sun (born in 1963), graduated in Mathematics (The Huazhong Normal University, Wuhan,
China) and received Ph.D. in Computer Science (Wuhan University of Technology, Wuhan,
China). He is currently a professor, and one of Editorial Board Guest Members of World Sci-
Tech R&D, one of the paper reviewers of Chinese of Journal Electronics, Journal of Electronics &
Information Technology, High Technology Letters, and KGCM, ICCIT program committee mem-
bers, and also an International Standard Draft organizing members of ISO/IEC JTC1/SC6. His
research interests include QoS multicast routing, distributed computing, network optimization and
ad hoc networks. Professor Sun has published over 100 research papers and is the author of four
books. He was awarded the Province Special Prize by the Hubei Province Government in 2007.
Chao Gui (born in 1964), graduated in Physics (The Wuahn University, Wuhan, China) and re-
ceived M.S. degree in Electrical Engineering (Wuhan University of Technology, Wuhan, China).
He is currently a associate professor in Computer Science. His research interests include wireless
communication, performance analysis and analytical modeling. He has published over 30 research
papers.
Qifei Zhang (born in 1977), graduated in Computer Science (The Jianghan University, Wuhan,
China) and received Ph.D. in Information and Communication Engineering (Huazhong University
of Science and Technology, Wuhan, China). He is currently a lecturer in Computer Science. His
research interests include wireless local area networks, Mesh Networks, Network Protocol and
Congestion Control. He has published over 30 research papers.
Hua Chen (born in 1977), graduated in graduated in Mathematics (The Huazhong Normal Univer-
sity, Wuhan, China) and received M.S. degree in Mathematics (Huazhong University of Science
and Technology, Wuhan, China). She is currently a lecturer in Mathematics. Her research inter-
ests include wireless communication, performance analysis, network model and algorithm. She
has published over 20 research papers.