Agent Approach QoS Routing in Manets Based on Fuzzy Priority Scheduler.

Krishna. G. Deshpande, V. R. Budyal

Abstract: As mobile computing gains popularity, the need for Ad-hoc network will continue to grow. The mobile Ad-hoc network is an autonomous system of mobile wireless nodes connected dynamically without any preexisting infrastructure. Here, the nodes are mobile, hence the network topology changes rapidly and unpredictable over time.Here we proposed the mesh based On Demand multicast Routing protocol (OMDRP), in OMDRP the mesh created by using route request and route reply packets. After the mesh we create the cluster, Clustering is a method which aggregates nodes into groups. These groups are contained within the network and they are known as clusters. In mesh we nd the multipaths from source to destinations. In that multipaths again we choose one QoS path depending on minimum delay required to reach the destination. In this paper we use the fuzzy scheduling algorithm to schedule the data packets based on their respective priority index,which will improve the network performance. The presented fuzzy scheduler for mobile Ad-hoc networks,determine the priority of the packets. The proposed fuzzy logic scheduling method is simulated by using Cprogramming language. The propsed scheme is better the without fuzzy priority scheduling system. Key words : MANETs, Multicast,ODMRP, Mesh, Fuzzy, Software Agents. I. I NTRODUCTION Mobile wireless ad hoc networks (MANETs) are a rapidly evolving telecommunications technology. Their popularity is connected with their easy deployment and fast conguration [1]. A mobile ad hoc network is an autonomous collection of mobile devices (laptops, smart phones, sensors, etc.) that communicate with each other over wireless links and cooperate in a distributed manner in order to provide the necessary network functionality in the absence of a xed infrastructure. These features make them ideal for average users, Internet service providers, and reacting to emergency situations in which normal communication is impossible. They can be used with success in disaster areas (earthquake, ood, and hurricane), military training grounds, and schools; at conferences, hotels, airports, houses, and so on. When it became clear that group-

oriented communication is one of the key application classes in MANET environments, a number of MANET multicast routing protocols have been proposed[2][3]. In this paper we used the on demand multicast routing protocol (OMDRP), ODMRP applies on-demand routing techniques to avoid channel overhead and improve scalability. It uses the concept of forwarding group [4], a set of nodes responsible for forwarding multicast data on shortest paths between any member pairs,and to build a forwarding mesh for each multicast group. Using mesh instead of a tree, the drawbacks of multicast trees in mobile wireless networks (e.g., intermittent connectivity, trafc concentration, frequent tree reconguration, on-shortest path in a shared tree, etc.) are avoided. A soft state approach is taken in ODMRP to maintain multicast group members. Multicast mesh creation involves two phases ; a request phase and a reply phase. Request phase invokes a route discovery process to nd routes to the multicast group. Different routes to the multicast group are setup during the reply phase. There are two types of nodes dened based on whether they are multicast group members or non-group members. Group members include all multicast sources, receivers and that of non-group members include intermediate nodes that help to create multicast routes from source to receivers. Non-group members help in forwarding the data packets. Both group members and non-group members help in recovery of failed links due to mobility of nodes and other interferences in route maintenance phase. After mesh we create a cluster for group communication, a successful method for dealing with the maintenance of mobile ad hoc networks is by partitioning the network into clusters. In this way the network becomes more manageable. Clustering helps aggregate topology information since the number of nodes of a cluster is smaller than the number of nodes of the entire network. Therefore, each node only needs to store a fraction of the total network routing information. In clustering techniques nodes are selected to play different roles according to a certain criteria. In general, three types of nodes are dened as Ordinary node or Forwarding node, Cluster Head, and Gateway node. After cluster we select receivers and for those receivers we select the designations in same cluster or may be in different clusters. A multicast multipath are found, in those path we select the shortest path which has a low delay to take to reach from source to destination.

The transmission of packets is to be performed, for this here we use a fuzzy scheduler. A scheduler should schedule the packets to reach the destination quickly, which are at the verge of expiry. Scheduling discipline manages the queue of requests awaiting service. Here we not applying the fuzzy logic at route nding, but we used the fuzzy logic at packet transmission after the QoS path nding. Without a scheduler, packets will be processed in FIFO manner and hence there are more chances that more packets may be dropped and hence the network may not meet the QoS target. In this paper, we designed and implanted a fuzzy-based priority scheduler.the fuzzy schedules the data packets based on its priority index.The fuzzy-based scheduling algorithm is coded in C language. Here we used the software agent in routing and also in packet scheduling stages. The software agents are autonomous programs situated within an environment,which sense the network environment and act upon it to achive the goals[12].the software agents have special properties such as mandatory and orthogonl properties[13].The orthogonal properties provides strong notion of the agents such as mobility,callaborative,learning. The mandatory properties ,such as autonomy,decision making,temproral continuity,goal oriented.It is found from the results that the proposed fuzzy scheduler improves the packet delivery ratio and decreases the end-to-end delay The route identication and management Agency and fuzzy agencies are used to improve the Packet deliver ratio, reducing the end-to-end delay, and reducing the control overhead. Organization of the paper: Section I provides the introduction on the manet and fuzzy scheduler . Section II discuses an proposed work. In section III, Simulation and Results.Section IV provides the conclusion on the proposed scheme. II. P ROPOSED W ORK A. Protocl Description Reactive routing techniques, also called on-demand routing. Routes are only discovered when they are actually needed. When a source node needs to send data packets to some destination, it checks its route table to determine whether it has a route. If no route exists, it performs a route discovery procedure to nd a path to the destination. Hence, route discovery becomes on-demand. The route discovery typically consists of the network-wide ooding of a request message. Once a route has been established, it is maintained by some form of route maintenance procedure until the destination becomes inaccessible or until the route is no longer desired. B. On Demand Multicast Routing Protocol (ODMRP) In this paper we use the mesh based on demand multicast routing protocol(ODMRP)[4] instead of tree based multicast routing protocols. There are two basic drawbacks of tree-based protocols. The rst drawback is ease of ree structure fragile because of unpredictable topology changes due to mobility of nodes. The second drawback is tree reconstruction delay. So, a

new topology concept called Mesh-based[5] was established. It has the possibility to provide multiple paths between any source-receiver pair. In the ODMRP, the mesh creating bt Route Request and Route reply Packets. C. Clustering in MANET A successful approach for clustering for dealing with the maintenance of mobile ad hoc networks is by partitionating the network into clusters[6]. In this way the network becomes more manageable. Clustering is a method which aggregates nodes into groups. These groups are contained by the network and they are known as clusters. we created the clusters as shown in gure1. The following steps have to be followed for forming the
Source Node

D2 D3 D1

D4
Destination Node

Cluster Head

Gateway Node S1

Forwording Node

D5

D8 S3 D6

D7 S2

Fig. 1.

Clustering in MANET

cluster [7]. Step 1: After deployment, the nodes broadcast their id value to their neighbors along with the HELLO message. Step 2: When all the nodes have discovered their neighbors, they exchange information about the number of one hop neighbors. The node which has minimum mobility is selected as the cluster head. Other nodes become members of the cluster or local nodes. The nodes update the status values accordingly. Step 3: The cluster head broadcasts the hello message so as to know its members. Step 4: The members reply with the reply message and in this way clusters are formed in the network. Step 5: If a node receives more than one hellow messages, it becomes Gateway which acts as a mediator between two clusters In gure1 four clusters are formed, for group communication each group consist number of destinations from different clusters or same clusters. we start transmitting the data from source s1 to Group 1 that contains a destinations D2,D3,D4. The D2 is within the cluster were the source present. So the transmitting path are the S1 -cluster head- D2.Suppose the source and destinations are in Different clusters then data transmitting through the forwarding node Gate way. For example from S1 to D3, the path is S1 Cluster Head -Gate

way-Cluster Head-D3. D. Scheduling Methods Without scheduler the packets are scheduled in First in First out (FIFO).For improving the performance of the mobile ad-hoc multimedia networks, a scheduler can be used. A scheduling algorithm determine which packet is delivered next among the packets are in the queues[8]. The scheduling algorithm is placed after routing agent. If the queue is full the scheduler will drops the packets from the queue. The fuzzy scheduling algorithm give the highest priority index to the control packets. And for data packets it schedules the data packets depending on those data rate, Queue length, and expiry time. Figure2 . Shows the priority scheduler for data packets. In this paper we determine the priority index of the data packets by the fuzzy logic concept. The three input variables namely, expiry time of packet, queue length of the node, and data rate of the source, are considered and the application of fuzzy logic to combine these variables and hence nd the priority index of the packet is found to be suitable. This led to the design of a fuzzy-based priority scheduler[9]. The most popular defuzzication method is used Centroid ,which returns the center of area under the curve. By cetnroid method of defuzzication ,the crisp output the crisp output is calculated using the formula =
allrules

E. Fuzzy Logic Scheduling Scheme The fuzzy logic implements human experience and preference by using membership functions and fuzzy logic rules[11]. The fuzzy scheduler proposed in this project, it calculates the priority index of the incoming each packets as shown in gure3. Here, the fuzzy priority scheduler considered as multiple input and single output(MISO). The linguistic variables associated with the input variables are low (L), medium (M), high (H), and. For the output variable, priority index, ve linguistic variables are used. They are, low (L), very low (VL), medium (M), high (H), and very high (VH). The membership functions of these variables are shown in the Figure4. (priority index g) The rst step to take the inputs and determine the degree to which the inputs belongs via membership functions. The membership values always between 0 and 1.If fuzzy system has n number of inputs and a single output, here inputs are data rate, Expiry time, Queue Length ,and output is Priority index. The outputs depends upon the fuzzy rules here twenty seven rules are shown in Fuzzy rule table 1.
Fuzzy Inputs Outputs Data Rate ( At Packet) Priority Index Expiry Time ( At Packet) Fuzzy System

xi (xi ) allrules (xi )

(1)

where, is the fuzzy cost, xi is the element and (xi) is its membership function. This is the most widely adopted defuzzication strategy, which is reminiscent of the calculation of the expected value of probability distributions.
Control Packets

Queue Length (At Node)

Data Packet 2

Fuzzy
Data Packet 3

Scheduler

Data Packet N

Fig. 2.

Priority Scheduler for data packets

The some of the rules are illustrate as follows. 1. If (DR is Low ET is Low QL is Low) Then Priority index is Low 2. If (DR is Low ET is Medium QL is Low) Then Priority index is Very Low 3. If (DR is Low ET is High QL is Low) Then Priority index is Very High The fuzzy scheduler uses three input variables and one output variable. The three input variables to be fuzzied are, the expiry time, and data rate of the packet and Queue length of the nodes to which the packet is associated with. The inputs are fuzzied, implicated, aggregated and defuzzied to get the output i.e., the priority index.

Medium Low

High

TABLE I F UZZY RULE TABLE Case 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Data Rate Low Low Low Low Low Low Low Low Low Medium Medium Medium Medium Medium Medium Medium Medium Medium High High High High High High High High High Expiry Time Low low low Medium Medium Medium High High High Low Low Low Medium Medium Medium High High High Low Low Low Medium Medium Medium High high High Queue length Low Medium High Low Medium High Low Medium high Low Medium high Low Medium high Low Medium High Low Medium High Low Medium High Low Medium High Priority Index Low Low Very Low Very Low Very Low Very Low Low Very Low Very Low Medium Medium Low Medium Medium Low Medium Medium Medium Very High Very High High High Medium Medium High High Medium

0.5

0
0 0.2 0.4 0.6 0.8 1.0 (a) 1.2 1.4 1.6 1.8 2.0

Medium Low

High

0.5

0 0 5 10 15 20 25 30 35 40 45 50 55 60

(b)

Low

Medium

High

0.5

0 0 15 30 45 60 75 90 105 120 135 150

(c)

Very Low
1

Low

Medium

High

Very High

0.5

Route Data Base

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

(d)

Fig. 4. Member Functions a) Data rate in Mbps b) Expiry time in msec c)Queue length d) Priority index

Node Manager Agent (NMA)

Packet Transfer Agent (PTA) Route Discovery Agent (RDA)

Packet Arrival Agent Migration

Mesh Management Agent (MNA)

F. Agency In this project we use the both static agents and mobile agents for route identication and fuzzy scheduler to improve the Quality Of Service (QoS). These are explained in bellow sections[10]. 1) Route Identication and Management Agency: Here we describe the stable route identication agency used in our proposed work. The different agents interactions of this agency is shown in gure5 it consist of a different agents such as Node Manager Agents (NMA), Mesh Management Agents(MMA), and Route Discovery Agents (RDA). These agents are responsible for collecting and determination network informations. the agency also consist of Data Base that is used for inter agents communication. Node Manager Agent( NMA) : It is a static agent that runs on a node, create the mesh based on Route Request (RP) and Route Reply (RP) packets. The mesh

Fig. 5.

Route Discovery Agent

creation based on data base, controls and coordinate the activities of the route identication agency. This agent advertise the route request packets to its neighbors periodically and also updates the data base. By using the Route Request (RP) and Route Reply (RP) packets we create a multicast mesh. This agents also maintain the mesh ,if any route fails in mesh the agents nds the alternative route. This agent triggers the route discovery agent to nd the the routes from source to destinations. The nal route is announced to all the nodes in the path between source and destinations. Route Discovery Agent (RDA) :It is a mobile agent, this agent migrate through the network in search of destinations when a route discovery process is initiated.

It collects the Queue length of the nodes along the available paths from source to destinations, and it also collects the end to end delay between the nodes of that paths. Finally this agent selects the best path which as a low delays and updates the data base. This agent also creates clusters for group communication, and choose cluster head based on which node has low mobility. Mesh management agent(MMA): It is a static agent that monitor the link connecting to neighbors in regular intervals of time. And it is always updating the data base. It checks the delay between the nodes for best path choosing. Route Data Base:It consist of information of the node i.e node id, node neighbors, node queue length ,end to end delay between nodes. this data base is used for updating the route and maintaince. 2) Fuzzy Agency:This agency is used for nding the priority index of the incoming packets based on packet parameters such as expiry time and data rate and node parameter like Queue Length. This agency consist of only one agent that is fuzzy scheduler and management agent.(FSMA). As shown in gure6.

Parameters Network Area Number of nodes Mobility Mode Speed of mobile nodes Distribution of nodes Communication range Packet Size Data Rate of Packet Expiry Time of the Packet Queue Length at nodes Learning constant()

Values 100x100 meters 10 to 50 Random waypoint 5-15 m/s Random up to 1o meter 512 0-2 Mbps 0-70 msec 20-100 0.1

TABLE II S IMULATION PARAMETERS

ET

DR

QL

ET Expiry Time DR Data Rate QLQueue Length

Fuzzy Data Base

Fuzzy Scheduling Agent (FSA)

Priority Index

Fig. 6.

Fuzzy Agency

Fuzzy Data Base: It consist of information of the packet parameter such as a expiry time ,data rate ,and node information such as Queue length. Fuzzy Scheduling Agent (FSA):This is a static agent, and acting at the node and using for packet scheduling i.e. which packet is scheduled rst based on fuzzy data base parameter maintained in Fuzzy data base. The proposed scheme Algorithm is as follows. step 1: Begin Mesh Creation and maintaince step 2: Randomly deploy the mobile nodes. step 3: Mesh Creation by Send Route Request Packet (RR) Route Reply(RP) Packets with the help of Route Identication and Management Agencies.

step 4: If RR packet Sequence No is Duplicate At the node then Discard the packet else, step 5: Receive the Packet. step 6: If mesh fails then multicast Routing Agent nd the alternate path, is available in Mesh. Cluster Creation for Group comunication step 7 The nodes broadcast their id value to their neighbors along with the HELLO message . step 8 If the node within cluster has a Low Mobility ,then step 9 Choose as a Cluster Head else step 10 Other nodes become members of the cluster or local nodes. The nodes update the status values accordingly. step 11 If a node receives more than one hellow messages, it becomes Gateway which acts as a mediator between two clusters. Data Transmission step12 Find the Feasible Multipaths from Source to Destinations step13 Find the QoS path based on low Delay step14 If QoS path is not found ,then update the route data base. step 15 Start Sending the packets from source to destinations using QoS path. Fuzzy scheduler step16 Apply the fuzzy priority scheduler at QoS path Nodes. step 17 The packets parameters such as data rate, expiry time and node parameter Queue length are the fuzzy scheduler inputs. step 18 calclate the priority index of the data packets by using fuzzy scheduler. step 19 End.

III. S IMULATION AND R ESULTS This proposed multicast routing protocol (ODMRP) has been implemented by using the C language. we used the simulation parameters are shown in table 1 . The following metrics are used for the calculation of fuzzy scheduler.

A. simulation parameters
Packet Delivery Ratio

mobility Vs Packet delivery ratio 1 0.8 0.6 0.4 0.2 0 with FPS without FPS

1) 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 number presents the effectiveness of the protocol No. of packets that are received. Packet Delivery Ratio = No.of packets sent. (2) 2) Average end-to-end delay of data packets: It represents the average value of the time that the received data packets take to reach the destination from their origin. This parameter includes the time the nodes stay in the internal queues, the retransmissions, and the forwarding through multiple intermediate nodes. 3) Agent Control Overheads: It is dened as the ratio of bandwidth utilized by mobile agents to the available bandwidth. B. Performance evaluation Using C language The fuzzy scheduler is implemented bu using C language. we rst assign the input variable then nding the priority index of the packets by applying fuzzy logic and using agents in our proposed work. Then the calculated priority index is used for the scheduling the data packets, which are the highest priority that means that packet to expire. Here we give the highest priority to the control packets.

20

40

60 mobility (kmph)

80

100

120

Fig. 7.

Mobility Vs Packet delivery ratio

b) Multicast group size Vs Packet delivery ratio: Figure 10 shows the Multicast group size Vs Packet delivery ratio. The number of senders is taken from two to ve, the mobility at 5-15 m/s and the multicast group size is varied from 2 to 20 members. Here the number of receivers increases the number of forwarding group node also increases, this in turn increases the connectivity of the links. So this further increases the PDR with FPS as compared to the without FPS.
multicast group size Vs Packet delivery ratio 1 0.8
Packet Delivery Ratio

with FPS without FPS

1) Simulation Results: a) Mobility Vs Packet delivery ratio: Figure 9 shows the mobility Vs Packet delivery ratio .Packet delivery ratio of ODMRP as a function of mobility speed. The size of multicast group is varied to examine the scalability of the protocol. The result indicates that ODMRP delivers high portion of data packets in most of our scenarios. As the number of members increases, the forwarding group mesh creates richer connectivity among members. The mesh makes the protocol scalable and robust to speed. In this simulation transmission rate at 2mbps, the mobility will be taken from 0 to 100 in kmph As the mobility speed is increasing, the PDR is start decreasing slowly in with FPS but PDR is decreasing rapidly as compared to FPS . The result shown in gure is clear that with Fuzzy scheduling priority system (FPS) provides the higher packet delivery ratio as compared to the without fuzzy scheduling priority system (without FPS).

0.6 0.4 0.2 0

10

15 20 multicast group size)

25

30

Fig. 8.

Multicast group size Vs Packet delivery ratio

c) Agent overhead Vs Number of nodes: As the number of nodes increases ,more number of mobile agents are used to nd the stable route and nding the packet priority index. Thus the agent overhead ratio increases with increase in the mobile nodes.this is shown in gure 11.

d) Number of nodes Vs End to END DELAYs: The end to end delay indicates how long it took for a packet to travel from the source to the destination.

Agent overhead Vs Number of nodes 1 0.8

communication range=10 meter communication range=15 meter communication range=20 meter

scheduling stages.The software agents are improves the PDR, decreases the end-to-delay.In future ,adding the another QoS parameter in simulation such as channel rate(Bandwidth). R EFERENCES

0.6 0.4 0.2 0

10

15

20 25 30 Number of nodes)

35

40

45

Fig. 9.

Agent overhead Vs Number of nodes

For low dense networks ,the ODMR gives the low end to-end delay. The average ned-to-delay increases with increasing the number of nodes. The gure 12 shows the the Number of nodes Vs endto-end delay.
Number of Nodes Vs EndtoEnd delay 70
EndtoEnd delay(in msec)

60 50 40 30 20 10 0 10 20 30 40 Number of nodes

without scheduling with FPS

50

60

70

Fig. 10.

Number of Nodes Vs End-to-End delay

IV. C ONCLUSIONS In this paper, we have analyzed the performance of the fuzzy-based priority scheduler for data trafc and evaluated the effect of inclusion of this scheduler with multicast mesh based routing protocol ODMRP. In this paper, we present a fuzzy controller based QoS routing algorithm with a multiclass scheme in mobile ad hoc networks. . Here we group the nodes for group communication by clustering method. So in this paper we attach the fuzzy scheduling algorithm at packet in the path nodes. The fuzzy scheduler nds the priority index of the queued packets. It combines the input parameters such as data rate, expiry time, and queue length for nding the priority index. Without fuzzy scheduling the packets are scheduled in rst in rst out.Herebwe used the software agents for both in routing and fuzzy priority

[1] 1. Subir Kumar ,T.G Basavraj ,C Puttamadappa,1. Ad-hoc mobile wireless Networks:Principles.Protocols and Applications,AUERbach Publication.2008. [2] Padmini Misra, Routing Protocols for ad hoc mobile wireless Networks, http://www.cse.ohio-state.edu/ jain/cis78899/ftp/adhocrouting/TDRP, Nov-1999. Moukthtar A.Ali, Ayman EL-SAYED, Ibrrahim Z.MoRSi,A Surve of Multicast Routing Protocol for Mobile Ad-Hoc Networks,. IEEE communication surveys tutorials, Vol 11.No 1,rst Quarter 2009. [3] Sung-Ju Lee, Mario Gerla, and Ching-Chuan Chiang, On-Demand Multicast Routing Protoco,in proc.IEEE Wireless Communication and Networking Conference(WCNC99),vol.3,pp.1298-1302, New ORLEANS,lA,usa,sEPTMBER 1999. [4] 9. Rajashekhar Biradar, Sunilkumar Manvi, Member, IACSIT , Mylara Reddy, 9. Mesh Based Multicast Routing in MANET:Stable Link Based Approach, proc. International Conference on Mobile Computing, vol. 353, pp. 153181, 1996. [5] 10. Blanca Alicia Correa, Laura Ospina, Roberto Carlos Hincapie, 10. Surve of clustering techniques for mobile ad-hoc networks, Proc. Manuscript, received January 20, 2006; revised April 18, 2006. This work was supported by the IEEE. [6] R11. R. Murugan, A. Shanmugam, Badji , 11. A Cluster Based Authentication Technique for Mitigation of Internal Attacks in MANET, proc. European Journal of Scientic Research ,ISSN 1450- 216X Vol.51 No.3 (2011), pp.433-441 EuroJournals Publishing, Inc. 2011 [7] Byung-Gon chun and Marry Baker,Evaluation of Packet scheduling algorithm in Mobile adhoc networks, proc. in mobile Computing and commun.review,, Vol.6, no.3, pp no. 36-49. [8] C.Gomathy and S.Shanmugave, Supporting QoS in MANET by a Fuzzy Priority Scheduler and Performance Analysis With Mixed Trafc, The 2005 IEEE International Conference on Fuzzy Systems [9] 13. M. S. Kakkasageri, S. S. Manvi, B. M. Goudar, 13. An Agent based Framework to Find Stable Routes in Mobile Ad hoc Networks (MANETs), International Journal of Computer Science and Security, volume (2) issue (3), pp no. 48-56, 2003. [10] http://www.computing.surrey.ac.uk/courses/cs364/FuzzyLogicFuzzySystem 3.ppt. [11] Michael Luck, Peter McBurney and Chris Preist,Agent Technology: Enabling Next Generation Computing A Roadmap for Agent-Based Computing , January 2003. [12] T. Magedanz, K.Rothermel and S. Krause, Intelligent agents: An emerging technology for next generation telecommunications, IEEE Proc. Globecom, pp. 464-472, San Francisco, CA, 1996.