Adaptive Selfish Scheduler with M/M/1 and M/M/n Queuing Mechanism in MANET

Lakshmi.S
Research Scholar Sathyabama University Tamilnadu, Chennai slakmy@yahoo.co.in,,Chennai. Professor &Head, SSN College of Engineering Tamilnadu,Chennai radhas@ssn.edu.in

Radha.S

Abstract A wireless Ad-hoc network consists of wireless nodes communicating without the need for a centralized administration, in which all nodes potentially contribute to the routing process. In this paper, we analyze packet scheduling algorithm for selfish nodes to find those that most degrade the performance in the network. Hence, a scheduling algorithm to schedule the packet based on their priorities will improve the performance of the network. Packet schedulers in wireless ad hoc networks serve data packets in FIFO order. Here, we present a M/M/1 and M/M/N Queue scheduling mechanism in an selfish aware scheduler to schedule the packets for selfish nodes in the mobile ad-hoc networks, to determine the priority of the packets using Ad hoc on demand distance vector protocol as the routing protocol. The performance of this scheduler has been studied using ns simulator and the performance can be measured using performance metrics such as packet delivery ratio, end-to-end delay and throughput. It is found that this scheduler provides overall improvement in the performance of the system when evaluated under different load and mobility conditions. In this paper, the scheduling scheme, thus improves the QoS parameters in wireless ad hoc network. Hence, introducing a scheduling algorithm to determine which queued packet to process next will improve the overall end-to-end performance. This scheduler is designed so that it should schedule the packets to reach the destination quickly, which are at the verge of expiry. Keywords: Ad-hoc networks, selfish nodes, ASSQM, AODV, Trust value. I.INTRODUCTION The growth of laptops and 802.11/Wi-Fi wireless networking have made MANETs a popular research topic since the mid- to late 1990s. A MANET (Mobile Ad-hoc Network) [1] is a self configuring system of mobile nodes connected by wireless links. In a MANET, the nodes are free to move randomly, changing the networks topology rapidly and unpredictably. MANETs are decentralized, and

therefore all network activities are carried out by nodes themselves. Each node is both an end-system as well as a relay node (router) to forward packets for other nodes. Most of the routing algorithms designed for MANET such as DSR [2], [3] and AODV [4] are based on the assumption that every node forwards every packet. But some of the nodes may act as the selfish nodes. These nodes use the network and its services but they do not cooperate with other nodes. Such selfish nodes do not consume any energy such as CPU power, battery and also bandwidth for retransmitting the data of other nodes and they reserve them only for themselves. This paper discusses Trust based technique and credit based technique to detect selfish nodes in MANET. In reputation based scheme, network nodes collectively detect and declare the misbehavior of a suspicious node. Such a declaration is then propagated throughout the network. Credit based schemes provide incentives for nodes to faithfully perform networking functions. In order to achieve this goal, virtual (electronic) currency or similar payment system may be set up. Nodes get paid for providing services to other nodes. II. RELATED WORK In the last few years, different routing protocols for Ad-hoc networks have been proposed. But most of them tend to ignore the fact that all the nodes in the network will not necessarily fully cooperate in routing the packets from source to destination. As discussed by Ning and Sun (2005), in general, many Ad-hoc devices operate on battery power. Consequently, power consumption for each transmission has a certain cost and significance. So, in reality, the assumption that all nodes perform the task of forwarding data, from which they do not directly benefit, while consuming their own battery power, is not always achievable. There is little reason to assume that some nodes will not try to achieve the benefits of participating in thenetwork and avoid the disadvantages it involves. This could mean that some nodes may refuse to forward packets as expected and thereby decrease the efficiency of the network. Due to the dynamic nature of Adhoc networks, identifying nodes that express such malicious behaviour is a

difficult task. The node originating the transmission might be out of range for detecting the malicious act. Collaborative Reputation Mechanism or CORE [1]. It is similar to the distributed IDS by Zhang et al. and consists of local observations that are combined and distributed to calculate a reputation value for each node. Based on this reputation, nodes are allowed to participate in the network or are excluded. In their work, the authors specify in detail how the different nodes should cooperate to combine the local reputation values to a global reputation and how they should react to negative reputations of nodes. For the actual detection of selfish nodes, they only refer to the work of Marti.A similar approach is conducted by Buchegger et al. with the CONFIDANT system [3]. Again, they only marginally describe their detection mechanism and rely mostly on promiscuous overhearing. On the other hand, the authors of [13] proposed that supporting a MANET is a cost-intensive activity for a mobile node. Detecting routes and forwarding packets consumes network-bandwidth, local CPU time, memory, and energy. Therefore there is a strong motivation for a node to deny packet forwarding to others, while at the same time using their services to deliver own data. The Ad Hoc On-Demand Distance Vector routing protocol (AODV) is an improvement of the DestinationSequenced Distance Vector routing protocol (DSDV). It is based on distance vector and also uses the destination sequence numbers to determine the freshness of the routes. It operates on the On-demand fashion. AODV requires hosts to maintain only active routes. The advantage of AODV is that it tries to minimize the number of required broadcasts. It creates the routes on an on-demand basis, as opposed to maintain a complete list of routes for each destination. Therefore, the literature on AODV [2], classifies it as a pure on demand route acquisition system. The usage of the AODV protocol for mobile ad hoc networking applications provided consistent results for large scale scenarios [3]. The authors of [14] highlighted that the monitoring algorithm (PCMA) which can detect node misbehavior in term of selfishness (i.e. none forwarding nodes). Because most of the other mechanisms give the suspicious node, i.e. selfish node, some degree of trust, we completely avoided any trust for the selfish node by relying only on information form the neighboring nodes of the suspicious node, and not all of the neighboring nodes, but only neighboring nodes that sent/received a direct information from or to the suspicious node. By doing this, we will save power which is important in MANET. Furthermore the new mechanism can detect the selfish node in the presence of partial dropping when the selfish node does not drop all packets but sends some of them and drops other. Furthermore, the authors of [15] highlighted the MAC selfish misbehavior and proposed extensions for the

detection system of [16] under the assumption that at least one of the parties involved is honest. Their approach follows that of [16] where in the receiver assigns a back off value for the sender; however, both sender and receiver will exchange some additional commitment information to ensure complete randomness and to verify that none of the hosts is misbehaving prior to the assignment. Any detected misbehavior (whether from the receiver or the sender) is reported to a reputation management system. III. SELFISH
SCHEDULER

QUEUE MANAGEMENT

In our work, the main focus surrounds on demand routing protocols, where the route is discovered only when a node wants to send data to another node. The routing protocol used in this study is the AODV protocol. Selfish aware scheduling gives higher weight to data packets when there are selfish nodes through which the packets are transmitted from the source to the destination in a AODV protocol. Here we consider that the only the nodes which acts as a router nodes can be of selfish manner by dropping the RREQ. The striking feature is that packet needs to traverse, the more potential it has to reach its destination quickly and the less queuing it incurs in the network. The selfish aware data packet scheduler serves packets either in M/M/1 or M/M/n . We use a selfish aware scheduler instead of a static priority scheduler since the fair queue scheduler guarantees all service classes at least the configured amount of service chances, thus avoiding starvation. If we consider packet length variations to allocate the correct proportion of bandwidth, we can use weighted fair queuing. After examining the effects of giving priority to control traffic, we look at the effects of setting priorities in data traffic.

FIGURE 1.

ADAPTIVE SELFISH SCHEDULER QUEUE MANAGEMENT SCHEME

In this methodology, we implement two queues, one having the RREQ packets alone and it is the M/M/n Queue because the RREQ packets are broadcasted and the other having the data packet and the RREPs are scheduled in an M/M/1 Queue fashion, because the scheduler first unicasts the data packets when the node is detected a selfish node Thus the packet can propagate through the selfish node to attain the network throughput. Algorithm: Detection Of selfish nodes Notations: TV1: Trust value calculated for the first past-recent delivery of data packets. TV2: Trust value calculated for the second past-recent delivery of data packets. TVC: The current trust value of the node. ATVC: The average trust value of the node. SN: Source Node RN: Intermediate Node DN: Destination Node NHN: Next hop Node 1. SN broadcasts RREQ to establish the forward route with DN addresses. 2. SN receives RREP through the reverse route from the DNs 3. If (packets are from DNs or a Reliable Node) { 4. Compute (ATVC = (TV1+TV2)/2) 5. If (ATVC>=TVC) { 6. Route data packet through RN 7. else if (ATVC<TVC){ 8. RN is a selfish node 9. Call ASSQM Scheme ( ) 10.} do 11. Current RN =NHN 12.}. Adaptive selfish scheduler Queue Management scheme contains a scheduler which is supported by the packet classifiers and the buffer manager. The packet classifier differentiates whether the packet to be transmitted is a data packet or a control packet. The scheduler monitors the buffer status so that it can make decisions to forward a data packet first or a control packet first. Here the detection is based on the current trust value and the average trust value. if the average trust value is higher than the current trust value then the node is a reliable node else the node is a selfish node. Algorithm: Adaptive Management scheme selfish scheduler Queue

TV2: Trust value calculated for the second past-recent delivery of data packets. TVC: The current trust value of the node. ATVC: The average trust value of the node. SN: Source Node RN: Intermediate Node DN: Destination Node NHN: Next hop Node 1. if (ATVC<TVC){ 3. RN is a selfish node. 4. Monitor the buffer status. 5. Notify the packet type to the packet classifier. 6. Call higher priority M/M/1 Queue when packet routed is a data packet and RREP. 7. ELSE if (ATVC>TVC) 8. Enable Call higher priority M/M/n Queue when packet routed is a RREQ. 9. Forward the RREQ first to the RN 10. do 11. Current RN =NHN 12.} 13. End. IV SIMULATION AND RESULT Network Simulator-2 is used for simulations. 50 mobile nodes are randomly placed in a rectangular area 1000 m1000 m.The wireless channel capacity is 2 Mb/s. Each simulation can run for 50 seconds. Selfish aware queue management mechanisms are applied for varying number of the selfish nodes. A Simple algorithm is compared with ASSQM under the simulated conditions. A Constant Bit Rate (CBR) source is used as the data source for each node. Each source node transmits packets at a certain rate. Packet size is fixed at 512 bytes. The random waypoint model is used. The maximum allowed speed for a node is 10 meters per second. The following performance metrics are used to compare the two scheduling algorithm are the packet delivery ratio, control overhead and total overhead. PERFORMANCE METRICS For performance evaluation of secure selfish queue management scheme the following metrics are used: PACKET DELIVERY
RATIO

Packet delivery ratio is defined as the data packets delivered divided by the data packets expected to be delivered.
END TO END DELAY

Notations: TV1: Trust value calculated for the first past-recent delivery of data packets.

The average end-to-end delay of data packets is the interval between the data packet generation time and the time when the last bit arrives at the destination.

THROGHPUT It is the fraction of useful transmission beginning of the whether or not data the TABLE I SIMULATION PARAMETERS
Parameter No. of mobile nodes Type of channel Type of propagation Type of network interface Type of interface Queue Type of antenna Value 50 Description Simulation nodes

the channel capacity used for selects a destination at the simulation i.e., information packets correctly delivered to
PDR

94 92 90 88 86 84 82 10 20 30 40 50 NO OF MOBILE NODES WITHOUT ASSQM WITH ASSQM

Wireless Channel type Two Ray Ground

Channel Type Radio propagation model

FIGURE 1. PACKET
Phy/WirelessPhy Network interface type

DELIVERY

R ATIO

FOR

ASSQM

M/M/n and M/M/1 Queue Antenna/Omni Antenna AODV

Interface queue

Antenna model

The chart depicts that the packet delivery ratio decreases during the selfish node behavior can be increased on an average by 23% when algorithmic solution is provided to prevent the attack. The packet delivery ratio obtained after solution is better compatible with the results of existing literatures. B. END
TO END DELAY

Type of protocol

Ad-hoc on Demand distance vector Maximum simulation time Data packet size x-dimension of motion y-dimension of motion

Simulation time Packet size Terrain dimensions

50 512bytes 1000m

Figure 3 shows the comparison between end to end delay and number of selfish nodes for two scenarios namely without ASSQM and with ASSQM. From the figure, it is clear that end to end increases when there is selfish behavior but decreases when the solution is provided to the range of end to end delay with ASSQM
0.6 0.5 0.4 0.3 0.2 WITHOUT ASSQM WITH ASSQM

1000m

VII PERFORMANCE EVALUATION OF SSQM A. PACKET DELIVERY RATIO Figure 1 shows the comparison between Packet delivery ratio and number of Mobile nodes for two scenarios namely without ASSQM and with ASSQM.From the figure, it is clear that packet delivery ratio decreases when there is a selfish behaviour but increases when the SSQM is implemented to the range of packet delivered through the selfish nodes.

0.1 0 1 0 20 30 40 50 NO OF M OB ILE NODES

FIGURE 2.

END TO END DELAY FOR ASSQM

The chart depicts that the end to end increased during the selfish node behavior can be decreased on an average by 24% when algorithmic solution is provided to prevent the attack. The end to end delay obtained after solution is better compatible with the results of existing literature. C. THROUGHPUT

Figure 3 shows the comparison between Throughput and the number of mobile nodes for two scenarios namely without ASSQM and with ASSQM .From the figure, it is clear that Throughput decreases when there is attack but increases when the solution is provided with ASSQM.

[3].

Sonja Buchegger and Jean-Yves Le Boudec. Nodes Bearing Grudges: Towards Routing Security, Fairness, and Robustness in Mobile Ad Hoc Networks. In Proceedings of the Tenth Euromicro Workshop onParallel, Distributed and Networkbased Processing, pages 403410, Canary Islands, Spain, January 2002.

[4].
900 800 THROUGHPUT 700 600 500 400 300 200 100 0 10 20 30 40 50 WITHOUT ASSQM WITH ASSQM

[5].

[6].

[7].

[8]. [9].

NO OF MOBILE NODES

FIGURE 3. T HROGHPUT

FOR

ASSQM [10].

VIII CONCLUSION In this paper, a adaptive selfish scheduling algorithm is presented. This paper takes the nodes behavior load into account when scheduling. By means of simulation studies, the performance of this algorithm is compared with that of simple priority algorithm. Simulation results show that this algorithm performs better. The scheduling scheme of this paper mainly deals with best-effort traffic, but the quality of service provision is becoming more and more important to the deployment of MANET. In the future, this scheme will be used to support QoS. REFERENCES [1].
Pietro Michiardi and Refik Molva. A Collaborative Reputation mechanism to enforce node cooperation in Mobile Ad Hoc Networks. In Proceedings of the 6th IFIP Communication and Multimedia.Security Conference, Portorosz, Slovenia, September 2002. ]Charles E. Perkins, Elizabeth M. Royer, and Samir Das. Ad Hoc On Demand Distance Vector (AODV) Routing. IETF Internet draft, Mobile Ad-hoc Network Working Group, IETF, January 2002.

[11]. [12].

[13]. [14].

[15]. [16].

J. Broch, D.A. Maltz, D.B. Johnson, Y.C. Hu, and J. Jetcheva. A performance comparison of multi-hop wireless ad hoc network routing protocols.In Proceedings of ACM/IEEE MOBICOM, Dallas, TX, October 1998. Samir R. Das, Charles E. Perkins, and Elizabeth. Royer. Performance comparison of two on demand routing protocols for ad hoc networks. In Proceedings of the IEEE INFOCOM, TelAviv, Israel, March 2000. wih-Chun Hu, Adrian Perrig, and David B. Johnson.riadne: A secure On-Demand Routing Protocol for Ad hoc Networks. In Proceedings of MobiCom 2002,Atlanta, Georgia, USA, September 2002. Panagiotis Papadimitratos and Zygmunt J. Haas. Secure Routing for Mobile Ad hoc Networks. In SCS Communication Networks and Distributed Systems Modeling and Simulation Conference (CNDS 2002), San Antonio, TX, January 2002. V. R. Ghorpade, Y. V. Joshi and R. R. Manthalkar,Fuzzy Logic based Trust Management Framework for MANET, DSP Journal, Volume 8, Issue 1,December, 2008. S.Anuradha , G.Raghuram, K.E.Sreenivasa murthy,B.Gurunath Reddy, New Routing Technique to improve Transmission Speed of Data Packets in Point to Point Networks, ICGST-CNIR Journal,Volume 8, Issue 2, January 2009. Kai Chen and Klara Nahrstedt, iPass: An Incentive compatible Auction Scheme to Enable Packet Forwarding Service in MANET, In Proceedings of 24th International Conference on Distributed Computing (ICDCS04), Tokyo, japan, Mar 2004. Pandey, A. K. and Fujinoki, H., 2005. Study of MANET Routing Protocols by GloMoSim simulator.International Journal of Network Management; 15(6):pp. 393410. Hu, Y. C., Perrig, A., and Johnson, D. B., 2004. Secure Routing in Ad hoc Networks: Securing Quality-of- Service Route Discovery in On-demand Routing for Ad hoc networks. Proceedings of the 2nd ACM workshop on Security of ad hoc and sensor networks SASN '04. [P. Papadimitratos and Z. Haas. Secure routing for mobile ad hoc networks. In Proc. of CNDS, 2002. Demir, C and Comaniciu C, An Auction based AODV Protocol for Mobile Ad Hoc Networks with Selfish Nodes. Communications, 2007. ICC '07. IEEE International Conference in June 2007. TaragFahad &Robert Askwith A Node Misbehaviour Detection Mechanism for Mobile Ad-hoc Networks ,PGNet 2006. [P. Kyasanur and N. Vaidya. Selfish MAC layer misbehavior in wireless networks. IEEE Transactions on Mobile Computing., April 2004.

[2].