Developing Expected Forwarded Counter (EFW)

Developing Expected Forwarded Counter (EFW) Abstract Wireless mesh networks have emerged as adaptable and low cost networks. Expected forwarded Counter (EFW) is a cross layer metric introduced to deal with the problem of selfish behaviour in order to provide reliable routing. This paper proposes an enhancement to the EFW, by considering congestion incurred due to selecting only high quality paths. The performance of proposed metric is evaluated through simulation. Simulation results show that overall routing performance is increased in terms of throughput and packet delivery ratio. Introduction Wireless Mesh Network (WMN) is a promising technology for the next generation wireless technologies. The Mesh Networks are self-organized, self-configured and easily adaptable to different traffic requirements and network changes. Routing is a challenge in Wireless Mesh Network (WMN) due to unpredictable variations of the wireless environment. Initially, to select a path with highest delivery rate in wireless mesh network, metrics that capture link quality have been introduced. But, most of these metrics are designed by assuming that each wireless mesh router participates honestly in forwarding process. While this assumption may not be valid in presence of selfish routers which may get profit from not forwarding all traffic. Selfish users utilize the network resources for its own benefit but unwilling to spend for others. Such selfish behavior reduces network delivery reliability. Metrics have been introduced to detect and exclude selfish nodes in a route to destination. These metrics do not consider quality of links, hence cannot select best path from source to destination. Cross layer metrics were used to consider both link quality and selfish behaviour of node [2] in order to select a high performance path. This kind of solutions may cause only high quality to get used and other links will get unused. This will make links to be congested seriously and in turn cause performance degradation. In this paper we propose metric that combines link quality and congestion information from MAC layer and forward reliability of node from routing layer. The rest of this paper is structured as follows. Section II discusses related work. Section III illustrates proposed work. Section IV presents results obtained through simulating proposed metric in comparison with ETX and EFW. Related work: Several works presented in the recent research literature focus on reliable data transmission in wireless multi hop networks with selfish participants. In recent years, several routing metrics have been proposed to select the path with the highest delivery rate in wireless mesh networks. The essence of all these metrics lies in the selection of reliable network paths, avoiding lossy wireless links prone to transmission errors. Some of these are discussed below. ETX (Expected Transmission Counter): Routing metrics for wireless mesh networks like ETX adopt a probabilistic model to represent the transmission reliability of a wireless link. Specifically, ETX measures the expected number of transmissions, including retransmissions, needed to correctly send a unicast packet over a wireless link. In order to compute ETX, it is necessary to estimate the packet loss probability in both directions since, in wireless networks based on the IEEE 802.11 protocol, the destination must acknowledge each received data frame Let (i, j)be a wireless link established between node i and j;Pij and pji denote the packet loss probability of the wireless link(i, j) in forward and reverse directions separately. The probability of a successful transmission on the wireless link(i, j)can therefore be computed as Ps,ij= (1−pij)(1−pji). Then, the expected number of transmissions necessary to deliver the data packet, considering both its transmission and the successive acknowledgment as required by the IEEE 802.11 protocol, can be evaluated according to expression Despite the purpose of selecting the most reliable paths, ETX does not model accurately the delivery rate of a network link, since it does not consider the forwarding behavior of the nodes that have established that link. In particular, ETX and its derived metrics do not take into account that a selfish node might discard the packet after its correct reception, if it benefits from not forwarding it EFW: To address the problem caused by the dropping behaviour of selfish participants, we combine the link quality measured by the ETX routing metric with the forwarding reliability of a relaying node j by improving the probabilistic model on which ETX is based. Let pd,ij be the dropping probability of a network node j((1−pd,ij)represents its forwarding probability). Since a network node can drop selectively the traffic sent by its neighbours, the dropping probability of any node j is identified both by the sending node i and the relaying node j. The probability that a packet sent through a node j will be successfully forwarded can be computed as pfwd,ij=ps,ij(1−pd,ij).Then, the expected number of transmissions necessary to have the packet successfully forwarded (Expected Forwarding Counter, EFW) can be measured according to the following equation. The first part of equation, which coincides with the ETX metric, considers the quality of the physical and MAC layers, whereas our contribution takes into account the network layer reliability. Therefore, EFW represents a cross-layer metric that models both the physical conditions of the wireless medium and the selfishness of the node with which the link is established. In addition to detecting the misbehaving nodes, the representation of the link reliability provided by the EFW metric permits to use the network paths with the highest delivery performance, without pruning the alternative routes that contain selfish nodes. Proposed work: The disadvantages of this solution are that nodes wanting to transmit packets will attempt to use the same high quality link and cause it congested seriously. And at the same time other links will get unused. MAC layer metric: Our proposed metric is based on the retransmission mechanism in MAC. The first part of this method is the success rate of transmitting frames based on the average number of retransmissions which we call Frame Transmission Efficiency (FTE) [13]. Fig. 1: Illustration of the retransmission mechanism at the MAC Layer The number of retransmissions of RTS and Data frames for each Successful transmission in MAC layer is supposed to represent the quality of that link and congestion instance. The success rate of sending frames is therefore a good estimate of both the quality and congestion of a link. From it the best quality links may be selected. The success rate of each link (FTE) is updated when a node forwards a Data packet to its neighbour and passes it up to the routing protocol. ACK Failure Count denotes the number of Data retransmission and RTS Failure Count denotes the number of RTS retransmission. The kth packet will send from Node S to Node D. The number of retransmission is assumed as Failure (k) and denotes below: Failure (k) = ACK Failure Count (k) + RTS Failure Count (k) (j) Thus FTE (k) between Node S and Node D is formulized as equation (4). It reflects the link quality and congestion situation of links. FTE (k) = We are using this frame transmission efficiency to represent link quality and congestion. It is MAC layer information. From routing layer we consider forward probability estimation. In cross layer fashion we combine information from both MAC and Routing layer to obtain congestion aware EFW. It is computed as follows Enhanced EFW or congestion aware EFW= Simulation Simulation scenario We performed simulations with NCTUns6.0 simulator that evaluates performance of the metric in comparison with EFW using OSPF (Open Shortest Path First) routing protocol. Performance Evaluation: To evaluate the performance of proposed metric in comparison with existing metrics ETX and EFW, the following variables are analysed Throughput Packet Delivery Rate Drop Rate From fig1. We can observe that the proposed metric has more throughput than the other routing metrics in wireless mesh networks. By this we can understand that the proposed metric selects better path in presence of selfish nodes in comparison with other metrics. From fig2. It seems that congestion aware EFW has more packet delivery rate when compared with other metrics. Conclusion: In this paper we introduced an enhancement to existing cross layer metric called Expected forward counter (EFW). In this metric we replaced link quality metric obtained from ETX metric with Frame Transfer Efficiency (FTE) metric which considers not only link quality but also congestion of link. As the proposed metric in cross layer fashion combines MAC layer observations of link quality and congestion with routing layer observations of forward probability estimation, it gives better performance in comparison with ETX and EFW metrics. Simulations results show that routing performance of OSPF in terms of throughput, packet delivery rate and drop rate has been improved in proposed metric. References: S. Paris, C. Nita-Rotaru, F.Martignon, and A. Capone, †Cross-Layer Metrics for Reliable Routing in Wireless Mesh Networks â€Å", in proc. IEEE/ACM TRANSACTIONS ON NETWORKING, VOL. 21, NO. 3, JUNE 2013. N. Nandiraju, D. Nandiraju, L. Santhanam, B. He, J. Wang, and D.P. Agrawal, â€Å"Wireless mesh networks: Current challenges and future directions of web-in-the-sky,† IEEE Wireless Commun., vol. 14, no. 4, pp. 79–89, Aug. 2007. S. Paris, C. Nita-Rotaru, F.Martignon, and A. Capone, â€Å"EFW: A cross layer metric for reliable routing in wireless mesh networks with selfish participants,† in Proc. IEEE INFOCOM, Apr. 2011, pp. 576–580 D.S.J De Couto, D. Aguayo, J. Bicket, and R. Morris. A High-Throughput Path Metric for Multi-Hop Wireless Routing. Wireless Networks, 2005. D. Johnson and G. Hancke, â€Å"Comparison of two routing metrics in OLSR on grid based mesh network,† Ad Hoc Netw., vol. 7, no. 2,pp. 374–387, 2009. Da Guo, Jun Li, Mei Song, Junde Song,â€Å"A Novel Cross-Layer Routing Algorithm in Wireless Mesh Network† in the proc. of IEEE International Conference 2007,pp 1-3,vol-07. Golnaz Karbaschi, A Link-Quality and Congestion aware Cross layer Metric for Multi-Hop Wireless Routing.2nd IEEE International Conference on Mobile Ad-hoc and Sensor Systems Washington, DC USA November 2005. M.E.M.Campista, P. M. Esposito, I.M.Moraes,L.H. M. Costa,O. C.M. Duarte, D. G. Passos, C. V. N. de Albuquerque, D. C.M. Saade, and M. G. Rubinstein, â€Å"Routing metrics and protocols for wireless mesh networks,† IEEE Netw., vol. 22, no. 1, pp. 6–12, Jan.–Feb. 2008. OSPF