978-1-5386-2640-5/17/$31.00 ©2017 IEEE Toward Improving Coding-aware Routing in Multi- hop Wireless Networks Parisa Rafiee, Ghasem Mirjalily Department of Electrical Engineering, Yazd University, Yazd, Iran p.rafiee@stu.yazd.ac.ir, mirjalily@yazd.ac.ir Abstract—Due to the effectiveness of network coding on improving the performance of wireless networks, network coding- aware routing algorithms have been emerged. In line with this new generation of routing schemes, an efficient coding-aware routing approach is proposed in this paper; in which a fuzzy system is designed to be used in each node. In this approach, some effective factors are considered for selecting proper routes, such as the links bandwidth and the coding ability of nodes. The designed fuzzy system brings about an efficient trade-off among all factors. In proposed routing algorithm, according to the overall status of the nodes, a path consisting of suitable nodes in terms of the considered factors will be selected for data transmission. Simulation results confirm that our proposed method outperforms the previously proposed schemes, in terms of throughput, delay, and packet delivery ratio. Keywords—routing, network coding, multi-hop wireless networks. I. INTRODUCTION Network coding (NC) [1] has been proposed as an elegant approach for improving the performance of wireless networks in recent years. By using NC, instead of sending data packets in native form, each node tries to encode two or multiple packets with each other, using an appropriate logical or mathematical operations; such as XOR [2,3], or (random) linear coding [4,5]. For example, in the case of XOR coding, various data packets will be XORed to each other bit by bit, if coding conditions are true. NC can be employed in both inter-session [6,7] and intra- session [8] scenarios to enhance the performance of the wireless networks. To illustrate the advantages of NC more clearly, a 2-hop chain network, consisting of 3 nodes named A, B and C, is presented in Fig.1. In this sample network, node A aims to deliver data packet  ଵ to node C. Simultaneously, node C aims to transmit  ଶ to node A. Due to the restrictions of the transmission range of each node, node C acts as a relay node in this transmission. If NC is not employed in this scenario, 4 time slots will be required to complete data transmission. A possible scheduling, described in Fig.1, can be as follows: (1) node A sends  ଵ to B, (2) node C sends  ଶ to B, (3) node B sends  ଵ to C, (4) node B sends  ଶ to A. Figure 1. Required time slots without using network coding Figure 2. Required time slots while using network coding However, if NC is employed in this scenario, packets can be encoded together. Under the assumption that XOR-type NC is employed, data transmission can be completed through 3 time slots as described in Fig. 2: (1) node A sends packet  ଵ to node B, (2) node C sends packet  ଶ to node B, (3) node B XOR- encodes the contents of packets  ଵ and  ଶ and broadcasts the resultant coded packet  ଵ ⊕ ଶ to both A and C. Given the fact that destination node A (C) has the native packet  ଵ ( ଶ ) it has formerly sent, the intended packet  ଶ ( ଵ ) can be extracted by a simple XOR. In this way, one time slot is saved, and can be used for other data transmissions. In this regard, NC takes the advantage of broadcast nature of wireless medium to improve the throughput of wireless networks. In addition to improving throughput, other benefits of using NC is elaborated in [9]. Given diverse privileges, NC has been considered as an efficient technique to be utilized in routing mechanisms of wireless networks. In general, combining both routing and NC improves the performance of the networks. However, if the routing metric used in the routing procedure is not coding-aware, the maximum coding gain will not be acquired. In other words, in order to establish routes containing coding opportunities, the routing mechanism must be coding aware to discover the coding opportunities.  ,EEE 4 th International Conference on Knowledge-Based Engineering and Innovation (KBEI) I Dec. 22 QG , 2017 ,ran University of Science and Technology) – Tehran, Iran I 0689