QoS routing in MANET through Cross-Layer Design with BER and modifying AODV Ram S. Dahal Information and Communication Technologies FoS Asian Institute of Technology Pathumthani, Thailand ram.dahal@gmail.com Teerapat Sanguankotchakorn, Senior Member IEEE Telecommunications FoS Asian Institute of Technology Pathumthani, Thailand teerapat@ait.ac.th Abstract-Many shortest path routing protocols have been proposed neglecting the effect of lower layers. Here, the Modified Ad-hoc On Demand Distance Vector routing based on Bit Error Rate (MAODV-BER) is proposed, where the route discovery of AODV has been modified to achieve the stable route by obtaining Bit Error Rate (BER) information from physical layer through cross-layer approach. Due to the use of multimedia applications in Mobile Ad-hoc Networks (MANET) strict Quality of Service (QoS) is required. So, for QoS fulfillment, the bandwidth and delay requirements are added in each route message. Finally, the path having minimum BER as well as minimum hop count and fulfilling QoS requirements is selected. I. INTRODUCTION Wireless communication due to its ubiquitous nature is the main research area in the communication world. Due to the varying bandwidth nature and strict QoS with limited battery life in ad-hoc network, constraints arise for supporting multimedia communication [1]. MANET as ad-hoc network is a distributed network where shortest path routing is not suitable because it selects the minimum hop-count route which is not of best quality [2]. Most of the routing protocols like AODV, DSDV, DSR, etc. use hop-count as a metric for route selection in MANET, so these routes selected by considering hop-count may not be the good quality link, because these links usually have poor SNR, higher Frame Error Rate (FER), low throughput, etc [2]. As well as due to the availability of multimedia traffic in MANET, finding the feasible route and fulfilling the multiple constraints (e.g. Bandwidth, delay, etc) of the link creates the main issue for QoS routing and achieving efficient utilization of network resources at same time [3]. For the optimization of wireless networks, the challenges that arise from the link and the QoS requirement that is generated from the application have to be considered. The requirements that are generated from the applications can be satisfied by adopting the rate, power and coding at the physical layer given the condition for current channel and network. Thus, to obtain the highest possible adaptivity, knowledge has to be shared between all the layers. This is called cross-layer design [4]. In this research, bandwidth and delay QoS requirements of applications are studied by modifying existing AODV and cross-layer design is made to provide the physical layer information i.e. BER to the network layer. RREQ and RREP message of AODV are added with the fields: 1) Bandwidth 2) Delay and 3) BER as well as HELLO message is modified to contain available bandwidth which is used for updating the routing table in intermediate nodes. The rest of the paper is organized as follows. Section II describes about related works and section III explains about proposed routing algorithm MAODV-BER. Section IV explains about simulation parameters and performance evaluation metrics. In section V analysis of result is done and finally conclusion is given in section VI. II. RELATED WORKS Several works have been carried out to integrate QoS with AODV and cross-layer method has been done to provide BER information to network layer. A. Akther and T. Sanguankotchakorn [5] proposed NQoS AODV to support multi-constraints QoS. Here Route Request (RREQ), Route Reply (RREP) messages are modified to fulfill the QoS as per the application requirements. H. Sun, and H.D. Hughes [6] proposed Adaptive QoS Routing where DSR is modified for routing function and local statistical computation is carried out by cross layer mechanism considering different parameters from physical layer to network layer and finally based on this knowledge, adaptive QoS routing decision is made. MAODV (Modified AODV) where the route is selected with the shortest possible longest hop length which has better link quality i.e. by selecting the path with lesser end to end BER was proposed in [7]. III. PROPOSED MAODV-BER In the proposed routing algorithm MAODV-BER, AODV routing policy is being used to find the route from source to