Modeling Wireless Channel for Ad-Hoc Network Routing Protocol Merlinda Drini, Tarek Saadawi Dept. ofElectrical Engineering, City College and Graduate Center of City University ofNew York New York, NY 10031, USA {mdrini@gc.cuny.edu, saadawi@ccny.cuny.edu} Abstract. Enabling transmission over ad-hoc networks is more challenging than over conventional mobile networks because a connection path in an ad-hoc network is highly error-prone and the path can go down frequently. Although it is well-known that wireless channels are time varying because of user mobility and multi-path propagation effects, experiments for various types of channels show that the basic channel parameters can be stable for short time intervals. Therefore, a wireless channel can be adequately represented by a set of stationary channel models. There are number of wireless channel models proposed in literature. For simplicity, we used a two state Markov model known as a Gilbert-Elliot model and then in order to model more appropriately a real communication channel, a three state Markov model is analyzed here. We demonstrate our concept by applying it to the Dynamic Source Routing protocol (DSR). In our proposed modified DSR, both the route discovery and route selection are based on physical layer parameter and the link monitoring function located at each node. Since routing selection based on the link quality is implemented, the minimum transfer delay from source to destination and the maximal throughput may be obtained Simulation results show that according to the channel quality, the delay and throughput performance show remarkable performance when compared to traditional DSR. Introduction The major problem with transmitting information over wireless channels is the issue of link reliability. Wireless channels have high channel bit error rate and limited bandwidth.. The high bit error rate degrades the quality of transmission and the network performance. The notion of a link in wireless ad hoc networks is different from that in the wired network, where the links exists between two nodes only if they are connected by a physical medium. On the other hand, in mobile wireless ad hoc networks, the links exist theoretically between any pair of nodes. Prepared through collaborative participation in the Communications and Networks Consortium sponsored by the U.S. Army Research Laboratory under the Collaborative Technology Alliance Program, Cooperative Agreement DAADI9-01-2-0011. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation thereon. 978-1-4244-2703-1/08/$25.00 ©2008 IEEE Each wireless node can communicate with any other node within its transmission range, which depends on the SNR at the receiver and the coding scheme used by the transmitter. Routing in MANET (Mobile Ad Hoc Network) is challenging due to the dynamic nature of network topology and the resource constraints. To maximize the channel resource utilization and minimize the network transfer delay along the path, the shortest path with minimum hops scheme is often adopted. Reference [I] shows that routing in multi-hop wireless networks using the shortest-path metric is not a sufficient condition to construct good quality paths, because minimum hop count routing often chooses routes that have significantly less capacity than the best paths that exist in the network. However, the quality of wireless channel among the mobile nodes is time varying due to fading, shadowing and pathloss. Given that the shortest-path metric does not take into account the physical channel variations of the wireless medium, it is desirable to select the routes with minimum cost based on some other metrics which are aware of the wireless nature of the underlying physical channel. In self-organized network, there are many other metrics to be considered: Power, Packet Loss, Maximum available bandwidth etc. These metrics should come from a cross-layer approach in order to make the routing layer aware of the local issues of the underling layers. In this paper we perform the stochastic analysis of a Wireless communication channel. Our study is based on a Finite State Markov Model. The channel switches between different states. Each state corresponds the probability that a packet sent by the transmitter will be received by receiver or will be lost. The transition between the different states of the channel is administrated by a Markov chain; this Markov chain is not observed directly, but the received packet provides some probabilistic information about the current state of the channel. Related work The idea of studying varying channel conditions has been widely used in communications. There are a number of wireless channel models proposed in literature. But generally these models capture behavior of the network at MAC layer. Reference [2] has introduced an enhanced MAC protocol for multi-channel and multi-rate IEEE 802.11 which enables wireless ad hoc networks to 549