An intelligent wireless ad hoc routing protocol

2006 
Ad hoc networks are self-organizing and self-configuring networks that do not have fixed infrastructures. The wireless ad hoc networks do not require a predefined configuration. Mobile stations in wireless ad hoc networks move around in the network, thus requiring routing to handle dynamically the constant changing of network topologies. We propose an algorithm that presents a new intelligent routing protocol for wireless ad hoc networks. The Intelligent Wireless Ad Hoc Routing (IWAR) protocol selects optimal routes, and also provides load balancing and fault tolerance. The IWAR selects routes based on available bandwidth, error rate, queue delay, process and propagation delay, and route stability. The concept of the artificial intelligence technique called Reinforcement Learning is used to heuristically define credits to routes based on throughputs, available queue buffer spaces, and route lengths. In addition, IWAR uses multiple routes to distribute the traffic among them to create load balancing and fault tolerance. This research includes the implementation of the IWAR algorithm and studies the protocol's performance and reliability. The protocol uses two heuristic functions that are used during the routing discovery and routing maintenance stages to select routes. The protocol evaluates the impact of the different components of each function to fine-tune the performance outcomes. Our study also compares the performance of the IWAR protocol with the well-known Ad Hoc On-Demand Distance-Vector (AODV) protocol. AODV protocol is a wireless ad hoc protocol that makes its routing decision based on the shortest path. The performance evaluation of both protocols studies lightly, moderately, and heavily loaded networks. In addition, the research includes the evaluation of both protocols with stable and unstable network configurations. Throughputs and delays are used to measure the performance of both protocols. In our study, the IWAR protocol outperformed AODV protocol by range from 10% to 21%.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    5
    Citations
    NaN
    KQI
    []