Performance Challenges and Issues in Wireless Mesh Networks

Abstract

Wireless Mesh networks have the potential to provide inexpensive and newlinequick access to the internet for military communications, surveillance, education, newlinehealthcare and disaster management. This work caters to the growing newlinehigh-bandwidth demands by providing low delay and high throughput by designing newlineefficient, robust scheduling algorithms for wireless mesh networks. newlineThis thesis focuses on link scheduling in wireless mesh networks by taking newlineinto account physical layer characteristics. The assumption made throughout newlineis that a packet is received successfully only if the Signal to Interference newlineand Noise Ratio (SINR) at the receiver exceeds a certain threshold, termed newlineas communication threshold.The thesis also discusses the problem of genetic newlinealgorithm and routing. newlineFirst, we consider various problems on link scheduling algorithms like ALS, newlineGP, SGLS. We propose link scheduling algorithms based on Ant Colony Optimization. newlineOur algorithms work is evaluated and compared with the existing newlineones and shown that the proposed algorithm works better than the existing. newlineNext, we investigate a related scenario involving genetic algorithm. In newlinethis we have considered the basics of any genetic algorithm i.e. mutation newlineand crossover. We have considered the Binomial and Erlang distribution to newlinepredict at what level the mutation or crossover should be done. newlineFinally, we consider the routing metrics to analyse the performance of newlineWireless Mesh Network. In this we have considered various routing metrics newlineand evaluated on varying size of node. A comparative analysis is done and newlinewe realized that in order for a metric to perform well, it does not necessarily newlineneed to meet all the requirements as many metrics are designed deliberately newlineto favor certain criteria and ignore others. newlineKeywords: Wireless Mesh Network, Scheduling algorithm, Genetic Algorithm, newlineMutation and Crossover, Network performance, Ant Colony Optimization, newlineRouting newlineVI newline