Research On Performance Evaluation Methods Of Transmission Control Algorithms In Ad Hoc Network

With the development of modern wireless communication technologies, it is possible for human beings to communicate with each other anywhere and anytime. Ad hoc network, one of the modern wireless communication technologies, can work without prefixed infrastructure. However, problems such as energy constraints, instability of radio channel and node mobility become bottleneck in the process of wide application of ad hoc network. Performance deterioration of traditional TCP is one of the problems encountered over ad hoc network. How to work out a reliable and efficient transmission algorithm becomes a focus nowadays.The thesis analyzes and compares fundamental ideas under different transmission control algorithm optimizations. The results indicate tradeoffs between algorithms. Absence of an overall performance evaluation method is the key reason why so many optimizations exist. Fuzzy hierarchy analytic process, a sound performance evaluation method, integrates uncertainty of decision maker into the evaluation progress while retaining the simplicity of hierarchy analytic process.The performances of four transmission control algorithms, Tahoe, SACK, Reno and NewReno are simulated. The overall traffic loads are collected and reasons are listed out. The goal of hierarchy analytic structure in this thesis is to find out the most energy efficient algorithms. The hierarchy analytic process is used to compute weights of different algorithms. The applications of fuzzy eigenvalue method, fuzzy least square method and fuzzy linear programming method in fuzzy comparison matrixes give different results. Advantages and disadvantages of fuzzy eigenvalue method, fuzzy least square method and fuzzy linear programming method are compared and the final evaluation model is carried out on these three methods. The performances of Tahoe, SACK, Reno and NewReno are given at last based on the evaluation model.