Congestion Control Using Polynomial Window Size Adjustment Algorithms for Wired and Wireless TCP networks

This paper introduces and analyses non-linear congestion control using polynomial window size adjustment algorithms. These algorithms generalize the Additive Increase Multiplicative Decrease algorithms used for the TCP Networks. They provide additive increase using a polynomial of the inverse of the current window size and provide multiplicative decrease using the polynomial of the current window size. They are further parameterized by á and â. There are infinite numbers of TCP-friendly polynomial algorithms by assuming polynomial of different order. This paper analyses the performance of two polynomial models, for the wired and wireless TCP networks. TCP compatibility of these algorithms for the two models are evaluated using ns2, a discrete event simulator. One such model is also proved to be TCP-friendly. The results of simulation are compared with that of TCP/Tahoe, TCP/Reno TCP/New Reno, and TCP/Fast. The comparison shows that both algorithms perform better in terms of throughput.