End-to-End Differentiation of Congestion and Wireless Losses using a Fuzzy Arithmetic based on Relative Entropy

Abstract

This paper, according to the traditional transfer control protocol(TCP)exposures the performance decline in the wireless environment.. presents a formulation for aggregating opinions about the alternatives into a single consensus or compromise one. The approach which is used to derive the aggregation formula is minimum relative entropy inference. Based on the relative entropy, we introduce an end-to-end arithmetic. The main idea of our arithmetic is to regard the average packet loss ratio and RFD as two decision makers. Those two decision makers give opinions to the network status separately; utilize the principle of relative entropy to infer the best choice of the network status. Therefore, we can differentiate the character of losses using decision-making. Our simulation measurements suggest that this arithmetic can work well in differentiating losses and throughputs.