Markov Decision Process, a Dynamic System for Enhancing the TCP Throughput in CR Network

In cognitive radio network TCP Throughput is one of the measure issues to improve the performance of the CR network. However most research work concentrated on performance improvement of TCP has weaknesses as follows:-The underlying parameters are only considered to increase the TCP Through-put, keeping the transport layer parameter unchanged. Hence to solve the above problem markov decision process base Algorithm is used in this paper. Formal decision analysis has been increasingly used to address complex problems in cognitive radio Network. The proposed work provides a study on the construction and evaluation of Markov decision processes (MDPs), which are powerful analytical tools used for sequential decision making under uncertainty. In this proposed work each CRN users autonomously decides modulation type and power to be transmitted in PHY layer, channels which is to be selected in MAC layer to get best TCP Throughput. Simulation result shows that the network can learn the optimal policy to improve the TCP Throughput in cognitive radio network.