基于Q学习算法的Ad Hoc网络自适应DSR协议研究

Study of Ad Hoc Network Adaptive DSR Protocol Based on Q-learning Algorithm

提出了一种用于Ad Hoc网络改进的基于Q学习算法的自适应动态源路由(DSR)协议。该协议能在网络状态较好的情况下减小开销,在网络状况较差的情况下改善端到端时延和吞吐量性能。仿真结果表明,自适应Q-DSR协议能够提升传统DSR协议的性能,是简单高效的路由协议。

This paper puts forward an adaptive dynamic source routing（DSR） protocol based on Q- learning algorithm for improving Ad Hoc network. The protocol can reduce the redundant routing spending in better network status and improve the end to end time delay and the throughput performance in worse network status. The simulation result shows that the adaptive Q-DSR protocol can advance the performance of traditional DSR protocol,is a simple and effective routing protocol.