Improvement On Quality Of Transport Layer Strategy In CRAHNs Communication

Cognitive Radio Ad Hoc Networks bring more challenges, which come from independent spectrum transmission link, multi-hops delivery, changing topology and node mobility etc, than classical Cognitive Radio Networks. Recently, the typical research works of CRAHNs are mainly integrated in opportunity spectrum access and spectrum utilization. While the core functions of spectrum perception, exchange and sharing are rapidly developing, but cognitive radio technology has a significant effect on the upper protocol, especially in mobile ad hoc network. At the same time, the research of upper protocol layers, such as Transport Communication Layer, remains in the infancy stage and the research degree cannot satisfy the existing network application demands.The researches of TCP versions in high speed wireless networks indicate that the performance of Vegas version is better and more stable than others in the Ad Hoc networks of high dynamic characteristic of small groups. Nevertheless, Vegas cannot make full advantage of available bandwidth to transmit packets since incorrect bandwidth estimates may occur due to frequent topology changes caused by node mobility. This paper proposes an improved TCP Vegas based on the grey prediction theory, named TCP-GVegas, for multi-hop ad hoc networks, which has the capability of prediction and self-adaption. GVegas has three enhanced aspects in the phase of congestion avoidance. The first aspect, the lower layers’ parameters are considered in the throughput model to improve the accuracy of theoretical throughput. Secondly, the prediction of future throughput based on grey prediction is used to promote the online control. The third aspect, the optimal exploration method based on Q-Learning and Round Trip Time quantizer are applied to search for the more reasonable changing size of congestion window. Besides, the convergence condition and rate of actual throughput of grey prediction is briefly analyzed by lyapunov method, and the time and space complexity of improved algorithm-GVegas is also analyzed. In the end, we implement the improved method in NS2, and set up classical line Ad Hoc network and group ad hoc model under reference point scenarios to testify the performance of GVegas. The simulation results show that the TCP-GVegas achieves a substantially higher throughput and lower delay than Vegas in multi-hop ad hoc networks. Besides, the improved algorithm can friendly coexistence with other TCP flows.