Research On Optimization Mechanism For Network Coding

As the mobile internet era is coming, it leads to much higher requirements for the performance of wireless communication which brought great convenience to our daily life and work. Most of the researchers lay their emphasis on how to avoid the disadvantage brought by the unavoidable interference which is caused by the broadcasting nature of wireless network. Based on the idea of the avoiding interference, the research community develops a serious of techniques to improve the performance of network. However, the bottleneck still shows up with the traditional idea. Until the idea of network coding is introduced to the research community, the performance of wireless communication could step into a bright future.Because of the fact that the design goal of network coding is to make use of the unavoidable interference in wireless communication, the traditional wireless system architecture could not satisfy with idea of network coding. Therefore, how to implement the network coding with current wireless system architecture by lower cost becomes a hot topic in recent years. This thesis focuses on the design of optimization mechanism for COPE. The main contributions of this thesis are listed as follows:(1) Based on the fact that the key idea of classical routing protocol is the shortest path which could not cooperate with COPE to find and make use of network coding opportunity, a method to customize to the classical routing information data is proposed. After inserting the uplink and downlink information together with preserved packets reports of the nodes, our mechanism could find and judge the network coding opportunity by cooperating with original opportunistic listening and neighbor state’s guessing algorithm to achieve system gain. It could be observed by the result of simulation experiment that our mechanism could improve the system throughput gain effectively by very low cost to be compatible with current network architecture.(2) Aiming at the fact that there existed more multi-rate topology in reality, an optimization mechanism for this situation is proposed. By importing the three new variable: required transmission number, matched coding rate, congestion avoidance rate, and inserting the information into routing control packet, the mechanism could split the high rate data flow into lower rate ones and create more network coding opportunity while avoiding the occurrence of congestion by choosing a variety of path. The simulation result point it out that our mechanism could improve the system gain further than what COPE did particularly in multi-rate topology. Meanwhile, it could also be proved that our mechanism could help network to avoid congestion and maximum the ability of each node.