Research On Optimal Retransmission Times Algorithm For Delay-Sensitive Networks

Since the development of network coding theory research, it has received extensive attention of scholars and a lot of research both in the domestic and abroad. Network coding theory research mainly divided into two aspects at present, one is theoretical knowledge research, the other is application technology research. Network coding has obvious advantages to improve the reliability of data throughput and data transmission in the actual application of the network environment. It is a kind of information exchange technology in coding and the integration of routing. In multi-hop or single-hop network, network coding changed the ability of original node which can only store and forward data. What’s more, it also make the sending node has capacity for sending data packet group coding according to specific conditions.This thesis studied the relevant theoretical knowledge and the results of research, according to monitoring and retransmitting the data packets of the delay sensitive single-hop wireless network. Besides, we put forward a kind of orderly optimal retransmission times of network coding algorithm which called OORT algorithm. This algorithm uses the values of complementary hash value field to find and divide all data packets to constructs the orderly send coding combination. Compared with the previous network coding algorithm, it can generate more opportunities for coding, thus effectively reduce the retransmission times.Due to network coding need to receive enough coded packets when decoding, it lead to a problem in decoding into delay sensitive network. So it needs to be further studied how to design the network coding algorithm in the delay sensitive network environment to adapt to the network limit of the decoding delay and the characteristics of low cache. OORT algorithm in this paper, on the one hand, it can further improve the efficiency of data packet retransmission to effectively reduce data packet retransmission times. On the other hand, when the initial listening probability is low, the initial listeners can optimize the overall network delay according to the change of the monitor probability of the primary nodes, in order to adapt to the delay sensitive network characteristics. Compared with other algorithms, OORT can respectively reduce about 12% of the average retransmission times and 11% of the average delay in the actual network environment simulation.