A Research Of Adaptive Systematic Random Network Coding Scheme On Satellite Communications

Satellite communication has many advantages, such as wide coverage,flexible communication mode and fast construction. Satellite communication has become a vital link in modern global communication network. However, satellite communication is also faced with a series of problems such as high delay of communication and susceptibility to interference of communication link. As TCP is widely used for communication reliability, TCP congestion control mechanism is applied to satellite link. But the throughput of TCP is seriously affected by the high delay of satellite link and the packet loss. In recent years, many scholars studied this problem, proposed the use of network coding theory to optimize the satellite link layer transmission scheme. But the existing research focus on the choice of static coding scheme. There is few dynamic coding strategy research for changing environment.The association in link Bit Error Rate (BER), coding block size and communication quality in the system using random linear coding is introduced in this paper. After simplifying and abstracting the system, the relation in them is given based on probability theory. It proves that the encoding window size affects the performance of Systematic random Network Coding (SNC) on the satellite link when the bit error rate of the satellite channel is constant. There exists a local optimal solution when the coding window size determines the value range. Furthermore, the optimal solution changes with the satellite channel error rate changing.Based on the above conclusions, this paper proposes an adaptive SNC coding strategy. The strategy consists of two core parts: one is to detect and track the error rate of the channel accurately in real time. The other is to adjust the SNC coding strategy adaptively according to the detected dynamic channel error rate while maintaining reliable and stable communication. In the simulation experiment, the TCP throughput of the satellite link is up to 30% higher than that of the traditional satellite link.The mathematical model of satellite channel BER variation is established for the purpose of accurate estimation and track of the bit error rate of satellite link in real time. A Kalman filter algorithm for continuous time non-uniform sampling scene is proposed according to satellite communication characteristics. The algorithm is universal so that convenient for subsequent research. Under the premise of certain bit error rate, this paper proposes the adaptive SNC coding strategy of satellite communication. In addition, it carries on the system design as well as verify the positive effect of the design. It shows that the heuristic significance to introduce the adaptive concept into network coding and apply it to satellite communication.