End-To-End Performance Oriented DTN Network Data Compression Algorithm

In extreme environment, the characteristics of high bit error rate, long communication delay, intermittent connection, asymmetric data rate mak e the traditional network system based on TCP/IP protocol can not work normally. In order to solve the problem of communication, the DTN protocol system were put forward in 2002 and have been widespread attention in the following years. The DTN add the Bundle layer above the Bundle layer to form a kind of network architecture. The Bundle protocol ensure the reliable transmission of the data through the “custody-to-carry forward” mechanism. However, the redundant information in the basic Bundle block waste t he precious link and bandwidth resources. Therefore, compressing the Bundle header files is an effective method to slow the progress of the protocol overhead.In this paper, we analyze the bundle protocol format detailed. According to the distribution of the characters stored in the dictionary of the basic block, this paper designs a stateless compression algorithm based on Huffman encoding. It can reduce the basic block size of 37% of the bundle protocol. According to the “storage-carry-forward” transmission characteristics, the stateful compression algorithm is designed in this paper based on the link state, introducing the concept of the background of the link and the background identifier. Initializing the node with the basic information required for the transmission through the first bundle, and the information is stored in the node memory as the background file. In the next data transmission process, the bundle protocol contains only the link information to ensure the reliable transmission of the curren t, using the unique identification to extract the background, so as to achieve the purpose of compression. The stateful compression is effective, and it can reduce the basic block size of 60% or so. It is better than the stateless compression. The implementation process of the state compression, as well as the required assumptions, is discussed in the fourth chapter. Combined with the stateful compression process, we design the graded compression algorithm in this paper. It achieve the purpose of compression through removing the information has been used.In this paper, we study the transmission of data in single hop and multi hop links, and establish the transport model. Selecting the delivery delay and the effective throughput as the main performance index, the mathematical expressions of the delivery delay and the effective throughput are established based on the physical model of the link transmission. Choosing the two performance indexes as the main criteria, we compare the transmission performance of the link with the data before and after the compression. The simulation results show that the stateless and stateful compression can shorten the delivery delay and improve the effective throughput of the data in the link. In extreme environments, the stateful compression can increase the effective throughput of 1.5% or so. In contrast, the improvement of the transmission performance using stateful compression in better than that using the stateless compression.