Research On Congestion Control And Routing For Space Information Network

Against the backdrop of the rapidly growing global economy and the constant space technological breakthroughs, the construction of space information networks (SIN) is speeded up. Human exploration requirements of the universe makes high demands on SIN performance. However, differences among the information transmission environments of different spatial areas increase the difficulty of SIN construction. Therefore, it has guiding significance to improve network capabilities and optimize the allocation of resources that to grasp the network architecture from the overall layout. Due to the tremendous growth of worldwide data quantity, the network congestion and optimal routing problems are becoming more and more apparent. As the tie and bridge of learning the space information, exploiting space resources by terrestrial users, the researches on congestion control and routing for SIN is very important. By means of Queuing Game, Stochastic Differential Game theory and so on, this paper carries out the thorough studies of congestion control and routing in the satellite backbone network, congestion control in the deep space communication network and congestion control in the stratospheric communication network. It is committed to building efficient, reliable SIN. The main works are summarized as follows:(1) The SIN model based on hierarchical structure is designed. According to different requirements of the space business and characteristics of the communication environment, the SIN is consist of the satellite backbone network for connection, the deep space communication network which is up supported by the satellite backbone network, and the stratosphere communication and terrestrial network which is down aided by the satellite backbone network. Since the important position nexus, the satellite backbone network achieves global coverage through the network structure composed of the LEO/MEO/GEO three layer satellites. Considering the influx of future spatial information, SIN model proposes requirements of reasonable resource allocation and routing to the satellite backbone network. To solve the access for spacecraft, such as high-speed deep space probe, the data transmission requirements of selectively receiving services for users are proposed in network layer of deep space communications. Finally, in stratospheric communications, the requirements of congestion control is proposed for assistive communication dealing with the large amounts of data from the terrestrial network. The SIN model provides clear goals for the algorithm design of each layer in SIN.(2) The congestion control and routing algorithm based on the business classification in the satellite backbone network is designed. Due to the satellite backbone network not only needs to connect deep space communications equipment, and undertake the various types of tasks delivered by the users from the stratosphere and terrestrial network, to distinguish the levels of heterogeneous users is the premise of guarantee for network efficiently running and improving the service quality. In this paper, through the establishment of the satellite backbone network model based on queuing game theory, the heterogeneous users with different levels of business are described by different gains of users after getting services. Under the condition of the optimal system welfare, the optimal admission fee is obtained by solving the model. Users choose whether to join the node according to their own information. It is benefit for network traffic dispersion. Then the congestion control scheme for users with different levels in the satellite backbone network is determined. On this basis, routing factor of the satellite backbone network is introduced to design a optimal routing for heterogeneous users. This routing factor is consist of the optimal admission fee, the current waiting cost and the routing transmission cost. As a result, the system welfare reaches the maximum and the simulations show that this algorithm promotes the network performance.(3) The congestion control algorithm based on impatient users in deep space communications is designed. Although the deep space communication environment is similar to the satellite communications environment, the business needs of the deep space communication users have their uniqueness. Based on satellite congestion control model of satellite backbone network, To satisfy the request of fast access and accepting the service for the deep space communication users, the model needs to be modified. The degree of impatience is introduced to express the leaving probability of deep space users after joining into the node queue. Therefore, the queuing game based congestion control model for deep communication is presented. After calculating the threshold of the queue length, a preliminary queue is designed for the impatient leaving users. At last, a preliminary queue congestion control algorithm considering impatient users based on queuing game theory is proposed. It optimizes the payoff and eliminates the negative effect from the impatient leaving users. The simulation results show that this algorithm can promote the network performance.(4) The congestion control algorithm based on dynamic window in stratosphere communication is design. The communication congestion control of stratosphere communication in SIN mainly aims at the information transmission with the terrestrial network. Therefore, using the stochastic differential game theory, a dynamic window congestion control algorithm is proposed on basis of improving the TCP congestion control scheme. Firstly, according to the influences of network performance from data transmission rate, RTT and such elements, the network traffic equilibrium model is built. Then an optimal strategy of congestion window size is obtained by solving the built game model for designing the congestion control algorithm. At last, Simulation results show that the designed algorithm can improve the network throughput and reduce the resource waste.