Application Of Ant Colony Algorithm On LEO Satellite Network Routing

With the rapid development of Internet, the satellite communication systems begin the network era both in civil and military area. Traditional geostationary orbit (GEO) satellite communication has been unable to meet the demand of the market, LEO satellite mobile communications system attracts increasing attention because of its short propagation delay and small transmission loss. In LEO satellite networks, each satellite can be linked with many other satellites via ISLs that is messages can be passed on via many different routings. And for LEO satellite movement and dynamic satellite network topology, a lot of mature routing arithmetics in terrestrial network are not suitable in satellite networks for the characteristic of the satellite communications. But because of the reason of secret, at present there is not yet an internationally recognized standard of routing algorithm for LEO satellite network, the topic which this paper studies is under such background.First, the satellite constellation topology and the satellite network are introduced in this paper. A comprehensive overview on LEO satellite network routing algorithm research status is given, including the comparisons of several major LEO satellite network routing protocols like DT-DVTR(Discrete Time Dynamic Virtual Topology Routing),AR(Adaptive Routing) and FSA(Finite State Automata Route). The aim of the routing algorithm is introduced by following the satellite network routing characteristics.Secondly, the basic theory of the intelligent agent is studied, including the strong and the weak definition of intelligent agent, the definition, characteristics and classification of the software agent, the concept and calculation model of mobile agent, the architecture and key technology of the mobile agent system and the pertinence between the mobile Agent and the routing algorithm.Then, the paper gives a detailed introduction on the principle and the model of the ant colony algorithm ,the communication between routing agents, the description of routing nodes and the update of the information ,the mobile agent-based routing algorithm. For the traditional ACO has the disadvantages of stagnation behavior and slow convergence, a few modifications are made in satellite network with its own cyclical and regular characteristics.In the end, the traditional and the improved ACO are tested in Iridium satellite constellation. The results show that the improved ACO not only achieves fast convergence in dynamic topology networks, but also can avoid networks congestion and counterpoise the load of the network.