LEO(Low Earth Orbit) satellite network is more suitable for multimedia communication than GEO(Geosynchronous
Earth Orbit) and MEO(Medium Earth Orbit) satellite network because of short propagation delay. Routing computation
in communication network with several constraint factors has been proved to be a NP-C problem. Ant colony algorithm
as a bionics algorithm has good performances on solving NP-C problem. Existing research about ant colony algorithm
mainly concentrates on Ad hoc network, and application in satellite network is not completely developed. According to
the characteristics of LEO satellite network, this paper presents a new routing method based on ant colony algorithm
with forecast strategy and elitist strategy. Generally each satellite establishes two intra-ISLs with forward and backward
neighbors in the same orbit, and two inter-ISLs with left and right neighbors in the adjacent orbits. Thus the topology of
satellite network consists of many grids. According to the specific topology, forecast strategy is introduced to reduce
redundant information. On the other hand, pheromone is updated continuously to ensure elitist path chosen. Assuming a
reference satellite network model is given, the simulation results show that the improved algorithm converges more
quickly than basic ant colony algorithm.
|