Research On Routing Algorithms Based On Intelligent Optimization Algorithm For Aanet

With the rapid development of aerospace and space exploration,the influence of spatial information is becoming stronger.From a certain extent,the mastery of space information can affect the development of a country's comprehensive national strength and economy.Space information network is mainly composed of deep space network,terrestrial space network and terrestrial network.Terrestrial space network consists of different kinds of low-altitude aircrafts,including planes,helicopters,fighter jets and unmanned aerial vehicles(UAV).It's also known as aeronautical ad hoc network(AANET).Various aircraft nodes in AANET have real-time mobility and velocity of them is very high,which lead to network topology changing frequently.Therefore,the routing technology in AANET is facing a lot of problems,including short valid routing time,frequent communication routings disconnection,long data packet transmission delay,etc.So it's necessary to use routing algorithms that can adapt to the characteristics of AANET,which requests fast convergence,short route setup time,low average routing hops and fast path optimization.This paper mainly studies routing algorithms based on intelligent optimization algorithms among aircrafts in AANET.First,considering AANET's characteristics that nodes have high mobility and topology changes frequently,ant colony optimization algorithm,particle swarm optimization algorithm and genetic algorithm are applied to the network to establish intelligent optimization routing algorithms.These three intelligent optimization routing algorithms are simulated in MATLAB and their performances are compared with AODV protocol.Simulation results show that in successful link rate and average link hop,ant colony optimization routing algorithm performs best,particle swarm optimization routing algorithm and genetic optimization routing algorithm are also good,AODV protocol is the worst.Then,considering special scenarios when the cluster head node is no longer in the center position of network area,or cluster head node get failed or destroyed and is absent from network,cluster head node selection and updating algorithm based on graphs theory is designed.And this cluster head node selection algorithm is modified using probabilistic graphical model in order to get results more close to AANET scenarios,cluster head node selection and updating algorithm based on probabilistic graphical model is proposed.Analysis shows that the cluster head node selection and updating algorithm base on probabilistic graphical model although has slightlyhigher algorithm complexity than the method based on graphs theory,the average link hop is less between cluster head node and other nodes.Finally,discussing the situation when there is a new node joining network,using the theory of probabilistic graphical model to consider the relationship between new node and its adjacent nodes,the best access node is chosen for new node.Simulation results show that the access node chosen with this method sometimes is different from the one chosen with principle of closest distance.