The Capacity And Delay In MANETs Based On Multi-Source And Multi-Destination

Mobile ad-hoc networks(MANETs) is a temporary mobile network whichcomposed by many mobile nodes which do not have wired infrastructure. Because thereasons of power and frequency spectrum, communciation links will effect the others’communciation links, and then reduce the capacity of the network, so we shouldcontrol this interfence within certain limits if we want to gurantee the quality of thenetwork. Base on these theory, researching capacity, delay and correlation theory ofnetworks in differnet scenarios is very important and necessary for improving theactual communication services.The paper is based on a paper which reserached capacity and delay of one-sourceand one-destination communication network by transminssion-group and f-castalgorithm. Firstly, because our research capacity and delay of multi-source andmulti-destination communication network, we meet many difference, for examplecommunication conflict, the priority of caching and so on. In order to solving theseproblems we desiged a advanced transminssion-group and a advanced f-castalgorithm. Then by computing the number of sources and their delays, we make up amarkov model to figure out the value of expectation of capacity and delay of thenetwork. Secondly, we improved the advanced f-cast algorithm at forwarding rate ofmessage and priority of caching. Finally, we introduced a virtul sensor network in thenetwork, using a decision model to guide the behavior of destination, making itreceiving message better and improving the performance of the network.The results of capacity and delay of multi-source and multi-destinationcommunication network which composed by advanced transminssion-group andadvanced f-cast algorithm are (1n)and (n), and they are conform to the resultsof simulation generally. What is more, the improvement of advanced f-cast algorithmand the adaptive shifting algorithm are make a very good showing in the simulation.In general, our model, algorithm and process of deduce are right and efficient.