Research On The Congestion Control Of Automatic Self-Allocating System

With the deferent information becomes larger and network's structure becomes more complex, network's loads always overrun network's capacity, and it increases the possibility of the congestion. Because of the characteristics of the cluster, the demands of the congestion control becomes higher and higher. If we could not control the congestion effectively and make it back to the normal state when the congestion comes, it will result in the serious network congestions. And it could make the cluster collapsed. Therefore, it is more important to avoid and control the network's congestion.Firstly, this paper introduces the applications of the cluster, basis conceptions and theories of the congestion control. Then the artery congestion control algorithms of the network system are analyzed and contrasted. Then we discuss the equation of the combination between fuzzy logic and congestion control, and then we compare the fuzzy logic congestion control algorithms to the artery congestion control algorithms.Then we introduce the aspect of request schedule, system framework of ASAS, queue's theory, QoS, work mode of the dispatcher, organization of the mission cache. And we discuss the combination of fuzzy logic and the ASAS.According to the shortage of the artery congestion control algorithms, the indeterminacy of the state information of the network, and the characteristics of the ASAS in buffer cache and queue's theory, the combination of fuzzy logic and congestion control is proposed. We introduce the basis conceptions of the fuzzy reasoning and aggregation, and the method of fuzzy judgment. In order to deal with the fuzzy information, we design the fuzzy reasoning congestion control mechanism based on the Sugeno model. In order to perfect the fuzzy logic congestion control algorithms, a fuzzy logic congestion control algorithm is proposed which based on the buffer performance and full queue time. It scouts the state of the buffer performance and full queue time, and then it judges the congestion state of the queue by fuzzy reasoning, so we could deal with the data flows which are real congested. According to the diverse service type, it provides the Diff-Serv and ensures that diverse service type can get the corresponding service quality. It combines the whole congestion control and the part congestion control. Lastly we use NS-2 stimulation to stimulate the FLCCA, and the results are analyzed and estimated.At the last of this paper, we discuss the combination of FLCCA and the ASAS. We analyze the equation of the combination of FLCCA and the ASAS. Secondly, ASAS's schedule model is proposed which is combined with the FLCCA, and we design a distributed Diff-serv algorithm based on the Diff-serv theory of FLCCA.