Research Of Iot Congestion Control Based On SDN

As the rapid development of Internet of Things(IoT)technology and sustainable growth of access devices,the business data has been growing explosively in the mean time.which is short of global network information,nor well satisfy the demands of plenty different business type,Thus make the existing congestion control into a bottleneck.Software Defined Networking(SDN)get rid of the inherent restrict of traditional congestion control through decoupling to data forwarding layer and control layer.And make the congestion control mechanism more adaptive and optimized,then solute the demands of transport network of The Internet of Things.Aiming at the congestion problems of nodes and unbalance assignment problems of network resources which exist in traditional network,congestion control method of data forwarding layer and control layer base on SDN was well designed.The main work is as follows:After analyzing lack of fairness and grouping priority in the queue which exist in equity and packet priority aspect of mainstream algorithm like tail discarded algorithm,RED algorithm,SFQ algorithm,one type of Active Queue Management(AQM)that is based on differentiated data flow is designed to the problem of nodes congestion question in Internet of Things framework based on SDN.The condition of next moment can be predicted through the algorithm on the Markov decision,and the influence weight of different flow is calculated out through the Analytic Hierarchy Process in queue.And then build the return function base on switching action and state action,obtain long-term maximum expected reward switch strategy by adapt the successive approximation iterated mode,Finally,verify the scheme through the simulation analysis,the result shows that the method can reduce the congestion rate of data business flow in forwarding nodes and improve the reliability and utilization of network resources.?Pointing at the lower utilization rate and unbalance assignment problems of network resources,one type of virtual mapping algorithm is designed on minimum minimum cost flow is used to implement optimal control of the flow.Centralized control of internet is implemented through analysis of traffic forwarding feature of SDN controller,established a mathematical model of the virtual network mapping,and apply the minimum commodity flow to the traffic path optimization.The Tabu search algorithm is built on improving simplex method and K short circuit method,which is used to solve the planning problems mentioned above.Finally,the simulation analysis of the algorithm is presented.The result shows that this algorithm has better distribution balance and resource utilization in contrast to traditional Shortest Path First(PSF)algorithm.