Markov Prediction Based Dynamic Handover In Flexible Networks

To meet the requirements of higher capacity and better coverage,extremely high communication reliability and very low latency,and massive machine-type communication,new architectures such as Heterogeneous Network,Ultra Dense Networks and Software Defined Networks(SDN)are becoming candidate architectures for the next generation of wireless networks.Ultra-dense deployment of heterogeneous cellular networks improves system capacity by increasing the number of low power nodes,largely easing the demand pressure from the explosive growth of the number of wireless communications devices and the traffic needed,but the dense deployment of small cell leads to frequent switching and ping-pong effects.It causes the waste of channel resources and the decline of Quality of Experience.Therefore,in order to solve the problem of the communication and computational problems of massive machine-type communication in an ultra-dense cellular network,a Markov prediction based dynamic handover in flexible networks is proposed,the main research contents and innovation are as follows:1.In this thesis,the ultra-dense deployment of heterogeneous networks is a part of the integrated central SDN controller entity of the access network,and virtualize the wireless access point into a virtual node,using a semi-structured distribution.This thesis presents a SDN Heterogeneous network architecture,and global view of SDN controller is used to obtain the network parameters and terminal parameters in real time,and the Network Information Processing module extracts the network parameters and state information of Radio Access Technologies,and processes it,executes the handover prediction algorithm.Finally,the network resources can be flexibly controlled by selecting the appropriate network for user and executing the handover decision.2.In order to improve the handover performance of massive machine-type communication,a handover scheme based on Markov prediction is proposed by combining prediction with cell handover.Firstly,a Markov model is established to compute the transfer probability between the cells by using the non-homogeneous discrete time Markov chain,and forecast the target cell.Users can send certification in advance to the prediction cell,request and forward data,when the user received the signal strength of the current cell weakened,send connection requests to the prediction cell directly.The results of computer simulation show that the handover design based on Markov prediction is easy to predict the next access network of the users on the basis of ensuring thehandover performance in the low speed(2m/s)mobile situation.3.Context-aware computing is mainly used to improve system performance and user experience.This thesis presents an integrated context-aware Markov predictive handover system model and prediction scheme to further optimizes the algorithm.Add context information,that is,the load status of each cell around the user,and the aim is to further improve the precision and intelligence of the handover algorithm.The results of computer simulation show that the optimized algorithm can weigh the signal quality and the transmission load of each cell,and dynamically choose the best handover cell for the user.