Research On Key Technology Of Resource Management For Indoor Heterogeneous Networks

In order to meet the demand for user's high-speed communication from anywhere at anytime,heterogeneous integration of existing networks in indoor environment has become a research hotpot of wireless technology,particularly overlapping by low power network(eg.WLAN)assisted high power network(eg.LTE)has already become the mainstream networking mode for indoor scenario.However,according to the pre-existing investigations on the topic of heterogeneous network deployment,the situation of system users with low coverage,namely the low ratio between admissible users and total users in the target scenario caused by the none consideration of user's location distribution,has always been ignored.Furthermore,due to the user's low speed,the intensity obstacles and etc.in indoor scenario,the network selection results during the handover process for users in heterogeneous network could be affected by instantaneous parameter volatility,resulting as the misjudgments in network selections.Moreover,service disruption during the process of handover would lead to the packet loss.To realize the indoor seamless coverage by heterogeneous networks,above problems must be addressed urgently.Therefore,this dissertation develops new network resource management technology for indoor heterogeneous networks from three aspects of low power WLAN network deployment,parameter preprocessing for network selections and handover executive decision-making to expand coverage of indoor heterogeneous and decrease misjudgment rate of network selections,as well as reduce the packet loss in the process of handover.Firstly,network characteristics of LTE and WLAN have been studied,as well as the existing commonly used network selection algorithms.Meanwhile,as the basic of execution opportunity decision-making,vertical handover procedures for LTE/WLAN heterogeneous networks have also been dissected.These theoretical analyses are very important for later research and simulations.Secondly,a novel user distribution-based AP deployment algorithm has been proposed for indoor environment,to cope with the low user coverage rate caused by uneven distributions of users' locations.Taking distribution of users' positions as prior knowledge,the proposed algorithm transforms it into position coordinats and the corresponding probability.Then AP deployment is transformed into C-mean clustering problem,and artificial fish algorithm is introduced to guarantee the convergence of the algorithm.The simulation results show that the proposed algorithm can improve the performances on the average number of admissible users,the user coverage and the degree of the load balance,especially when users' positions are in discrete distributions.Thirdly,to cope with the influence on network selection algorithm caused by the volatility of decisive parameters for low-speed users in indoor scenario,this dissertation has proposed a parameters preprocessing algorithm based on Hidden Markov Model for vertical handovers.The algorithm models the relationship between distributions of the parameters and the transfer of access characteristics by Gaussian Hidden Markov Model.The instantaneous decisive parameters would be transformed into normalized results which are associated with access probabilities,and then be used as inputs for heterogeneous network selection.Simulation results show that the proposed parameter preprocessing algorithm can effectively reduce the misjudgment rate for heterogeneous network selections,when both of the user's speed and the frequency of the mobile status changes is low in the indoor environment.Finally,in order to guarantee the packet loss requirement of Vo IP and packet data service during srvice disruption time in vertical handover,a business-oriented vertical handover execution opportunity decision algorithm has been proposed in this dissertation.Combined with the characteristics of dual-user Vo IP business and packet data business,the proposed algorithm is committed to reducing packet loss caused by service interruptions in the process of vertical handover,via starting L2 layer handover when the channel is idle.Simulation results show that the proposed algorithm would reduce packet loss in the process of vertical handover,while low handover delay is ensured.