Research On Madm Based Scheme Of Joint Resource Scheduling In Heterogeneous Wireless Network

With the rapid development of radio access communication technologies,the overlap of various radio access networks in a certain area makes the typical heterogeneous scenarios.Under this circumstance,it is difficult to satisfy every service request using single radio access technology,along with rising needs of differential service.The joint resource scheduling(JRS)is a promising solution for the cooperative evolution of heterogeneous wireless networks(HWN),A thorough JRS is established due to the various management mechanisms and resource implementations of different radio access technologies.This thesis proposes several solutions for key problems of the JRS.The thesis starts with the discussion of traditional management frame for combining HWN.On the basis of these works,it then proposes the implementation model for the following JRS system.This model is divided into user panel and control panel.The concrete content of different modules and the main algorithms executed on different modules will be introduced to explain how this system works.On the user side,a network selection strategy based on Multiple Attribute Decision Making(MADM)is proposed.After the detection of all accessible networks and their network parameters,the strategy lists all the candidate selection schemes.The decision of whether using JRS depends on the classification of the service which can whether be a large bandwidth demand service or a small bandwidth demand service.For large bandwidth demand service,the combining parameters for different candidate selection schemes will be calculated.After integrating deciding matrices for different service and pre-processing these matrices with weights corresponding to certain service request,the final selection scheme holding highest utilization through GRA algorithm in MADM will be chosen to access.Aiming at the best QoE,this thesis then defines the concept of satisfaction as one vital evaluation for QoE got by users.A streaming strategy based on queue theory is adopted on the network side for the integration of simulation.The definition of load and load balance index are made to evaluate the load balance situation on the network side.Simulation is made between the JRS selection strategy and the single network selection strategy based on vertical handover(VHO).The result of simulation shows that JRS can improve the satisfaction of users especially for those asking for large bandwidth demand service compared to VHO.On account of the proper distribution of the radio resource,the satisfaction of users who askfor small bandwidth service also rises slightly.Besides,the load balance situation between radio access networks improves while using JRS.On the network side,the following part of this thesis optimizes the former streaming strategy with a finer granularity model.The constraints are designed according to the demanding bandwidth of users and the using condition of network channels or resource blocks.The optimization goal is to minimize the load balance of the system.Lagrange relaxation is applied to simplify this optimization problem,and the solution can be figured out through the iteration of subgradient method and the monotony of the convex function.Furthermore,the congestion control module and the Mobile resource scheduling optimization(MRSO)module which aim at the flexible management are designed.The congestion control module defines two more constraint steps to limit the distribution of radio resource to avoid the congestion.There are totally three steps for the distribution:(a)Normal distribution;(b)Dynamically auto-adaptive adjust the distributing resource by producing a buffer factor as the preliminary constraint measures for the FSC step;(c)Stop distributing resource when the using rate breaks the threshold for the SSC step.The MRSO module can decrease the number of unnecessary handover caused by the MADM decision for mobile users.The simulation result shows that JRS can not only keep the balance of load,but also ensure the smooth of communication link.Users can get higher satisfaction when more users are involved in this system,and the average number of handover among mobile users can be highly decreased.Aiming at the classification of multiple attributions service on the user side and the guarantee of load balance on the network side,this thesis respectively proposes the JRS strategy and advanced radio resource distribution mechanism.All the work has been proven to improve the QoE of users with different request.