Research On Spectrum Handoff Scheme In Mobile Wireless Cognitive Networks

The shortage of spectrum resources is a bottleneck limiting the sustainable development of mobile wireless communications and service applications.As a revolutionary spectrum sharing technology,cognitive radio has been opened up a new way to alleviate the contradiction between the lack of spectrum resources and the growing demand for wireless service through using the licensed bands opportunistically.The cognitive network must guarantee the quality of the licensed network when accessing the licensed bands,which means that when the primary user reuses the current channel,cognitive user must interrupt the transmission itself and switch to another channel for continuing the communication.This process is called spectrum handoff.Achieving fast and smooth handoff is important to improve the system performance.This thesis is based on a new fusion network architecture,which uses Software Defined Networking(SDN)to divide traditional base stations into two parts:control base stations and data base stations(also called small cell base station).In order to ensure the reliability of system,control base station uses the lower licensed frequency band to transmit the control signal,and data base station uses the licensed band,cognitive band and/or unlicensed band to provide data services.The urban vehicle network is regarded as the main application scenario.We study the spectrum handoff of the downlink in data base station,including the following two aspects:1.In traditional static cognitive network,the spectrum availability mainly depends on whether the licensed users are occupying channels or not.However,in the mobile scenario,the available spectrum resources also changes with the user's location.Therefore,a channel availability model associated with licensed users' activity status and cognitive users'movement is proposed.The status of a cognitive user is divided into three types:1)the cognitive user is located within the coverage of a licensed base station the channel belongs to,and the channel is busy;2)the cognitive user is located within the coverage of a licensed base station the channel belongs to,but the channel is idle;3)the cognitive user is not located within the coverage of a licensed base station the channel belongs to.The latter two types belong to the state that the channel is available for cognitive users.The probability of channel availability can be calculated by analyzing the transition rate of three states,then it can be used as the sorting factor for spectrum handoff mechanism.When a cognitive user needs to perform the spectrum handoff,it selects a channel with the largest probability of availability.It is applied to reactive mechanism,proactive mechanism and hybrid mechanism for simulation.The results show that the hybrid handoff algorithm based on channel availability probability can effectively reduce the number of spectrum handoff,and improve the system throughput.2.The key to achieve smooth handoff also depends on the choice of target switching channels.Therefore,this thesis proposes a switching channel selection strategy based on discontinuous spectrum aggregation,in which,data base station can accesses to cognitive and unlicensed bands simultaneously.For the sake of analysis,the unlicensed system is referred to as WiFi system specifically.Although the interweave mode to access cognitive band can effectively avoid co-frequency interference,the outband power leakage caused by OFDM could not be ignored.Thus firstly cognitive band interference problem was quantified.Secondly,to coexistence with WiFi system,we use the frame-based listen-before-talk mechanism(F-LBT).Through adaptively adjusting the number of idle frames,the F-LBT mechanism can guarantee the fairness between the two systems.Finally,with the assurance of the interference restrictions on licensed system and small cell users,data rate requirements,we build the objective function to maximize the system throughput and minimize the number of handoff times.Then we use the genetic algorithm to derive the handoff scheme for cognitive and unlicensed channels,and the number of idle subframes.We compared the simulation results with the algorithm of maximization throughput,the results show that this scheme can effectively improve the system throughput performance,reduce the handoff times,and guarantee the fairness on resources allocation with WiFi system.