Study And Analysis Of Network Performance Based On Channel Access In Cognitive Radio Networks

Cognitive radio technology can maximize the reuse of spectrum among users and improve the utilization of spectrum resources while maintaining the interference between users within an acceptable level.However,the time-varying characteristics of primary users in cognitive radio networks make the secondary users with high dynamics,which poses higher challenges and requirements for channel access technology.In order to ensure the service quality of secondary users,this paper studies channel access performance of cognitive radio networks from two aspects of spectrum resource scheduling and spectrum access.The performance research on secondary networks based on spectrum resource scheduling needs to propose resource scheduling and allocation schemes to meet the service requirements according to the application scenarios.This paper focuses on the cognitive radio networks in special scenarios that need to provide differentiated services such as emergency rescue,and studies the resource scheduling and allocation problem with service classification.Considering that scheduling schemes with differentiated services usually cause the starvation problem of low priority users due to lack of services for a long time,the purpose of this study is to improve the delivery efficiency of important information on the basis of providing fair allocation schemes.Besides,the high dynamics of secondary users makes the traffic has burst characteristic and network congestion problem in access process.Due to the difficulty of modeling burst traffic,previous studies mostly use the smooth Poisson process to simulate the arrival of the traffic flow.But the traffic in actual communication is bursty in any time scale.Ignoring the burstiness of the traffic will lead to over idealization of performance evaluation model.Therefore,in view of the factors affecting spectrum access in cognitive radio network scenarios,including burst traffic,network congestion,channel fading and finite buffer,this paper studies the performance evaluation problem of secondary users based on access mode.The main achievements are as follows:Firstly,considering that the classification schemes based only on heterogeneous services do not pay attention to the elastic services containing important information in some special scenarios such as the vital data of patients in emergency rescue.The priority of secondary users is divided into two types according to the concern information,and proposes a polling scheduling strategy with reserved channel(Ps-rc strategy).In priority guarantee stage,only the secondary users in high priority queue are served.In polling stage,while serving the secondary users in low priority queue,a part of channels are reserved for high priority users to ensure the service without causing the starvation problem of low priority users.The technologies of channel assembling and spectrum adaptive are adopted and the implementation method of Ps-rc strategy is presented.The continuous time Markov chain is established to depict the resource flow process in secondary system by dividing the trigger events of state transition into four types.All possible destination states,transition rates and transition conditions of the system starting from a general state under four kinds of trigger events are obtained.Then,the analytical expressions of performance metrics such as network capacity,spectrum utilization,blocking probability and forced termination probability are derived.Numerical experiments show that the proposed Ps-rc strategy can guarantee the service quality of the predefined priority users in cognitive radio networks with multiple types of services.Secondly,a more common scheduling scheme that based on the classification of heterogeneous services is considered in this paper.On the basis of using static factors to differentiate elastic services and real-time services,in order to improve the priority of elastic services containing vital data,this study proposes to use dynamic factors to redivide the elastic services that interrupted by primary users.The purpose is to strip the elastic services with high priority.Then,a new reserved queue is set for high priority elastic services and the scheduling priority higher than real-time services is also arranged.That is,on the basis of establishing queues of elastic services and real-time services,a new reserved queue is added for the interrupted elastic services with high priority.On this basis,the technologies of channel assembling and spectrum adaptive are also adopted to propose a priority scheduling strategy with reserved queue(Ps-rq strategy).The algorithms of resource scheduling and channel allocation are presented.Furthermore,the continuous time Markov chain under Ps-rq strategy is established,and the transition events,transition rates and transition conditions that can trigger the change of the system state are derived.Then the performance evaluation metrics of secondary system through solving the steady state distribution are obtained.Numerical experiments show that the proposed Ps-rq strategy can effectively improve the scheduling priority of the interrupted elastic services with high priority on the basis of ensuring fairness.Then,a two-state Markov-modulated Bernoulli process(MMBP-2)is introduced into cognitive radio networks for the first time,and a performance evaluation model of burst traffic with cross-layer consideration based on access mode is established.In data link layer,MMBP-2 is used as the input model to simulate heterogeneous services with different burst degrees,the correlation between them is reflected by Markov chain.In modeling process,this study considers the congestion problem caused by burst traffic and adopts a random early detection(RED)mechanism to control network congestion.In physical layer,the successful reception probability of secondary users on channel is determined by the Nakagami fading,the activity of primary users and the spectrum sensing of secondary users.By establishing the two-dimensional Markov chain,the above affecting factors are integrated into two-dimensional probability space and the state transition matrix is obtained.Besides,an iterative algorithm named forward reduction and backward expansion is constructed to solve the matrix steady state solutions,then obtains the performance metrics including system throughput,average delay,average queue length and total packet loss rate.Numerical experiment results show that the evaluation model established in this paper can ensure the stable system output in serious burst environments.Comparative experiments prove that adopting reasonable congestion control mechanism is helpful for the improvement of secondary system performance.Finally,considering that the RED mechanism used in evaluation model has the parameter sensitivity problem.Thus,this paper focuses on the node congestion,establishes a two-node queuing model with different priorities,and proposes to use the indexes of queue length state and transmission state to feedback network congestion,so as to reduce the dependence of RED on single parameter control.Accordingly,a dual threshold with double index congestion control mechanism,DI-RED,is proposed and the updated threshold position,packet arrival rate and packet dropping probability is derived.The implementation algorithm is also presented.Meanwhile,a discrete time queuing model is established to quantitatively analyze the data transmission performance with DI-RED,and the evaluation metrics of system performance is also obtained.Simulation experiments show that the proposed DI-RED mechanism can not only overcome the parameter sensitivity of RED,but also achieve a compromise among various performance metrics.Based on this,the proposed DI-RED mechanism is applied to the burst traffic evaluation model,and the effectiveness of DI-RED is verified by numerical simulation.