Performance Analysis Of Wireless Opportunistic Scheduling Based On Stochastic Network Calculus

With the extensive application of wireless opportunities scheduling and the need for high-speed data traffic and multimedia, there is an urgent need for a means to guarantee the quality of service of wireless data networks. Therefore, we use two basic tools of the theory of stochastic network calculus (arrival curve and service curve) to depict the stochastic arrival process of the traffic and node service process in the wireless data network. By these two basic tools, we qualitatively obtained delay and backlog in order to achieve the purpose of evaluation and protection of quality of service.In this paper, we firstly state the theory of network calculus and opportunistic scheduling, and then detail the existing three opportunistic scheduling algorithms. Last, for these three scheduling algorithms, we depict service curve of the node scheduler service process using network calculus theory and draw the expression of delay and backlog. In the end, they were analyzed and compared. However, these three algorithms in the design itself do not take into account the quality of security services. In theory, they are not ideal opportunities scheduling algorithms when we consider the study which application of the quality of service to networks. They have their limitations. Numerical analysis section compares the performance of the differences between them and verifies the correctness of the above theory.In this paper, we propose a new stochastic model using statistics bounded burst theory which limits the arrival process of downlink of wireless data network. According to the disadvantages of three above algorithms, modified proportional fair scheduling algorithm is proposed. Traditional proportional fair algorithm only considers the ratio of the average throughput of the channel of user at the any time and channel throughput before time window, but modified proportional fair algorithm also set up a weighting factor(weighting factor presents the priority of user at current link state, the more greater, the lower the priority of the user). The two together determine the scheduling rules. The theoretical analysis and experimental results show that the algorithm has great advantages in the terms of delay and backlog compared with three existing algorithms. And the stochastic model makes the delay and backlog rose to a statistically significant probability to achieve the purpose of protection of the quality of service.