Packet Scheduling Algorithm On MAC Layer For Long Term Evolution System

ABSTRACT:With the emergence of broadband wireless access, cellular mobile communications technology and broadband wireless access technology have shown a trend of mutual integration, the services demand of mobilization of broadband access and broadband of mobile communication increasing strongly, the user requirements for higher rate more urgent, hence,3GPP developed Long Term Evolution in2004. As spectrum resources of the wireless communication system are always limited, so research the issues in what way to use the limited radio resources more rational and effective, support as many users while ensuring quality of service of a variety of business has been an important significance, in order to achieve higher data transfer rates and reliable quality of service (QoS) guarantee, packet scheduling as a an important part of the LTE can be rational allocate and use the limited spectrum resources, and improve system performance.Firstly, this paper analyzes the overall system architecture features of the LTE system, Summarizes the radio resource management’s role in LTE, detailed study packet scheduling principles in radio resource management.Secondly, in-depth analysis key issues of packet scheduling algorithm of the MAC layer facing, in-depth study the improved M-LWDF scheduling algorithm, for video streaming services, simulate and compare the performance of the traditional packet scheduling algorithms with M-LWDF scheduling algorithm, simulation results show that the M-LWDF algorithm can meet the needs of users streaming video services because of its high throughput, better fairness and lower packet loss rate.Finally, this paper analyzes the LTE FDD downlink time-frequency structure and establish a joint time-frequency domain scheduling model, and propose PF-PF and PF-M-LWDF two kinds of joint time-frequency domain scheduling algorithms, for video streaming services, simulate and compare the performance of the joint time-frequency domain scheduling algorithms with PF and M-LWDF scheduling algorithms, the results show that the overall performance of the joint time-frequency domain scheduling algorithms has been further improved compared to only the frequency domain scheduling algorithm for its further use multi-user diversity gain.