Research On QoS-Based Downlink Resource Scheduling Algorithm In TD-LTE System

The user plane of the Time Division Long Term Evolution(TD-LTE)system is used to transmit data information of users between the base station and the terminal,which plays an important role in the system.The downlink resource scheduling process is in the MAC layer of the user plane protocol stack,which is becoming one of the research hotspots in the communication system because it can improve the transmission performance of different user services.The resource scheduling process allocates limited channel resources according to channel quality conditions and service quality QoS(Quality of Service)information in different scenarios to ensure the transmission performance of user services.However,when multiple types of services exist at the same time,the transmission performance of the user is different due to the difference in transmission characteristics of each service,and needs to be improved.This paper mainly studies the resource scheduling algorithm in MAC layer for service performance balance and improvement in TD-LTE communication system with multi-service coexistence.The main research contents are as follows:(1)A QoS-aware channel resource scheduling algorithm based on Shapley Value method is proposed.Focused on the balance problem of the transmission performance between Real Time(RT)service and Non Real Time(NRT)service in the process of downlink multi-service transmission in TD-LTE system,this paper uses the Shapley value method of cooperative game theory to solve the optimal solution of channel resources occupied by each service according to the transmission data volume of each user service.Then,it performs QoS-aware scheduling based on the optimal solution to control the transmission data of each service,which improves the service transmission performance.Extensive simulation experiments of the system-level simulation platform LTE-Sim show that the proposed algorithm improves the transmission performance of NRT services,such as throughput and fairness.Compared with DPVT-MLWDF(Delay Priority Virtual Token-Modified Largest Weighted Delay First)algorithm,the throughput of NRT services is increased by an average of 38.05% and the fairness is increased by 12.51%,while the number of users changed from 5 to 50.Compared with other algorithms,the performance of non-real-time services is obviously improved,and the performance of realtime services is also improved,which achieves the purpose of improving system performance.(2)A weighted average delay improvement algorithm based on MLWDF algorithm is proposed.Focused on the problem of QoS requirements of RT service in downlink multi-user service transmission,and fully considers the QoS parameters,channel quality parameters and buffer state parameters,which improves the transmission performance of the real-time service in the downlink multi-user service transmission of the TD-LTE system,and ensure the nonreal-time service transmission performance.Extensive simulation experiments of the systemlevel simulation platform LTE-Sim show that the proposed algorithm improves the throughput,fairness and real-time performance of RT service by fully considering the delay of users.Compared with MLWDF algorithm,while the number of users changed from 5 to 50,the average throughput is increased by 36.82% and the packet loss rate is reduced by 64.65% of RT service.The throughput performance of NRT services also outperforms the VT-MLWDF algorithm,which guarantees the transmission performance of this type of service.