Research On Resource Scheduling Algorithm And Performance Evaluation In LTE-A Relay-enhanced Cellular System

LTE-A system demands more on data rate, spectral efficiency and delay than LTE system. In order to achieve the performance requirements of LTE-A system, 3GPP proposed several enhanced technologies including multi-user MIMO, multi-point cooperation, carrier aggregation and relay. Relay technology which can expand the network coverage, effectively eliminate communication blind zone, improve system throughput and reduce the cost of network construction, has become a hot topic in LTE-A system. The application of relay technology introduces a new data link, increases the resource consumption of the system, and makes the resource scheduling and allocation more complicate, therefore rational resource scheduling should be researched in order to improve system performance.Firstly, this dissertation describes the development of relay and overviews relay-enhanced cellular system. Then according to some related 3GPP protocols and proposals, LTE-A TDD downlink relay system simulation platform is built up, including cell topology modeling, radio propagation channel modeling, antenna modeling, user generation and user access point selecting, receiving uplink feedback, resource scheduling module, link quality module and link performance module, which establishes the foundation for the following research of resource scheduling algorithms.Secondly, classical resource scheduling algorithms for traditional cellular system are researched, including Round Robin(RR), Max C/I and Proportional Fair(PF). The three scheduling algorithms are implemented in the established relay system simulation platform. The simulation results show that PF can obtain the compromise of system throughput and fairness, and finally achieve relatively good performance.Thirdly, resource scheduling algorithms for relay-enhanced cellular system are studied emphatically. A common resource scheduling algorithm, Adaptive Distributed Proportional Fair(ADPF), is studied in relay-enhanced cellular system. However ADPF can’t guarantee a good balance between throughput of backhaul link and throughput of access link, and has poor fairness. Therefore Two-hop Balanced Distributed scheduling algorithm based on Relay Cache(TBDRC) is proposed in this dissertation. In scheduling for the first hop, eNB adaptively adjusts the number of RBs occupied by backhaul link and direct link based on the cache information of relay, in order to improve resource utilization, fairness among the use of resource blocks and transmission throughput. In scheduling for the second hop, relay allocates RBs to each relay UE based on the size of each relay UE’ TB, in order to avoid a waste of RB resource, reduce interference between access link and direct link, and improve fairness among all users. Then ADPF and TBDRC are implemented on the established relay system simulation platform. By comparing simulation results of TBDRC with those of PF and ADPF, we can observe that the proposed TBDRC can significantly improve resource utilization and fairness among all users, increase system throughput to some extent, and achieve a good trade-off between system throughput and fairness among all users by means of balancing the throughput of backhaul and access link.Finally, the impact of the number of relay on system performance is evaluated in many aspects, based on the established relay system simulation platform and TBDRC.