Cross-Layer Optimaztion With Queuing Therory In LTE Packet Scheduling System

Long Term Evolution (LTE) system offers remarkably dynamic schedule scheme and provides abundant services.How to maintain the specific Quality of Service (QoS) for the services which are delay-sensitive and jitter-sensitive, especially for VoIP-based voice service? How to guarantee voice capacity without aggravate the overall performance? They are huge challenges emerging in LTE system which is a converged, all IP-based network.In this thesis, the Queuing Theory is applied to analyze the LTE system, and practical and efficient methods are suggested to optimize the capacity of system, which are time-varying linear prioritization and to bring in the traffic differentiation as real-time and non real-time additionally,taking variable prioritization. In the end, simulation results demonstrate that Queuing strategy with time-varying prioritization and traffic differentiation in LTE improves the subscriber's QoS, enlarges the capacity of system. Referring to this output, the potential space for improvement is discussed, and the further step to optimize is given, in hope of providing helping information for system planning and optimization during the network deployment.The first chapter is the introduction of LTE system's development and key technologies.In the second chapter, the queuing theory is brought in, and the successful case in communication is reviewed. In Third chapter, the packet scheduling part in LTE system is studied with application of the queuing theory, furthermore, the cross-layer optimization—the time-varying priority queuing is concluded. The forth chapter is the demonstration of LTE system dynamic simulation platform. The fifth chapter is the comparison and discussion of simulation results.The last part is the summary of research works and enlightenment...