Research On Packet Scheduling For The Downlink In LTE System

With the rapid growth of mobile communications, the demands of packet data service are augmenting daily. People expect that future mobile communication systems can provide valid packet data, higher transmission rates and lower packet delay. In order to increase the competitiveness and keep in step with the development trend of mobile communications, a Long Term Evolution (LTE) of the 3rd Generation Partnership Project (3GPP) radio technology was started. Effective packet data scheduling is required to improve the system performance, for LTE is a packet switch system and radio channel resources are limited. Packet data scheduling influences the system performance of LTE, for it can let user equipments make good use of the common channel. So packet scheduling technology is one of hotspots and key technologies in LTE.Channel Quality Indicator (CQI) is the most important parameter when scheduling, which indicates the radio channel environment of user equipments. Scheduler would not work well for radio channel can cause CQI parameter go wrong during data transmission between User Equipment (UE) and evolved Node B (eNode B). So a modified CQI based packet scheduling algorithm for LTE is proposed in this dissertation in order to let packet scheduling make good use of the characteristic of radio channel. This algorithm is based on the exponent smoothing forecast model and secondary moving average forecast model, combining with the priori parameters. The result of simulation indicates that it can improve the throughput when scheduling.Quality of Service (QoS) Class Identifier (QCI) priority is defined for services differentiation in LTE as a new parameter. In order to make use of the QCI priority and differentiate the classes of service, a MFQ-M-LWDF (Modified Factor Q for Modified Largest Weighted Delay First) algorithm is proposed in this dissertation, which is based on the M-LWDF (Modified Largest Weighted Delay First) algorithm and the QCI priority. MFQ-M-LWDF algorithm builds relationship between QCI priority and the services which UEs apply in the packet scheduling judgment. It contributes to achieve joint optimization of different UEs and different services which user equipments apply for. The result of simulation indicates that it can improve the effective throughput and guarantee packet delay when scheduling.Transmission delay and packet loss ratio are sensitive factors to real-time traffic in LTE. In order to guarantees the QoS performance and promote the user experience, an overtime and overstock based priority strategy of packet scheduling is proposed in this dissertation. It combined two kinds of monitor entities, queue selection mechanism and Proportional Fair (PF) algorithm to complete packet scheduling. Hereinto, the two monitor entities are queue time monitor entity and queue length monitor entity, which monitor the packet delay and the variation trend of queue length status. It provides much more scheduling opportunity for the queue which is overtime or overstock when packet scheduling, in order to improve the QoS performance. The result of simulation indicates that it can reduce the packet delay and packet loss ratio and improve the throughput when scheduling.