The System Level Simulation Platform Of Ofdma And Research Of Wireless Resource Scheduling Algorithm

Nowadays, all kinds of new wireless technologies emerge in endlessly and it is developing at an unprecedented rate. Along with the rapid development of communication technology, people demand for data rate and quality of service are higher and higher. Traditional voice service can not adapt to the increasing demands of customers. People are keen that future mobile communication systems can support high-speed and low-latency multi-media services, such as television broadcasting, video conferencing, etc. Since orthogonal frequency division multiplexing (OFDM) technology adopts orthogonal carriers and band overlapping design, it has high spectrum efficiency and strong capacity of combating with multi-path interference. Then Orthogonal Frequency Division Multiple Access (OFDMA) technology is widely adopted because of its sub-carrier adaptive ability which can increase the system capacity greatly. It is generally known that spectrum is the most valuable resource in wireless communications. How to improve the spectrum utilization and meet the Quality of Service (QoS) of various multimedia services? These are current researching focuses. Consequently, the system level simulation and efficient radio resource scheduling algorithm based on OFDMA are studied in this dissertation. The main context and work is as follows.First, an OFDMA system level simulation platform built with MATLAB is designed and developed. It includes initialization module, traffic stream module, TCP module, interference calculation module, resource scheduling module, ARQ module as well as performance statistics module. The whole simulation process and the realization of each module including model and parameter selection are detailed. The scheduling algorithms are simulated with four kinds of services:VoIP, Real Time Video, HTTP and FTP. The performance parameters of different scheduling algorithms, such as throughput, packet delay and packet loss rate, are compared. All of these provide powerful support for further study. The platform adopts modularization design which reserves a good interface for further development.Second, a new resource allocation algorithm that can provide QoS guarantee is proposed. Classical radio resource scheduling algorithms, including round robin, fair throughput, maximum carrier to interference (MAX C/I), proportional fair (PF), provide reference standard for new algorithms research. Round robin is the fairest scheduling algorithm. Whatever the channel quality, each user has the same chance of getting service. MAX C/I scheduling algorithm can reach the maximum system throughput. PF algorithm achieves a good balance between throughput and fairness. However, none of the above algorithms takes packet delay into account. A modified largest weighted delay first (M-LWDF) algorithm improves on this point. Compared with PF algorithm, the M-LWDF algorithm guarantees the QoS of real-time traffic with sacrificing throughput. This dissertation presents a resource allocation algorithm which providing assured QoS. The proposed scheme assigns bandwidth according to the priority weights and the system load, and then improves quality of low priority traffic in the condition of ensuring QoS of delay sensitive traffic. Moreover, it provides better performance in terms of fairness. Simulation results demonstrate the effectiveness of this algorithm.