A system level simulation study of WiMAX

Abstract

In this thesis, we implement a WiMAX system level simulator compliant with the evaluation methodology document published by the IEEE 802.16m Task Group. We study the PHY abstraction of polar codes and integrate polar codes into the simulator. We compare the system level performances of polar code and convolutional turbo code (CTC) and observe that CTC outperforms polar code. On the simulator, we study the downlink (DL) performance of WiMAX under various configurations such as scheduling methods, subchannelization methods, and frequency reuse models. We study there types of scheduling methods, namely round robin (RR) scheduling, proportional fair (PF) scheduling, and maximum sum rate (MSR) scheduling. We observe that MSR scheduling has the best throughput performance but does not support the users far from the base station. We study three frequency reuse models, namely 1×3×1, 1×3×3, and 3×3×1. We observe that 1 ×3×1 reuse model has the best throughput performance and maximum spectral efficiency is obtained in 1 × 3 × 3 reuse model. We study two subchannelization methods, namely PUSC and band AMC. We observe that in low mobility cases, band AMC outperforms PUSC and in high mobility cases, PUSC is better than band AMC.