An enhanced packet scheduling algorithm for downlink cognitive long term evolution-advanced

Abstract

The demand for Real Time (RT) and Non-Real Time (NRT) multimedia contents on mobile devices are increasing at a high pace as internet are becoming easier to access. These demands are mostly fulfilled by Long Term Evolution-Advanced (LTE-A) mobile communication standard. However, these exponential demands also will cause LTE-A to face a challenge to provide satisfactory Quality of Service (QoS) mobile users in the near future. This is due to spectrum scarcity because LTE-A operates at a fixed spectrum that will cause the spectrum soon to be congested on most frequency bands. Therefore, a study on increasing spectrum availability and efficiency is needed. Recently, Cognitive Radio (CR) technology is highly researched as a promising technology to overcome spectrum scarcity. Spectrum availability and efficiency of the network can be improved via implementation of CR into LTE-A. Furthermore, addition of Packet Scheduling (PS) in the downlink Cognitive LTE-A will further satisfy the required QoS of mobile users. However, the study on the stated implementation is very limited. Thus, this research aims to investigate packet scheduling performance in the downlink Cognitive LTE-A. The goal of this thesis is to provide satisfactory QoS which are system throughput, fairness and packet loss ratio whilst increasing the system capacity of RT and NRT multimedia contents users in the downlink Cognitive LTE-A. This thesis proposed an enhanced packet scheduling algorithm in the attempt to achieve the stated goal and using simulation of the enhanced algorithm in Cognitive LTE-A. The performance of the proposed packet scheduling algorithm is compared with well-known packet scheduling algorithms based on several performance metrics in a simulation environment limited to a single hexagonal cell to reduce complexity of simulation. Based on the simulation results, it is shown that the proposed packet scheduling algorithm supports 185%, 200% and 131% more Cognitive LTE-A users compared to SC M-LWDF in first, second and third scenario while supporting 9.6% and 15.4% more Cognitive LTE-A users at the required RT PLR compared to MM-LWDF at the expense of degradation of the NRT PLR which is considered acceptable due to the vacancy of radio spectrums when compared to the well-known packet scheduling algorithms.