Exp-bet packet scheduling algorithm for multiple traffic services in long term evolution (LTE) system / Ku Siti Syahidah Ku Mohd Noh

• Main Author: Ku Mohd Noh, Ku Siti Syahidah
• Format: Thesis
• Language: English
• Published: 2016
• Online Access: https://ir.uitm.edu.my/id/eprint/17664/2/TM_KU%20SITI%20SYAHIDAH%20KU%20MOHD%20NOH%20EE%2016_5.pdf; https://ir.uitm.edu.my/id/eprint/17664/

Most mobile devices today support multimedia services and legacy mobile services such as voice, short message service (SMS) and multimedia messaging service (MMS). Real-time services like video streaming and multimedia gaming are given higher priority when scheduled as compared to non-real-time services like web browsing and sending emails. However, both real-time and non-real-time services have equal demand which means that the urgency for such services to be scheduled cannot be neglected. Due to the different traffic characteristics and Quality of service (QoS) requirements of real-time and non-real-time services, the provision of the QoS requirements has become a challenge. It is found that packet scheduler can be used to fulfill the required QoS by services. An efficient packet scheduler in allocating resources to active flows is very important since it will affect an individual user's throughput and fairness. Thus, it is important to choose an optimal packet scheduler scheme to be used in order to schedule this active flow. In this thesis, the comparison of proposed algorithm namely the Exponential Blind Equal Throughput (EXP-BET) was made with the Exponential Proportional Fairness (EXP-PF) and Frame Level Scheduler (FLS) algorithms. The aim of this thesis is to determine the suitable packet scheduling algorithm in order to schedule the multiple services and ensure fairness regardless of the user's location in a cell. The simulation process has been done using the LTE-Sim simulator and the performance of these packet scheduling algorithms were evaluated based on the performance metric of fairness index, throughput, packet loss rate and delay for VoIP, video and best effort flows. From the simulation results, it was observed that EXP-BET delivers higher fairness and throughput and lower PLR and delay for real-time services. Instead, EXP-BET shows 17.72% improvement than FLS and 7.52% from EXP-PF in term of fairness index for the non-real-time services.