Comparative Empirical Analysis of Downlink Scheduling Algorithms with Emphasis on the User Satisfaction – Based Scheduler

Abstract

The increasing demand for mobile broadband services and multimedia applications has intensified the need for efficient radio resource management in Long Term Evolution (LTE) networks. Downlink packet scheduling plays a critical role in allocating limited radio resources among multiple users while satisfying diverse Quality of Service (QoS) requirements. However, existing scheduling algorithms exhibit varying performance under heterogeneous traffic conditions, making it difficult for a single scheduler to satisfy diverse user requirements simultaneously. This study presents a comparative empirical analysis of selected LTE downlink scheduling algorithms with emphasis on providing insights towards the development of a user satisfaction-based scheduler. Four representative algorithms were evaluated: Modified Exponential Rule (MEXP) and Enhanced Fair Earliest Due Date First (EF-EDDF), which represent unified scheduling approaches, as well as Downlink Scheduling Algorithm with Imperfect Channel Quality Indicator (DSA) and Enhanced Exponential Rule (EnEXP), which represent differentiated scheduling approaches. Simulations were conducted using the LTE-Sim platform under a single-cell LTE environment with heterogeneous traffic consisting of Video, Voice over Internet Protocol (VoIP), and Constant Bit Rate (CBR) services. Performance evaluation was based on packet delay, packet loss ratio, throughput, and Jain’s fairness index. Results show that differentiated schedulers perform better for delay-sensitive multimedia traffic, while unified schedulers exhibit superior performance for non-real-time traffic. The findings provide useful insights towards the development of a user satisfaction-based scheduler capable of improving fairness, QoS delivery, and overall user experience in LTE networks.