Quality of service aware scheduling in mixed traffic wireless networks

Abstract

Considering a traffic mix in a wireless network, the scheduler at the base station needs to be aware of the type of user packets and the state of each user buffer to satisfy the requirements of users. In such networks, the requirements vary in terms of latency, through put, and reliability. Thus, a trade-off is required between the prior performance metrics to enhance the overall network performance. In this thesis, a Quality of Service Aware Scheduler (QAS) is proposed with tuning parameters to achieve balanced Quality of Service (QoS) delivery. Moreover, a moderate fairness is imposed among full buffer users that are assumed to have infinite amount of data in packet buffer. An open-source modeling tool is usedto solve the Resource Blocks (RBs) optimization problem of QAS. System Level (SL) simulations are performed to investigate the performance of the proposed scheduler. In addition, QAS performance is compared to the standard scheduling strategies, namely Round Robin (RR) and Best Channel Quality Indicator (CQI) under different network loads and users positioning methods.