Clustering Based Packet Scheduling Adaptive to the Network Load in LTE-Advanced Networks

This paper investigates authors’ previously proposed novel clustering based packet scheduling algorithm for the downlink transmission of LTE-Advanced networks, under variable network conditions. Numerous simulations are run to investigate the performance and validity of this algorithm under different network scenarios such as equal number of real time and non-real time users, real time users more than non-real time users and vice versa. Under each scenario, the total number of users is also varied to validate the algorithm both for different network scenarios and for the variable overall network load. The key performance indicators are average delay, the delay viability and packet drop rate of real time users, minimum throughput of non-real time users, and system throughput and fairness among users. The simulation results show that the algorithm maintains the service level and system level performance under each network scenario and variable network load.