Harmonized Q-Learning for radio resource management in LTE based networks

The efficient management of radio resource is highly imperative so as to meet the vast application requirements in future high speed wireless networks such as Long Term Evolution-Advanced (LTE-A). The current research on applying machine learning algorithms either focuses on packet scheduling in infrastructure network or in cognitive radio in ad-hoc environment. Our study on spectrum usage indicates that there is a lot of room for optimization of spectrum in a multi-operator scenario of LTE systems which covers large customer over a vast geographical area. In this paper, we introduce the concept of Harmonized Q-Learning (HQL) for the radio resource management in LTE based networks that efficiently manage its resource pool dynamically. The multi-operator system is modeled on the game theory based Q-Learning. Our system level simulation of the proposed algorithm shows higher throughput while meeting the real-time resource requirement of each player.