Resource allocation and cell selection framework for LTE-Unlicensed femtocell networks

Abstract

The availability of enormous bandwidth in 5 GHz WiFi band has advocated its use for the LTE communication. For this, LTE-Unlicensed (LTE-U) technology is standardized in 3GPP Release 12. The use of coexistence mechanism such as Listen-Before-Talk (LBT) can help protecting the quality of service of mobile users operating in the WiFi band by guaranteeing them channel access. Additionally, allocation of wireless resources to each mobile user in the WiFi band should be controlled to facilitate interference-free transmissions to other mobile users in the same band. In this paper, we propose two resource allocation techniques for femtocell networks which efficiently allocate resources such as spectrum and transmit power among competing mobile users in the LTE and WiFi bands. For the LTE band, we propose the use of spreading based spectrum and power allocation, and model the throughput of femtocells using a non-cooperative game theoretic framework. To improve the co-existence performance of femtocells operating in the WiFi band, we propose a Q-leaning based resource allocation technique for femtocell transmit power and duty cycle calculation. In a scenario where multiple base stations are available for user-base station association, considering only the signal strength to make association decision is not an optimal approach, specially when user load and available resources at base stations vary significantly with time. Considering this, we suggest a cell selection scheme that assigns mobile users to femtocells considering the bitrate they are expected to receive in both LTE and WiFi bands. Our resource allocation techniques, when combined with the proposed cell selection scheme, show a minimum improvement of 23.4% in network throughput and 19.6% reduction in energy consumption. This, in turn, also improves the energy efficiency of the network by at least 35%.