Energy-efficient sensing mechanism for licensed-assisted access under non-saturated traffic condition

This paper investigates the energy-efficiency of licensed-assisted access (LAA) which enable long-term evolution systems to operate in unlicensed spectrum. Since Wi-Fi with 802.11 n/ac is a typical radio access technology in the unlicensed spectrum, we consider a scenario that the LAA small-cell base stations (SCBSs) are deployed together with Wi-Fi access points. We propose three different listen-before-talk (LBT) schemes according to the channel sensing mechanisms for LAA SCBC, and establish Markov chain models for each scheme under non-saturated traffic condition. Using these analytical models, the energy-efficiencies of LAA SCBSs that perform the different LBT schemes are investigated. We also propose the algorithm to obtain the optimal contention window size of LAA SCBSs by which total energy-efficiency of networks is maximized while satisfying the required energy-efficiency of both LAA and Wi-Fi networks. Numerical results show that the LBT scheme which has an efficient channel sensing mechanism outperforms the other LBT schemes for the energy-efficiency perspective.