Adaptive Self-Deferral for Carrier Aggregation of LTE-LAA With RF Power Leakage in Unlicensed Spectrum

The licensed assisted access (LAA) is the new feature of 3GPP long-term evolution (LTE) that uses unlicensed spectrum as an additional bandwidth to meet ever-increasing mobile traffic demands. For fair coexistence with other incumbent systems such as Wi-Fi, LAA specifies the listen-before-talk (LBT) mechanism for channel access. LBT of LAA is also designed to support multi-carrier operation, which is the key to capacity increase, but inherent RF power leakage to adjacent carriers ruins the multi-carrier LBT and deteriorates aggregation capacity considerably. Self-deferral is a solution to solve this problem by aligning carriers' transmission times via transmission deferring of each carrier after backoff, for which the key to success is to find how long self-deferral must be. In this paper, we propose an algorithm to adjust a self-deferral period of the multi-carrier LBT adaptively to carrier loads for enhanced carrier aggregation capacity under RF power leakage. We formulate the target problem as an optimization problem whose objective is to maximize the expected number of aggregated carriers derived as a function of the self-deferral period and carrier loads. Then, we derive the optimum of the aggregation capacity maximization problem for the case of homogeneous interference patterns between carriers in a closed form. Due to the computational complexity of finding the global optimum for the general case of realistic heterogeneous interference patterns between carriers, we develop a suboptimal algorithm to configure the self-deferral period. Through extensive simulation, we demonstrate that the proper configuration of the self-deferral period is of importance and the proposed algorithm outperforms various LBT options in a wide range of network configuration by up to 72% in a single-spot scenario and 47% in 3GPP's indoor deployment scenario, while meeting fair coexistence with the Wi-Fi systems.