Modeling and Analyzing the Coexistence of Licensed-Assisted Access LTE and Wi-Fi

Licensed-assisted access (LAA) is a feature of Long-Term Evolution (LTE) which enables operation in the unlicensed spectrum. LAA guarantees fair coexistence with Wi-Fi by implementing the listen- before-talk (LBT) mechanism. In this paper, we leverage stochastic geometry to characterize key performance metrics of neighboring LAA and Wi-Fi networks. Our analysis is focused on a single unlicensed frequency band, where the locations for coexisting LTE eNodeBs (eNBs) and Wi-Fi access points (APs) are modeled as two independent homogeneous Poisson point processes (PPPs). Based on an analytical modeling of the channel access procedure, we have derived the medium access probability (MAP), the signal-to- interference-plus-noise ratio (SINR) distribution, the density of successful transmissions (DST), and the data rate distribution for both LAA and Wi-Fi. We show that compared to the baseline scenario where Wi-Fi coexists with an additional Wi-Fi network, LAA can improve the DST and data rate performance of Wi-Fi by adopting more sensitive clear channel assessment thresholds and/or larger contention window sizes. Meanwhile, LAA is demonstrated to achieve acceptable data rate performance despite using LBT.