Performance bounds of different channel access priority classes in future Licensed Assisted Access (LAA) LTE networks

The growing demand on data to enhance the user experience has motivated research toward increasing the efficiency of spectrum utilization by extending cellular technology such as LTE toward the unlicensed bands. Several studies have been done regarding the fair coexistence of LTE with legacy technologies such as WiFi in the unlicensed bands. Most of these studies show that when LTE adopts appropriate technology such as LAA, it can be a good neighbor to WiFi. However, few studies have targeted the impact of coexistence of several LAA networks on the performance of each other in the unlicensed bands. Hence, within this scope, this is the first work that uses stochastic geometry to analyze the effect of channel access priorities of four traffic classes defined recently by 3GPP release 14, on the performance of the network and the degree of efficiency that can be achieved. In particular, we consider load based equipment (LBE) based MAC protocol where we use stochastic geometry to model and analyze the coexistence of four channel access priority classes of LTE-LAA networks with persistent downlink transmissions as defined by 3GPP. Mainly, we focus on a single unlicensed frequency band transmissions. The locations of LAA eNodeBs (eNBs) for the four classes are modeled as four independent homogeneous Poisson point processes. Furthermore, we derive analytical closed form expressions for a set of performance metrics which are the medium access probability (MAP), signal-to-interference-plus-noise ratio (SINR) coverage probability, and rate coverage probability. Our analysis demonstrates the impact of channel access priorities used by different classes on their network performance.