Modeling and Analysis on Hybrid-Duplex Networks with Dynamic Traffic

In this paper, we focus on the interference analysis in stochastic cellular networks with hybrid-duplex base stations (BSs). Most previous works on full-duplex networks are based on the assumption that all BSs are full-loaded, which overestimates the interference. By contrast, we consider a more practical scenario where BSs may be active in four states, i.e., silent, halfduplex downlink, half-duplex uplink and full-duplex. Dynamic traffic at the BS is considered and the corresponding queuing model is established to depict the variation of BS states. The proportions of different states of BSs are further derived, based on which the inter-and intra-cell interferences are analyzed. Instead of averaging intra-cell interference within the whole cell, yielding a rough interference analysis, we propose the concept of strong interference region which classifies the intracell interferers into strong and weak interferers. It presents that the downlink user experiences obviously different outage probabilities when the intra-cell interferer is located in or out of the strong interference region. Simulation results validate the superiority of the proposed analytical approach in increasing the precision of interference analysis.