Effect of Idle Mode Cells on the Ultra-Dense Dynamic TDD Networks

To satisfy the growing capacity demand of explosive traffic, ultra-dense network (UDN) is proposed as a key technology, where small cells are densified to fully exploit the spatial spectrum reuse gain. The decreasing coverage area of a small cell also leads to the obviously increasing traffic asymmetry in UL and DL within the same cell and among different cells. To adapt to the dynamic and asymmetric traffic within UDN, dynamic time-division duplex (DTDD) can be seen as a promising scheme due to its capability in the flexible adjustment of ratio of UL to DL subframes. In this paper, we propose a load-aware analytical framework to accurately characterize the network performance of a DTDD UDN with the emphasis on the effect of idle mode cells. With a practical multi-slope path loss model, we first derive the void probability of a random SAP and obtain the coverage probability and area spectral efficiency (ASE). Then we evaluate how the SAP void probability, UL/DL configuration and network density affect the network performance. Our numerical results show that the idle mode cells have significant effect on the network performance, which alters the variation tendency of coverage probability and ASE with the increasing network density.