Opportunistic Interference Mitigation for D-TDD in Ultra-Dense Networks

Dynamic time-division duplexing (D-TDD) is a promising solution to address the fast and varying traffic demands in ultra-dense networks (UDNs). However, in addition to conventional interference, the cross-link interference (CI) resulted by asymmetric uplink (UL) and downlink (DL) transmission in D-TDD operation may further degrade system performance. In this paper, we propose a novel opportunistic interference mitigation (OIM) framework to enhance D-TDD in UDNs by exploiting multi-user diversity in conjunction with interference-aware transmission. In particular, we first decompose the interference patterns experienced by D-TDD in a typical three-cell network, and propose novel interference alignment (IA) inspired opportunistic user scheduling and transmission schemes for each interference pattern. Our scheme mainly relies on the reference signal space (RSS) that guides the transmit beamforming and user scheduling to align the generated interference at a predefined subspace in each AP receiver with best effort. Numerical results show that our schemes can effectively mitigate the CI in D-TDD, and in conjunction with modified opportunistic IA for pure UL/DL transmission, our OIM framework achieves substantial performance again over the conventional schemes.