A Distributed Approach to Interference Alignment in OFDM-based Two-tiered Networks

International audienceIn this contribution, we consider a heterogeneous two-tiered network where a tier of cognitive self-organizing small cells is deployed over the coverage area of a preexisting orthogonal frequency division multiple access (OFDMA) macro cell. Neither cross- nor co-tier cooperation or communication is established, thus no joint signal processing techniques can be implemented by the macro and small base stations (MBS/SBS). We assume that each SBS disposes only of a perfect local channel state information (CSI) w.r.t. the interfering links towards the macro and small user equipments (MUEs/SUEs) operating inside its coverage area. We design a novel cognitive interference alignment (IA) based scheme to protect the macro cell from the cross-tier interference, while mitigating the multi-user interference in the second tier. The optimal precoder that maximizes the spectral efficiency of the link connecting each SBS to its served SUE is found by means of a distributed one-shot strategy. Our numerical findings reveal non-negligible spectral efficiency enhancements with respect to traditional time division multiple access (TDMA) approaches. We show that the peculiar dimensionality of the considered non cooperative scenario is such that, the channel information w.r.t. the interfering links towards the SUEs does not yield spectral efficiency enhancements to the second tier. Finally, the proposed technique exhibits significant robustness to channel estimation errors, achieving remarkable results for the imperfect CSI case, and yielding consistent performance enhancements to the two-tiered network