Exploiting block sparsity for joint mitigation of asynchronous NBI and IN in hybrid powerline-wireless communications

Transmission of the same information signal simultaneously over multiple physical layers, such as powerline and unlicensed wireless communication networks, results in higher reliability and/or enhances the coverage range compared to using a single physical layer due to diversity gains. However, each physical layer suffers from its own distinct impairments that can severely degrade performance. Unlicensed wireless communications suffers from narrow-band interference (NBI) while powerline communications (PLC) is impaired by impulsive noise (IN). With orthogonal frequency division multiplexing used for both communication systems, these two impairments, if not mitigated, can severely degrade performance. This paper proposes an approach for efficient joint estimation and mitigation of the NBI and IN signals in hybrid wireless and PLC systems. The proposed approach exploits the inherent sparse structures of the NBI and IN signals in the frequency and time domains, respectively, and is based on the compressive sensing (CS) principles. The paper also addresses the practical asynchronous NBI scenario that suffers from carrier frequency offset (CFO) with respect to the wireless received signal. In this regard, it investigates the use of time-domain windowing to enhance the NBI´s sparsity and, hence, improve its subsequent estimation and mitigation. Further, the paper enhances the estimation and mitigation of NBI and IN by modeling the burstiness of both impairments as block-sparse vectors. To this end, it investigates the performance of two block-sparse CS recovery algorithms with and without prior knowledge of the bursts´ boundaries. Finally, numerical experiments quantify the performance gains realized by exploiting both the burstiness and sparsity of the NBI and IN signals over exploiting sparsity alone.