Exploiting Sparsity in Amplify-and-Forward Broadband Multiple Relay Selection

Cooperative communication has attracted significant attention in the last decade due to its ability to increase the spatial diversity order with only single-antenna nodes. However, most of the techniques in the literature are not suitable for large cooperative networks such as device-to-device and wireless sensor networks that are composed of a massive number of active devices, which significantly increases the relay selection complexity. Therefore, to solve this problem and enhance the spatial and frequency diversity orders of large amplify and forward cooperative communication networks, in this paper, we develop three multiple relay selection and distributed beamforming techniques that exploit sparse signal recovery theory to process the subcarriers using the low complexity orthogonal matching pursuit algorithm (OMP). In particular, by separating all the subcarriers or some subcarrier groups from each other and by optimizing the selection and beamforming vector(s) using OMP algorithm, a higher level of frequency diversity can be achieved. This increased diversity order allows the proposed techniques to outperform existing techniques in terms of bit error rate at a lower computation complexity. A detailed performance-complexity tradeoff, as well as Monte Carlo simulations, are presented to quantify the performance and efficiency of the proposed techniques.