One-Bit Distributed Sparse Spectrum Sensing Based on the DQA-ZA-LMS and DQA-RZA-LMS Algorithms Over Adaptive Networks

Abstract

In this paper, we proposed the distributed quantization and sparsity aware zero attracting least mean square (DQA-ZA-LMS) and its reweighted version (DQA-RZA-LMS) algorithms that can perform sparse spectrum sensing with the lowest power possible. The usage of the quantization aware diffusion adaptive networks has recently been proposed and they can be used in many possible mobile communicative applications. The sparsity aware feature of the proposed algorithm can help the network to track and estimate sparse random vectors that are shown to be the case with the spectrum of the new generation wireless communication systems such as 4G, 5G, 6G, and beyond. The spectrum sensing is considered in this paper to be performed by small cell eNode Bs (SC-eNBs) for the 4^th generation long term evolution (LTE) and the next generation eNB (ng-eNB) networks for the 5^th and 6^th generation mobile communication systems that are scattered in an area collecting distributed quantized data from the environment and working collaboratively to estimate the sparse spectrum vectors. Our findings show that in comparison with the nonquantized version of the distributed ZA-LMS (DZA-LMS) and distributed regularized ZA-LMS (DRZA-LMS) algorithms, our proposed schemes perform considerably well using the quantized data and also reduce power consumption.