Swarm intelligent based detection in the uplink of large-scale MIMO wireless communication systems

Abstract

Large-scale multiple-input multiple-output (MIMO) system plays a vital role in realizing the ever-increasing demand for high-speed data in 5G and beyond wireless communication systems. MIMO systems employ multiple antennas at both the transmitter and receiver. These systems can achieve both the spatial diversity and the spatial multiplexing gain, which are required for enhancing the quality of service (QoS) and the capacity of wireless systems, respectively. Howbeit, reliable detection of the transmitted data streams is challenging due to the presence of inter-channel interference and inter-user interference. To address the above symbol detection issues, maximum likelihood (ML) (Van Trees, Detection, estimation, and modulation theory, part I: detection, estimation, and linear modulation theory, 2004, [34]) detection performs an exhaustive search over all the possible transmitted information symbols and achieves optimal bit error rate (BER) performance. However, being an NP-Hard problem, ML detection is practically unfeasible for large MIMO systems. Therefore, alternate low-complexity robust detection techniques are being devised for near-optimal detection in large MIMO systems. Nature-inspired algorithms have been an emerging choice to obtain a better solution for combinatorial optimization problems. Recently, nature-inspired algorithms has attracted the attention of researchers from wireless communication community, due to its simple implementation and low-complexity behaviour in solving research problems in communication. In this chapter, we have discussed some of the promising bio-inspired techniques such as ant colony optimization and social spider optimization, and introduced one of the key applications of these algorithms, that is, to solve the combinatorial optimization problem of symbol detection in large-scale MIMO systems. We have also compared the BER performance of different bio-inspired algorithms with the traditional low-complexity detection techniques such as zero forcing and minimum mean squared error detectors. © Springer Nature Singapore Pte Ltd. 2020.