Article Preview
TopIntroduction
In bandwidth constrained wireless systems, the multiple input multiple output communication has been considered as an important technology as it can utilize all the merits of multiple antennas without any additional spectrum for drastically increasing the capacity (Hua 2017). When MIMO systems are deployed, there are few critical factors such as growing complexity in signal processing and high cost of multiple analog chains (like analog to digital converter at the end of the receiver end, mixers and low noise amplifiers) (Yuan 2017). Also, there is an increase in possibility that at least few of the antennas go through deep fading as there is an increase in quantity of antennas (Eskandari et al. 2018). For handling these critical issues, many strategies involving selection of antenna subset have been proposed by many researchers. The main concept of selection of antenna is to utilize certain quantity of analog chains that are adaptively switched to a subset of the available antennas which is capable of minimizing the quantity of radio frequency chains needed, also maintaining the selection with diversity gains. MIMO is considered as one of the most prospective technologies as it is capable of enhancing the spectral efficiency which in return enhances the transmission reliability and security (Krishna 2015). Base station, antenna arrays with a few hundred components, serving many tens of thousands of active nodes simultaneously and frequency resources are used by massive MIMO systems (Choi et al. 2018). In traditional MIMO system, many antennas on both ends provide more reliable high speed connections utilizing wireless channel diversity. By utilizing hundreds of antennas that use advances in parallel digital signal processing and high speed electronics, massive MIMO focuses on further improving the high speed connections. For transmitting signal energy into smaller areas, additional antennas are used (Zhai et al. 2017). When the proposed approach is combined with the simultaneous scheduling of multiple user nodes, the results provide better performance in terms of energy efficiency. But so as to guarantee communication, every antenna has to be connected to a radio frequency chain that will increase the expensive cost and energy consumption of hardware implementation. This has attracted a lot of researchers' interest towards the antenna selection approach. Many antenna selection algorithms and conditions have been presented in the last decade for traditional small scale MIMO (Yuan et al. 2017). Most of these approaches have concentrated on the selection criteria based on capacity. Studies and extension on some of these approaches have been performed for the massive MIMO systems. But, in the sense of maximum capacity, none of these algorithms for antenna selection for massive MIMO are optimal. Because of increase in quantity of antenna subsets, the optimal traditional small scale MIMO approach such as exhaustive search approach becomes ineffective for massive MIMO system (Zhai et al. 2018). Though there are many advantages for MIMO system, adaptation of them in real time wireless system has been a very slow process. The main reason for this slow adaptation is the more efforts required by MIMO in terms of hardware necessity (Dong et al. 2011). Though the elements of antenna are cheap, the receiver side antenna elements need a complete RF chain that includes an analog to digital converter, frequency down converter and a low noise amplifier which further add to the complexity of hardware requirements. Optimal Antenna subset selection is a reliable method that deals with the problems on hardware complexity (Bana et al. 2019). Therefore an optimal approach for antenna selection that is faster than the traditional approach is very much needed for the massive MIMO antenna selection systems. There are several nature inspired techniques for optimization were developed. Which include: Particle Swarm Optimization (PSO), Gravitational Search Algorithm (GSA), Genetic Algorithm (GA), Evolutionary Algorithm (EA), Deferential Evolution (DE), Ant Colony Optimization (ACO), Biogeographic Based Optimization (BBO), Firefly Algorithm (FA), and Bat Algorithm (BA). The common aim of these algorithms is to find the best quality solutions and boost the efficiency of the convergence. In do so, an inspired version of nature should be fitted with discovery and extraction in ensure that maximum global finding is achieved. Ultimately, the goal of all variants inspired by nature is to combine discovery and exploitation capabilities capable of searching for the best optimal global solution in the quest space.