Engineering Applications of Artificial Intelligence
Low-complexity joint transmit and receive antenna selection for MIMO systems
Introduction
Multiple-input-multiple-output (MIMO) wireless communication systems have significantly higher channel capacity than the single-input-single-output (SISO) system for the same total transmission power and bandwidth (Telatar, 1999, Goldsmith et al., 2006). However, increased number of antennas in the MIMO system has a cost consequence. In practical wireless communication systems, signals of each antenna go through a separate Radio Frequency (RF) hardware known as RF chain (e.g., amplifiers, analog-to-digital converters, etc). In MIMO systems, combining signals carried by a larger number of antennas results in a high hardware cost due to the large number of RF chains. Molisch et al. (2005) showed that hardware cost can be significantly reduced by selecting a good subset of antennas from the set of physically available antennas and using the signals from the selected antennas only, without sacrificing the advantage of multiantenna diversity. A block diagram of MIMO system with RF chain is shown in Fig. 1. Sanayei and Nosratinia, 2004a, Sanayei and Nosratinia, 2004b provided a detail description of antenna selection problem and its benefits. In this paper, we discuss the algorithm to select a good subset of antennas on the basis of channel conditions in real-time.
The complexity of optimally selecting transmit and receive antennas can increase exponentially with the number of transmit and receive antennas. Exhaustive Search Algorithm (ESA) evaluates all possible combinations of transmit and receive antennas to select the antenna combination that gives the best performance (performance such as the channel capacity, the bit error probability, etc.). Enumerating over all possible combinations and finding the one that can give best performance is computationally inefficient. Therefore, in a real-time environment, in which the channel condition changes frequently and the antenna selection should quickly adapt to changing channels, a computationally efficient algorithm is needed. A number of antenna selection schemes, with a motivation to reduce computational complexity have been presented in literature (Chun et al., 2008, Gorokhov et al., 2003, Gore and Paulraj, 2002, Gorokhov et al., 2004, Lu and Fang, 2007, Phan and Tellambura, 2007, Sanayei and Nosratinia, 2007, Sanayei and Nosratinia, 2004a, Sanayei and Nosratinia, 2004b, Uchida et al., 2010, Gharavi-Alkhansari and Gershman, 2004). In the recent years, antenna selection schemes are also proposed for wireless LAN, Cognitive radio and MIMO relay networks—e.g., Hanif et al. (2010), Mestanov et al. (2010), Ming et al. (2010), Truong et al. (2010), Yangyang et al. (2010) etc. In most of the existing literature, antenna is selected on one side i.e., either transmit antenna selection or receive antenna selection. There is very little work for joint transmit and receive antenna selection. In this paper, we apply Binary Particle Swarm Optimization (BPSO) to the real-time joint transmit and receive antenna selection problem. We focus on the joint antenna selection problem because the computational complexity is typically more challenging in joint selection; however, the same scheme is applicable for separate receive and transmit antenna selections.
The rest of the paper is organized as follows. The system model is illustrated in Section 2. In Section 3, we present a new approach for joint transmit and receive antenna selection, which uses BPSO. We also discuss how to tailor the conventional BPSO to the joint antenna selection problem to improve the algorithms' speed of converging a good (nearly optimal) solution. Simulation results and computational complexities are presented in Section 4.
Section snippets
System model
We consider a MIMO system with NT transmits antennas and NR received antennas. Because of cost concern, we consider a system that has only Nt and Nr RF chains at the transmitter and receiver, respectively, where Nr≤NR and Nt≤NT. It is assumed that the receiver has the channel side information (CSI), but the transmitter does not. We denote the channel state by complex matrix . On the basis of this known CSI, the receiver selects Nr receives antennas to connect to RF chains from the NR
Binary particle swarm optimization for joint antenna selection
In this section we present a joint transmit and receive antenna selection scheme that utilizes Particle Swarm Optimization (PSO). PSO is a collaborative computational technique that is derived from the social behavior of bird flocking and fish schooling. The PSO is found robust in solving global optimization problems (Kennedy and Eberhart, 1995). Like other evolutionary algorithms, PSO is a population-based search algorithm. In PSO, each individual is termed as a particle and a collection of
Computational complexity and simulation results
For performance comparison, we present the simulation results of the proposed BPSO joint transmit and receive antenna selection and compare them with the results from existing antenna selection schemes. The channel is assumed to be quasi-static, and different channels in the MIMO are assumed statistically independent. The parameters used for simulations are selected such that the performance of BPSO is evaluated for different search space sizes and signal to noise ratios. All the simulations
Conclusion
In this paper, we presented BPSO algorithms for a joint transmit and receive antenna selection problem. The BPSO for the antenna selection problem requires very low computation and its performance is close to that of the exhaustive search. BPSO, as an evolutionary computing has a simple evolution model, resistance to trap in local minima, and low implementation complexity. This paper indicates that BPSO is a suitable candidate for solving complex communication problems like the joint antenna
References (23)
- Chun N., Zhifeng T., Mehta N.B., Molisch A.F., Jinyun Z., Kuze, T., war S., 2008. Antenna selection for next generation...
- et al.
An Introduction to MIMO Communication
(2006) - et al.
Receive antenna selection for MIMO spatial multiplexing: theory and algorithms
IEEE Trans. Signal Process.
(2003) - et al.
MIMO antenna subset selection for space-time coding
IEEE Trans. Signal Process.
(2002) - Gorokhov, A., Collados, M., Gore, D., Paulraj, A., 2004. Transmit/receive MIMO antenna subset selection. In:...
- et al.
Matrix Computations
(1983) - et al.
Fast antenna subset selection in MIMO systems
IEEE Trans, Signal Process.
(2004) - Hanif, Muhammad F., Smith, Peter J., 2010. On MIMO cognitive radios with antenna selection. In: Proceedings of the IEEE...
- Kennedy, J., Eberhart, R.C., 1995. Particle swarm optimization. In: Proceedings of the IEEE Conference on 13Neural...
- Kennedy, J., Eberhart, R.C., 1997. A discrete binary version of the particle swarm optimization. In: Proceedings of the...
Joint transmit/receive antenna selection in MIMO systems based on the priority-based genetic algorithm
IEEE Antennas Wireless Propag. Pap.
Cited by (22)
AutoML-GWL: Automated machine learning model for the prediction of groundwater level
2024, Engineering Applications of Artificial IntelligenceLow complexity based scheduling methods for multi-user MIMO systems
2020, Physical CommunicationJoint transmit antenna selection and precoding for millimeter wave massive MIMO systems
2020, Physical CommunicationCitation Excerpt :In real time scenarios, where the channel conditions are frequently changing, a computationally efficient antenna selection algorithm is required. In existing literature, minimum eigenvalue [17] and constrained cross entropy optimization (CCEO) [18] are proposed to sub-optimally find the best possible combination of available antennas. [19] and [20] demonstrates the application of binary particle swarm optimization (BPSO) algorithm for efficient transmit and receive antenna selection. [20]
Quantized feedback scheduling for MIMO-OFDM broadcast networks with subcarrier clustering
2017, Ad Hoc NetworksCitation Excerpt :Hence, GA delivers near optimal solutions swiftly with lower computational complexity [20,21,33–35] as compared to the ESP. Though GA cannot find the optimal solution immediately, GA usually finds very good optimal solution for the optimization problems in short span of time [20,21,33,35]. GA can also provide optimum solutions to convex, non-convex, and discontinued optimization functions with some abnormalities efficiently unlike other optimization techniques [33].
Quality of service aware antenna selection scheme for multi-hop relay networks
2017, AEU - International Journal of Electronics and CommunicationsCitation Excerpt :The intention of the entire antenna selection scheme is to choose the perfect and reliable subset of antennas from the available antennas and to associate them with the RF chains. In recent years, antenna selection schemes are also proposed for cognitive radio, wireless local area network (WLAN), massive MIMO and MHR networks [18]. In literature, the impact of different antenna selection schemes on transmit beamformed [29] and space time coded [30–33] systems are investigated.
A computationally efficient genetic algorithm for MIMO broadcast scheduling
2015, Applied Soft Computing JournalCitation Excerpt :Authors in Ref. [19] considered precoding matrices, Zero Forcing Beamforming (ZFB) and water-filling algorithm for optimum power allocation. Joint antenna and user selection scheme is implemented with different evolutionary algorithms for MU MIMO uplink system [20,21]. These works motivate us to consider the use of BGA for combined user and antenna scheduling (CUAS) for MU MIMO BC systems.