Optimal Design of Switchable Wearable Antenna Array for Wireless Sensor Networks

• Published in: Sensors (Basel, Switzerland) Vol. 20; no. 10; p. 2795
• Main Authors: Januszkiewicz, Łukasz, Di Barba, Paolo, Hausman, Sławomir
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 14.05.2020; MDPI
• Subjects: Algorithms; Antenna arrays; Antenna design; Antennas; Arrays; Computer Simulation; Design optimization; FDTD; finite-difference time-domain; Genetic algorithms; Heuristics; Human body; Humans; Internet of Things; Medical research; Optimization; Physiology; Sensor arrays; Sensors; signal-to-interference ratio; WBAN; Wearable Electronic Devices; Wearable technology; wireless body area network; Wireless networks; Wireless sensor networks; Wireless Technology; evolutionary computing; genetic algorithms; wireless body area network; finite-difference time-domain; radiation pattern synthesis; FDTD; switchable antenna; WBAN; signal-to-interference ratio; interference rejection; optimization; human body model; SIR
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s20102795

In the paper, we present a novel approach to the optimum design of wearable antenna arrays intended for off-body links of wireless body area networks. Specifically, we investigate a four-element array that has a switchable radiation pattern able to direct its higher gain towards a signal source and a lower gain towards an interference. The aim is to increase the signal to interference ratio. We apply a genetic algorithm to optimize both the spatial placement and the feed phasing of the elementary on-body antennas. We propose a simplified, computationally efficient model for the simulation of the array radiation pattern. The model is based on full-wave simulations obtained with a simplified cylindrical model of the human body. We also propose, implement, and evaluate four objective functions based on signal to interference ratio, i.e., min-max, nadir point distance maximization, utopia point distance minimization, and full Pareto-like. Our optimized design obtained with this approach exhibits a significant performance improvement in comparison to the initial heuristic design.