Discrete Beamforming Using Low-Cost Energy-Efficient ESPAR Antenna for Jamming Mitigation in IoT Devices
Mass deployment of low-power IoT nodes further demands the resilience to interference and malicious RF jamming as a first-order design requirement. This paper proposes a realistic anti-jamming beamforming approach using an electronically steerable parasitic array radiator (ESPAR) antenna equipped with 12 parasitic elements supporting binary (open/short) switching, which can change the parasitic state in 4096 discrete ways. We introduce Binary Simultaneous Perturbation Stochastic Approximation (BSPSA), which is a blind binary beamforming algorithm that maximizes the instantaneous value of any given link-quality indicator (e.g., SINR) without requiring any reference signal and can be implemented with microcontroller level complexity. By taking radiation patterns of an actual ESPAR and running a simulation with a single desired user and a single jammer, the method nearly achieves the brute-force upper bound, being orders of magnitude faster. An anechoic-chamber SDR testbed with a CW desired signal and Gaussian-noise jammer confirms the method providing results that are better than 11 dB in 90% of runs with around 16-17 dB median gain within 50 iterations. The measurements show that a single discrete ESPAR beam switching and BSPSA can provide a practical anti-jamming performance for the IoT devices and gateways, achieving a significant degree of interference reduction at low hardware and energy consumption.