Opportunistic RIS-Aided WiFi Imaging in Smart ElectroMagnetic Environments

Imaging and object detection at Radio Frequency refers to a set of techniques that aims to reconstruct extended or point-shaped objects from a set of complex measurements performed at radio frequency. The system to be investigated uses Reflective Intelligent Surfaces(RIS), which actively interact with the environment to improve RF imaging resolution. Each panel is made of many programmable cells. The RIS exhibits the breaking-through capability of coherently combining the electric field directly incident on a target with the reflected one by introducing a proper phase shift. The target is imaged by reconstructing its reflectivity profile based on the projections measured. The region of interest (ROI), which contains the albedo, is divided into pixels, and the system collects as many measurements as the number of discrete elements. The effortless integration of RIS-aided systems into current WiFi-based electromagnetic environments makes this research worthwhile. The ubiquity and flexibility of the WiFi protocol make integrating these systems much more accessible and far-reaching. Unlike radar or Synthetic Aperture Radar (SAR), which generally employ wide-band pulses, RF imaging at the WiFi level uses an 802.11 protocol to leverage spectrum partitioning. The total available bandwidth is arranged into 40-MHz channels to illuminate the scene. The target reconstruction relies on both bandwidth and measurement diversity given by spatial and standard frequencies proper choices. According to simple WiFi network protocol, two Access Points(APs) regulate the centralized communication over the 2.4GHz or 5GHz frequency domains. The AP, acting as the receiver, collects scattered electromagnetic field samples, which are processed to obtain images of the complex reflectivity profile of the surface under investigation. The use of RIS in image reconstruction at mmWave allows the estimate of the location and shape of the target, carrying out the main benefit of simplifying the hardware for the system. In this context, further investigation into the advancements in the mathematical inversion process, which transforms the multitude of measurements into a reconstructed image, is highly valuable. Simulated results are presented to assess the performance of various inversion algorithms, considering not only computational efficiency and reconstruction capabilities but also noise rejection and robustness of the results.