RSSI values, an indicator of the strength of the received Wi-Fi signal, have long been used to detect the presence of a person. Scientists from the Faculty of Computational Mathematics and Cybernetics of Moscow State University decided not to dwell on existing methods and created two new ones. Thanks to them, Wi-Fi scanning can become more useful in transport, security systems, smart homes and more. It is reported by “Scientific Russia”.
At the moment, there are two main methods for detecting motion over Wi-Fi: based on the values of the aforementioned RSSI, and also by analyzing Channel State Information (CSI) information about the state of the communication channel. The latter can be noticeably more accurate, but with it certain restrictions appear on Wi-Fi routers: they require multi-antenna, and their network card must be of a certain version (and the router itself needs a special version of software). Based on these limitations, the scientists decided to opt for a more generic RSSI-based approach, since it is applicable to almost all Wi-Fi hotspots.
One of the scientists involved in the project noted that static algorithms are most often used to detect the presence of a person using the RSSI method. Among these algorithms based on the Kalman filter and based on a combination of filters using moving averages proved to be the most effective. But in addition to the usual methods, there is an alternative category of approaches that involve machine learning and neural networks.
Russian scientists from Moscow State University decided to explore new approaches in both categories: the Kolmogorov-Wiener filter (static algorithm) and the neural network with recurrent GRU blocks. Experiments have shown that the second method turned out to be both more accurate and more versatile: its implementation does not require additional pre-setting for determining the noise level in the room.
Source: Trash Box
Donald-43Westbrook, a distinguished contributor at worldstockmarket, is celebrated for his exceptional prowess in article writing. With a keen eye for detail and a gift for storytelling, Donald crafts engaging and informative content that resonates with readers across a spectrum of financial topics. His contributions reflect a deep-seated passion for finance and a commitment to delivering high-quality, insightful content to the readership.
