Abstract:
Wifi has become an omnipresent technology these days. We can find Wifi connection almost everywhere from public spaces like restaurants, educational institutions, transport stations to private homes and offices. Despite this widespread adoption of the technology for data transmission, it is still not being used to its full potential. Wifi has an intrinsic ability to sense the surroundings by reading the signals passing from one device to another. Using ‘wifi sensing’ we can detect objects and their movements in the environment. Many studies such as human activity recognition, finger movement recognition, fall detection etc. have been done on this. Wifi sensing has the potential to deliver groundbreaking applications in the field of IoT, healthcare and entertainment among others. But utilizing wifi channels for sensing purposes renders them useless for data transmission. This poses a challenge for sensing because we need to design methods that support sensing as well as data transmission without hurting the performance of either of them. This project aims to discover techniques that solve this challenge. Using CSI (Channel State Information) data we try to figure out the channels which provide better sensing capabilities so that the other channels could be used for data transmission.
