Performance of Hand Gesture Recognition Based on Received Signal Strength with Weighting Signaling in Wireless Communications

In recent years, with increasing wearable devices, we apply a wearable wireless node to model the wireless channel for hand gesture recognition in this study. With transmission and reception of wireless signals, the received signal strength indications (RSSI) are obtained to model the wireless channel. When people wear a wearable device, the hand gestures make the different RSSI through the respective channels. After we calculate the Euclidean distance between the reference patterns to perform the pattern matching. Moreover, we propose a significant reference patterns (SRP) method to train reference patterns for the pattern matching. Thus, the received RSSI patterns for hand gestures recognition are retrieved by weighting sequence. The recognition on two gestures of up and horizontal movement is investigated. Experimental results show that the recognition rate for independent people is higher than 99% for three persons.