Enabling Technologies of Long-Range Aerial Acoustic Communications and Acoustically-Based Multi-Person Motion Recognition

Research in aerial acoustic communications has undergone extensive development over the years. However, achieving an effective communication range for aerial acoustic communications in indoor environments remains a critical challenge, particularly for smartphone applications. To overcome this limitation, firstly is presented a novel near-ultrasonic acoustic communication system on the Android platform. The system surpasses the range limitations by employing an innovative receiver design that incorporates fractional Fourier transform and 2-D mask filters to effectively denoise the signals. Through rigorous onsite testing, the proposed communication system demonstrates error-free communications up to a distance of 70 meters. Conventional aerial acoustic communication schemes are often vulnerable to privacy risks due to the broadcast nature of wireless communications. Moreover, as the communication distance increases, these risks become more prominent. In order to address this concern, secondly a secure cooperative aerial acoustic communication scheme based on distributed beamforming is proposed. By leveraging signal processing technologies utilized in long-range aerial acoustic communications, the proposed scheme enhances privacy and security. In the era of expanding intelligent environments, the imperative to detect and recognise multi-person resonates across diverse applications, encompassing domains such as smart homes, security surveillance, and human-computer interaction. Finally, a novel acoustically-based application for multi-person motion recognition and identification is introduced by utilizing sound of footsteps. The proposed approach involves extracting time-frequency distributed features using discrete wavelet transform, followed by training a classifier using supervised learning techniques. A comprehensive repository of multi-person audio recordings is meticulously curated, substantiating the approach's robustness and efficacy via extensive experiments conducted in diverse indoor settings. Overall, this thesis provides comprehensive solutions to address the challenges in achieving an effective communication range for aerial acoustic communication in indoor environments. It also explores secure cooperative communication schemes and introduces innovative applications for multi-person motion recognition and identification.