Compressed Sensing-Assisted Device Discovery for Low Terahertz Communications
Abstract
Compressed sensing-assisted device discovery techniques are useful to efficiently align TX and RX antennas precisely for low THz communications. Various alignment approaches were evaluated by simulation. Orthogonal matching pursuit was determined to be the most promising approach to alignment with the lowest computation time, and a sectorized application refinement improved this approach significantly. This method was validated at 304.26 GHz using a correlation-based m-sequence channel sounder. An average of 2.3° difference between expected and computed angles of arrival (AOA) and departure (AOD) were measured for various scenarios including paths with reflections from a nearby metal plate and wall.