Low-Complexity 1-bit Detection of Parametric Signals for IoT Sensing Applications

A detection method for parametric signals in noise is proposed. Since the detection is performed by processing 1-bit quantized data, the method can be implemented using low-complexity systems. A generalized likelihood ratio test (GLRT) is derived, which is based on parametric estimates of the amplitude and phase of a periodic signal. The asymptotic properties of the test statistic of the GLRT are derived analytically. Such properties are then employed to study the performance of the detector, by comparing them with the clairvoyant detector. The performance of the developed detector is evaluated by numerical simulations and validated by a simple experimental setup. The impact of noise and threshold selection on detection performance and the promptness of response are also numerically analyzed. Given its implementation simplicity, the proposed detector does not require high-resolution circuitry nor computationally intensive hardware. Therefore, it could enable novel applications in the Internet-of-Things domain, such as low-cost radar systems and sensor networks.