Suboptimal Receivers for Weak M-ary Chirp Signals in Non-Gaussian Noise

Inhalt:
In this paper, performance of suboptimal receivers for detection of weak M-ary chirp spread spectrum (MCSS) signals in non-Gaussian noise is considered. It is known that optimal receivers for detection weak signals in non-Gaussian consists of zero-memory nonlinearity (ZMNL) followed by the Gaussian correlator receiver. In order to implement the ZMNL, precise a priori knowledge of the first order probability density of the noise is required. However, in practice, it is difficult to know a priori the density of noise precisely. Therefore, it is important to study how an optimal receiver designed for an assumed or estimated noise density will perform when the actual noise density differs from that of the assumed one. In addition, it is useful to consider receivers with various nonlinearities, such as hard limiter, clipper, and hole puncher, that are easy to implement. In this paper, receivers with these suboptimal nonlinearities are analyzed and their symbol error rate (SER) estimates are obtained. The results are shown using the widely used ε-mixture noise model.