Analysis and detection algorithms for complex time zeros of bandlimited signals

Bandlimited signals can be expressed in terms of their time zeros, which leads to viewing these zeros as informational attributes of the signals in a similar way to Fourier coefficients and Nyquist samples. The zeros can be used to reconstruct the signal. It is shown that for any bandlimited signal, having a finite number of Nyquist samples, the time zeros can be uniquely used to describe the signal. Computer simulation provides illustrative examples concerning the different manifestation of real and complex zeros of bandlimited signals. Two novel techniques of locating the complex zeros of bandlimited signals are presented. The two techniques effectively are applied concurrently in an efficient algorithm which aims at minimising the amount of processing required to locate the complex zeros. Computer simulation results show the efficiency of the algorithm. Some practical implementations are also discussed.