Wavelet based signal processing algorithms for early target detection

Detection of radar target earlier, that is, at a longer distance, amidst noise is a major issue since the received echo signal has very low input signal to noise ratio of the received signal. Further, it becomes very difficult due to unknown statistics and nonstationary nature of the received signal. This paper deals with the wavelet based signal processing algorithms which could detect target even with very weak received signal. Since the signals with identical spectra may have different time-frequency image in time-frequency domain. The algorithms use optimum bandwidth relative to the signal spectral width to maximize the output signal to noise ratio and hence probability of detection. Simulation results show that the wavelet filtering algorithm and entropy based basis algorithms are able to detect target from the input signal to noise ratio of -10dB.