A Novel GNSS Weak Signal Acquisition Using Wavelet Denoising Method

With the increasing demands of precise positioning in weak signal environment, high sensitive GNSS receiver research and development has been pushed badly in need. Conventional GNSS signal acquisition techniques are considered inadequate when the incoming signal is too weak. In this paper we have mainly consider wavelet denoising algorithm applying in weak GNSS signal acquisition. Conventional wavelet de-noising algorithms include regional scale transformation method and threshold method. The first method requires less limitation about the noise type, but the latter one is applied only in Gauss noise conditions. Besides wavelet de-noising process is done when the signal is independent in time sequence, therefore our work has done based on the traditional correlation acquisition. When the noncorrelation or differential correlation has done, the noise distribution and property has been also changed. If the noise pre-processed is Gauss distributed, the postprocessed noise is no longer Gauss white noise. Under this circumstance we conduct statistics analysis to estimate the derivation of noise, and assume a new Gauss noise. Then the wavelet de-noising process is done. Our algorithm contains three key steps. Firstly, correlation and differential correlation method are used to acquire the very weak signal; secondly, noise derivation is estimated and noise model is established; then the wavelet denoising process is applied. The result turns out fine for the signal lower than other acquisition method. INTRODUCTION There are several weak GPS signal acquisition methods introduced in recent years. Most of them focus on how to increase the PIT (Predict Integrated Time) of the GPS signal and how to predict the data bit transit of GPS signal. Based on these researches and wavelet de-noising theory, we will introduce a new weak GPS signal acquisition method to increase at least 1dB the sensitivity of GPS acquisition. The typical GPS signal arrive at earth is about -160dBW. Typical GPS algorithm use 1ms PIT correlation time to detect this GPS signal. When in in-door usage, the GPS signal could reduce to -180dBW or even lower. By using coherent/non-coherent combined acquisition method and differential acquisition method, -182dBW GPS signal could be detect under 100ms PIT of GPS signal. But the weaker GPS signal detection need much longer PIT due to the signal square loss and doppler frequency change. So, in this paper we introduce the wavelet de-noising method to increase the sensitivity without increasing PIT time. We use wavelet de-noising method to decrease the background noise, thus to increase the signal peak of weak signal acquisition. WEAK GPS SIGNAL ACQUISITION 1. Coherent/Non-coherent (NC) Coherent integration is referred as the regular correlation between the received signal and local generated replica. Usually there are three methods to do a coherent integration: sequential acquisition, parallel phase domain acquisition (IFFT) and parallel frequency domain acquisition (FFT). For regular GPS received signal a 1ms coherent integration can contribute 30dB energy accumulation, which is enough to reach the threshold. But that is not the condition of a weak GPS signal environment. Long time coherent integration is limited by the unknown message of navigation data transferring. For example if the bit edge occurs in the middle of a data sequence, the total coherent integration can be zero due to two magnitude equally integration blocks that have the opposite signs. To eliminate the effect of navigation data on integration non-coherent process is used by squaring the in-phase and quad-phase coherent integration results. This is called coherent/non-coherent combined method.