適用於衛星導航接收機窄頻干擾消除之低複雜度訊號處理架構

Abstract

全球導航衛星系統(GNSS)在現今已普遍的使用在各個領域，如車用導航、航海的定位上…等，而手機也愈來愈強調具有定位功能。其中，又以美國的GPS衛星系統最為大家熟知及使用。 GPS為展頻系統，其最大優點便是對干擾具有較佳的抵抗力，但若干擾能量太強大，GPS還是沒辦法正常操作，也因此我們需要額外的抗干擾技術來解決干擾問題。傳統上，最常見的方法是從時間維度(FIR Filter)和頻率維度(FFT)來解決這類問題，但這些架構的成本太高，因此我們需要在系統效能可接受的條件下，採用另外的方法來處理干擾問題。 本論文採用一可適性IIR陷波濾波器(IIR Notch Filter)來消除窄頻干擾。我們利用一低複雜度的演算法來預測干擾頻率，並提出一演算法搭配前端AGC來預測干擾訊號的能量大小。此外，還提出一演算法來偵測干擾訊號的存在，以便干擾不存在時，進入後端的GPS訊號能量不會被陷波濾波器所衰減。更進一步，我們分析了陷波濾波器的群延遲效應(group delay)在信號擷取及追蹤部分的影響，並將干擾訊號由單頻(single-tone)推廣到多頻(multi-tone)，分析多級IIR陷波濾波器在信號擷取與追蹤部分的影響。 此外，我們定點分析IIR陷波濾波器架構、可適性演算法及所提出的預測干擾能量、偵測干擾存在演算法，並在定點設計下驗證所提出演算法的效能。

Nowadays GNSS (Global Navigation Satellite System) has been widely used in many places, for example, In-Car Navigation System, navigation on the sea, and so forth. Especially, there are more and more mobile phones that have navigation capability. And among all the navigation systems, GPS is the most widely used and well-known to people. GPS is the spread spectrum system and it has an advantage to resist the interference and jamming. But once the jamming power is too large, GPS still cannot work well. So, we must have anti-jamming techniques to overcome this problem. Traditionally, we usually deal with the interference or jamming from the time domain (FIR Filter) or frequency domain (FFT), but both the complexities are too high. So we try to find a low-complexity solution as long as the performance of the system is acceptable. In this thesis, we utilize an adaptive IIR notch filter to eliminate the narrowband interference. We use a simple algorithm to adaptively estimate the jamming frequency, and propose a method to estimate the jamming power. Furthermore, we propose a method to detect the jamming existence so that GPS signal will not attenuate by the notch filter once there is no jamming. We also analyze the group delay effect in the acquisition and tracking, and see the results caused by the multi-pole notch filter. Besides, we re-evaluated the proposed algorithms by using fixed-point analysis of IIR notch filter and adaptive algorithm.