Efficient Techniques for Sparse Multipath Channel Estimation and Strong Interference Cancellation in FM Passive Radars

• Main Authors: Tzu-Hao Su, 蘇子皓
• Other Authors: Jeng-Kuang Huang
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/39363153179109569943

碩士 === 元智大學 === 通訊工程學系 === 98 === Differing from the traditional active radar, the passive radar can detect the unknown targets by processing the FM broadcasting signal. However, the echo signal power for the unknown target is always much weaker than the direct path signal. Hence an direct path interference (DPI) cancellation scheme should be adopted to makes the echo signal detectable. To achieve this purpose, we first build up the transmit signal model including the ambiguity function, the echo formulation and the signal model with direct path interference. Then the Wiener solution for the DPI channel estimation can be derived and simulated under the ultra low signal to interference power ratio (SIR). To be more computationally feasible, adaptive filter and algorithms including least mean square (LMS), normalized least mean square (NLMS), variable step size least mean square (VssLMS) are investigated. Then for the sparse channel environment, we propose an efficient scheme based on the error surface analysis (ESA) etc, which greatly improve the DPI cancellation performance. The above DPI cancellation schemes are based on an integer-delay assumption. However, the DPI delay may contain fractional part. The DPI can not be perfectly eliminated by the former algorithms. Thus, we re-formulate the time domain and frequency domain signal model of the DPI with fractional delays. Then we divide the problem into two parts: fractional interference delay estimation and factional interference delay interpolation. We then use the alternating notch periodogram algorithm (ANPA) estimate the DPI with fractional delays. For signal interpolation, we will apply the sinc interpolator and Farrow interpolator. Finally the unknown aircraft target can be detected from the passive radar, as shown by the simulation results.