Research On Interference Suppression Technique In The GNSS Receiver Under High-dynamic Circumstances

Although global navigation satellite systems (GNSSs) such as the Global Positioning System (GPS) have been widely adopted in various applications, they also exhibit severe system vulnerability, i.e., poor anti-interference capability. This is particularly true for high dynamic range receiving systems; interference suppression techniques designed for static or low dynamic range conditions cannot be directly adapted to them due to the complex electromagnetic environment under high-dynamic conditions. Therefore, enhancing the anti-interference capability and reliability of high-dynamic receiving system has recently become a hot topic of interest in the research community. This paper presents a theoretical study on suppressing blanket jamming under high-dynamic conditions. The main contributions of this research are summarized as follows.1) A MUSIC-PI anti-interference model was established. Conventional PI-based anti-interference techniques cannot distinguish between interference and the signal, which results in a rather large loss of the desired signal. However, the MUSIC-PI algorithm applies the basic principle of the original MUSIC algorithm to the estimation of the actual steering vector and subsequently facilitates nulling by applying the principle of minimum energy when the noise subspace and signal subspace are orthogonal. Therefore, this model is capable of estimating the optimal weight and effectively suppressing blanket jamming while minimizing the loss of satellite signal. When the signal-to-noise ratio was lower than 20 dB, the LMS and PI algorithms failed, whereas the MUSIC-PI method still achieved accurate nulling to reach a suppression depth of-120 dB with improved robustness.2) The problem of high-dynamic time-varying interference was investigated by developing an HMM-Multi-PI cascade interference suppression model, where the HMM algorithm is the first stage and PI & Multi-PI guidance algorithms are the second stage. The PI&Multi-PI model is a semi-blind guidance model that uses the PI algorithm to carry out preliminary interference suppression before employing the Multi-PI algorithm to further enhance the interference suppression performance. The performance levels for satellite capture have the following order:HMM-Multi-PI algorithm, Multi-PI algorithm, HMM algorithm and PI algorithm. The test results showed that the HMM-Multi-PI algorithm is advantageous in terms of suppressing time-varying interference and narrowband-wideband mixed interference.3) The estimation of the DOAs of interference sources in the high-dynamic circumstances was studied. A DOA-based constrained convex optimization model (CCOB) can effectively improve the DOA performance under mismatch conditions to strengthen the system robustness. An NMLC model that only requires the DOA of the interference suppression is proposed to address the difficulty of determining the DOA by beamforming with weak satellite signals. Based on this, the more advanced NMLCB algorithm was developed by introducing virtual nulling to the neighborhood of the DOA and implementing power maximization processing for the rest of the space. The performance levels for the suppression depth and nulling width of the related algorithms have the following order: NMLCB algorithm, DN algorithm, PI algorithm, and NMLCB algorithm, DN algorithm, PI algorithm, respectively. The test results showed that the NMLCB algorithm is effective when dealing with the high-dynamic DOA mismatch problem of a GNSS receiver.4) The undersampling problem of interference sources in high-dynamic circumstance was examined by developing a CCOB-CS model of the sampling signal. This model carries out CCOB processing of the sampling signal to enhance DOA robustness before performing signal reconstruction by utilizing the OMP method. The test results showed that the problem of insufficient snapshots during sampling can be successfully solved by applying the principle of sparse reconstruction in the field of high-dynamic anti-interference array processing. Compared with traditional anti-interference models, this proposed model was proven to remarkably improve the signal-to-interference-plus-noise ratio (SINR).5) The problem of undersampling DOA estimation was studied. A DOA estimation framework with l0-norm continuous CS (ZNM) was established. Because it is a meshless model, it breaks the original framework of signal subspace model. Therefore, this method is suitable for DOA estimation problems with insufficient snapshots. The errors of this proposed method were compared with classic spatial spectra analysis algorithms to give the following order of reduction performance:TLS-ESPRIT algorithm, TAM algorithm, MUSIC algorithm, root-MUSIC algorithm and ZNM algorithm. The test results demonstrated that the ZNM algorithm yields the smallest error. Moreover, a code capture test under zero delay also verified that this method provides higher DOA estimation accuracy and thus promotes a better interference suppression effect.6) A FPGA+DSP hardware/software integrated array GPS anti-interference system was developed. The main modules of the system include an interference source, eight-way RF front end, and baseband processor. Three separate tests were designed to investigate the impacts of the interference type, sampling frequency, and interference source DOA on the GPS receiver and effect of interference suppression.In summary, an in-depth study was conducted on the interference suppression performance of high-dynamic GPS receiver. Theoretical measures were proposed to address the time-varying interference, randomness of interference DOA, and undersampling. A prototype GPS anti-interference device was built and practically tested. The theories and test results reported in this paper are expected to lay a foundation for future research on high-dynamic interference suppression.