A Novel Detection and Tracking Algorithm of Chirp Type Civilian GNSS Interference

C.H. Kang, S.Y. Kim, C.G. Park

Abstract:	As the number of wireless applications relying on global navigation satellite system (GNSS) for localization and navigation is increasing, the potential threat of GNSS operation caused by intentional jamming is increasing. In recent years, South Korea has experienced several jamming attacks on the global positioning system (GPS). By investigation, the types of jamming signals were single-tone continuous wave interference (CWI) on L1, and chirp type interference on L2 and L5 frequency bands. Furthermore, previous works surveyed the signal properties of commercial GPS jammers purchased online. The test of jammers provided information on the characteristics of current civil GPS jammer signals. The majority of the jammers used chirp signals, all jammed L1 band, only six jammed L2 band and none jammed L5 band. The sweep rate of jammers is on average about 1~2000 GHz/sec. In order to deal with the threat, detecting and characterizing interference, and giving timely alert, are important for safe GPS operation in all countries, as well as in South Korea. Most of previous GNSS interference detection methods have focused on detecting the existence of CWI using an adaptive notch filter, without identification of the interference type. These approaches have a limitation on detection and mitigation of chirp type interference, because its sweep rate degrades the signal tracking performance of the adaptive notch filter. Thus, it is necessary to classify the types of CWI, and use different detection parameters. In this paper, GNSS interference detection and tracking algorithm using the instantaneous frequency of GNSS interference is proposed to detect the chirp type interference whose sweep bandwidth is 4MHz, sweep period is 4 and sweep rate is 1000GHz/sec as well as single-tone CWI. Instantaneous frequency (IF) provides important information about the change of frequency in non-stationary signals. The concept of IF is popularly used in various technical fields and applications such as radar, sonar, and communications. There are many approaches for IF estimation: time-frequency representation approaches (linear, quadratic approaches), IF estimation methods using analytic signal through the Hilbert transform, etc. One of time-frequency representation approaches is the short time Fourier transform (STFT) which highly depends on window size. The wavelet transform is linear time-frequency representation similar to the STFT, but with more flexibility in time and frequency resolution. The Wigner-Ville Distribution (WVD) is a quadratic time-frequency analysis method, and it does not need to specify a window type like the STFT method. In estimation approaches using analytic signal, there are many variation methods for practical calculation using a finite-impulse response (FIR) differentiator. In this paper, IF of a chirp type civilian GNSS interference is computed by numerical summation of received signal samples through properties of trigonometrical functions in the time-domain because it would generate IF estimation results rapidly and could be relatively easily interpreted in the real-time system. However, this value has a lot of error caused by measurement noise and the frequency change of the interference. In order to reduce the error, the frequency tracking algorithm based on adaptive fading Kalman filter is proposed with the interference detection algorithm. The proposed algorithm is expressed as follows. In the first, the proposed interference detection algorithm based on the property of AGC gain is activated. The interference is a main source that changes the AGC gain because the power of the GNSS signal is below the thermal noise floor. Thus, AGC can be a valuable tool for detecting interference. The value of AGC gain was examined according to the increasing power of interference in previous work. Based on these results, the value of AGC gain is used for detecting the existence of interference sources in this paper. If the magnitude of AGC gain is below the threshold, which means interference signal exists in the received signal, the proposed tracking algorithm is activated to track the interference frequency. The interference tracking algorithm is designed to estimate the instantaneous frequency of the received interference. Pattern recognition algorithm with the low pass differentiator (LPD) is also used to estimate the sweep period of chirp type GNSS interference which is needed to reset the filter state and initial error covariance to track the chirp type interference accurately. The concept of adaptive fading Kalman filter is to apply a factor to the Kalman gain to adjust tracking performance of the filter. In this paper, the factor is applied to compensate the measurement error by decreasing the Kalman gain and is calculated by the relation between the calculated innovation covariance and an estimated innovation covariance because the innovation covariance of the filter can present the effect of unaccounted errors. In order to estimate the sweep period, the measurement passing through LPD is normalized. The normalized signal has a peak at the beginning of the sweep period where the frequency of interference changes radically. If the time index at the peak is measured, the sweep period is able to be estimated. However, the peak is hard to be recognized by a certain threshold because the signal passing through LPD has a lot of noise due to the measurement error. Thus, in this paper, pattern enhancement algorithm is used to detect the peak effectively for reducing the effect of noise. The algorithm is based on an autoregressive (AR) model and the AR coefficients are estimated with the use of Kalman filter. The performance of the proposed algorithm is verified by simulations which are compared to another method such as a conventional discrete Kalman filter (DKF), adaptive fading filter without period estimation (AKF) whose models are the same as that of proposed algorithm. By analyzing the results of comparisons, the proposed algorithm has the better tracking performance than conventional algorithms when the jammer to signal ratio (J/S) of GNSS interference signal is more than 30dB, which seriously affects the accuracy of the GNSS.
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	2910 - 2915
Cite this article:	Kang, C.H., Kim, S.Y., Park, C.G., "A Novel Detection and Tracking Algorithm of Chirp Type Civilian GNSS Interference," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 2910-2915.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In