Inter‐radar interference analysis of FMCW radars with different chirp rates
2019; Institution of Engineering and Technology; Volume: 2019; Issue: 19 Linguagem: Inglês
10.1049/joe.2019.0167
ISSN2051-3305
AutoresYuya Makino, Takuya Nozawa, Masahiro Umehira, Xiaoyan Wang, Shigeki Takeda, Hiroshi Kuroda,
Tópico(s)Advanced SAR Imaging Techniques
ResumoThe Journal of EngineeringVolume 2019, Issue 19 p. 5634-5638 IET International Radar Conference (IRC 2018)Open Access Inter-radar interference analysis of FMCW radars with different chirp rates Yuya Makino, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorTakuya Nozawa, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorMasahiro Umehira, Corresponding Author masahiro.umehira.dr@vc.ibaraki.ac.jp Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorXiaoyan Wang, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorShigeki Takeda, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorHiroshi Kuroda, Hitachi Automotive Systems, Ltd., Hitachinaka-shi, JapanSearch for more papers by this author Yuya Makino, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorTakuya Nozawa, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorMasahiro Umehira, Corresponding Author masahiro.umehira.dr@vc.ibaraki.ac.jp Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorXiaoyan Wang, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorShigeki Takeda, Graduate School of Science and Engineering, Ibaraki University, Hitachi-shi, JapanSearch for more papers by this authorHiroshi Kuroda, Hitachi Automotive Systems, Ltd., Hitachinaka-shi, JapanSearch for more papers by this author First published: 25 July 2019 https://doi.org/10.1049/joe.2019.0167Citations: 2AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onEmailFacebookTwitterLinked InRedditWechat Abstract Millimetre-wave frequency-modulated continuous wave (FMCW) radar is expected to be widely deployed for advanced driver assistance systems (ADAS) and self-driving cars. Considering that automotive radars will be deployed in a huge number of cars in future, inter-radar interference will be a significant problem since inter-radar interference can cause miss-detection and/or false detection of the target. As millimetre-wave automotive radar is not standardised at this moment, it is required to evaluate inter-radar interference among FMCW radars with various chirp-rates and chirp durations. This study describes inter-radar interference analysis in FMCW radars with different chirp rates and simulation results of inter-radar interference to confirm the validity of the interference analysis. 1 Introduction In recent years, aiming at safer and more efficient transportation, research and development of ADAS and automated driving technologies is widely carried out all over the world [1, 2]. In ADAS and automated driving car, car driving is partly performed by the computer on-board the car using sensor information and other information such as location information and MAP data. Therefore, ADAS and automated driving car need various types of sensors such as optical cameras, laser imaging detection and ranging (LiDAR), radar, ultra-sonic sonar to recognise the surroundings of a car. As each sensor has advantages and disadvantages, multiple types of sensors are used in combination. Regarding radar, millimetre wave radar can use wide spectrum of 4 GHz in 79 GHz band, thus distance measurement accuracy of 10 cm can be achieved. One of the promising radar for automotive applications is millimetre-wave FMCW radar because it achieves fairly high distance resolution of about 10 cm, and it can simultaneously measure the relative distance and relative speed of the target. Furthermore, direction of the target can be measured by using multiple-input multiple-output (MIMO) configuration and it is relatively inexpensive compared with LiDAR. Therefore, it is expected to use several millimetre-wave FMCW radar in an automated driving car. Considering that automotive FMCW radars will be densely deployed in future, inter-radar interference can be a serious problem since it causes miss-detection of the target and/or false detection of the so-called ghost target. It is well-known that there are two types of radar interference, i.e. narrowband interference and wideband interference [3]. Narrowband interference occurs when the interference radar use the same chirp rate and chirp direction as the observation radar, and the observation radar detects a target that does not exist as a ghost target. On the other hand, wideband interference occurs when the interference radar has different chirp rate from the observation radar, resulting in an impulse-like interference signal in time domain. Fourier transform of impulse-like interference signal causes noise floor increase in frequency domain. Thus, SNR degradation due to wideband interference causes miss-detection of the target. Regarding wideband interference, various inter-radar interference mitigation methods have been proposed, e.g. frequency hopping random chirp (FHRC)-FMCW radar to avoid narrowband interference [4] and weighted envelope normalization (WEN) algorithm to reduce wideband interference [5]. The authors also proposed the interference detection and suppression method in the time domain to mitigate wideband interference [6]. As wideband interference mitigation methods are based on interference suppression in time domain, SNR improvement by interference reduction depends on received level and design parameters such as chirp rate of interference radar. As there is no standard of millimetre-wave FMCW radar so far, a large number of FMCW radar with different design parameters can be used simultaneously. Therefore, it is necessary to evaluate SNR degradation due to inter-radar interference among a large number of FMCW radars with various design parameters. This paper describes inter-radar interference analysis of FMCW radars with various design parameters to estimate SNR degradation caused by wideband interference. The results of wideband interference analysis are confirmed by computer simulations. This paper reveals that inter-radar interference becomes significant when the chirp rate of interference radar is similar to that of interference radar. The inter-radar interference analysis implies that standardisation of FMCW radars is desirable to mitigate wideband interference. 2 Interference in fast chirp FMCW radars 2.1 Fast chirp FMCW radar Fig. 1 shows a block diagram of millimetre-wave FMCW radar. Frequency modulated linear chirp signal is transmitted, and reflected signal from the target is received with delay time. It is mixed down by multiplying the transmitted radar signal to extract beat signals, whose frequency is in proportion to the distance of the target. To detect the frequency of beat signals, AD conversion and fast Fourier transform (FFT) processing are performed and the peak of the frequency, i.e. distance of the target, is detected as the peak in frequency spectrum. Fig. 1Open in figure viewerPowerPoint Block diagram of FMCW radar There are two kinds of modulation waveform of FMCW radar, i.e. triangular wave modulation and saw-tooth wave modulation. The triangular wave modulation uses both up chirp and down chirp to calculate the relative speed and relative distance of the target, however, peak matching of up-chirp and down-chirp becomes difficult when there are many targets to be detected. On the other hand, the saw-tooth wave modulation uses either up-chirp or down-chirp, and the fast chirp modulation with short chirp period is considered to be promising for multi-target detection [7]. In the fast chirp radar, each peak is tracked target by target and velocity of each target can be measured from the distance change divided by the observation time, thus peak matching is not required. Fig. 2 shows the frequency modulation pattern and the beat frequency of the saw-tooth wave modulation using up-chirp, where the beat signal frequency is the frequency difference between the transmitted radar signal and the received radar signal. Lower frequency component than the cut-off frequency of LPF appears at the output of LPF. After AD conversion and FFT, the peak frequency, which has sufficiently larger level than noise level, is detected as the target, and the beat frequency fB is also detected. If there are multiple targets, multiple frequency peaks are detected. The relative distance and relative velocity with respect to the object can be calculated as follows: (1) (2) where R is the relative distance, c is the speed of light, Δf is the frequency bandwidth of chirp signal, fB is the beat frequency, v is the relative speed, ΔR is the measured distance difference between Ri and Ri +1, and T is the chirp duration [8]. Fig. 2Open in figure viewerPowerPoint Fast chirp FMCW radar using sawtooth wave modulation 2.2 Inter-radar interference This paper focuses wideband interference in fast chirp FMCW radar, and it is illustrated in Fig. 3. The above figure shows frequency as a function of time of the transmitted chirp signal, the received chirp signal reflected by the target, and the interference radar signal with different chirp rate. The lower figure shows beat frequency of the desired radar signal reflected by the target and the interference radar signal. Supposing that f LPF is cut-off frequency of LPF in a FMCW radar, a pulse-like interference signal appears when the beat frequency is less than f LPF and its frequency changes from −f LPF to +f LPF. As the received level of the interference radar signal is much larger than that of the received radar signal reflected by the target, the beat signal of the interference radar signal is much larger than that of the desired radar signal, thus the pulse-like interference radar signal with the desired beat signal appears at the output of LPF. As frequency spectrum of pulse signal is flat over the frequency, the noise floor increases in the frequency spectrum of the received radar signal, thus SNR of the peak corresponding to the distance of the target is degraded. This SNR degradation causes the increase of miss detection of the target [4, 5]. Fig. 3Open in figure viewerPowerPoint Wideband interference in Fast Chirp FMCW radar 3 Inter-radar interference analysis 3.1 Inter-radar interference signal power As described in Chapter 2, wideband interference is a pulse-like interference radar signal added to the desired beat signal and it is equivalent to white noise; thus, SNR of the peak frequency corresponding to the target is degraded. It is necessary to estimate the interference power for inter-radar interference analysis. The interference power is in proportion to the received power and the duration of the interference radar signal. The duration of the interference radar signal changes according to the chirp rate of the observation radar signal and the interference radar signal. The duration of the interference radar signal between FMCW radars with different chirp rates is shorter as the difference between the chirp rates of the two radars is larger. It is assumed that LPF is an ideal filter with bandwidth of f LPF. As continuous wave (CW) signal is the signal with chirp rate = 0, the interference radar power for CW signal is used as reference power. Let us suppose that the frequency bandwidth, Δf of the observation radar signal is the same as that of the interference radar signal, the chirp rate of the observation radar is CR and that of the interference radar signal is CRint as follows: (3) (4) where the chirp period of the observation radar signal is T, and that of the interference radar signal is nT. The beat frequency, fB, of the interference radar signal is given by the following equation: (5) As the interference radar signal appears at the output of LPF when |fB | < f LPF, the time duration of the interference radar signal, Δt, is given by the following equation. (6) Supposing that the interference signal is CW, the interference signal also appears at the output of LPF when |fB | < f LPF and the time duration of the interference signal, Δt CW is given by the following equation. (7) Let us suppose that the reference interference signal power for CW signal is P 0, and the observation radar signal with CR is interfered by the interference radar signal with CRint. As the interference radar signal power is in proportion to the time duration given by (6) and (7), the signal power, Pn for the interference radar signal with chirp duration, nT is given by: (8) When the chirp direction of the interference radar is the same as that of the observation radar, n > 0. Equation (8) indicates that the interference radar signal power, Pn increases from and becomes infinity as n approaches to 1. When Δt is equal to or >T, as given by (6), the interference radar signal appears all the time during the chirp duration. When n approaches from 0 to 1, the interference radar signal power, Pn also increases and approaches to infinity in the same manner. The interference radar signal appears during the chirp period, T of the observation radar signal when n is given by the following equation: (9) When the chirp direction of the interference radar is opposite to that of the observation radar, n < 0 in (8) and interference radar signal power becomes 0 when n = 0, i.e. CRint = ∞. As n decreases from 0 to −∞, the interference radar signal power increases from 0 to P 0. Fig. 4 shows the wideband interference power as a function of chirp period, nT of the interference radar. As seen from this figure and indicated by (9), the interference radar signal power becomes significantly large when the chirp rate of the observation radar is close to that of the interference radar. Thus, it is desirable not to use the FMCW radar with different but similar chirp rate at the same time when the chirp direction of the interference radar is the same as that of the observation radar. However, it is expected that the occurrence probability of the wideband interference is small. On the other hand, when n < 0, the interference radar signal power is less than that of P 0. Therefore, even if |n | is almost equal to 1, the interference radar signal power is relatively small. This implies that if it is necessary to use the radar with chirp period of n = 2 for example, its chirp direction should be opposite to that of another radar. It should be noted that interference radar interference is 0 when n = 0. This means that no wideband interference occurs when multiple saw-tooth FMCW radar with the same chirp rate, the chirp duration, T and the chirp bandwidth, Δf are operated at the same time. Fig. 4Open in figure viewerPowerPoint Interference signal power according to chirp period 3.2 Occurrence probability of inter-radar interference Let us suppose that the chirp period of the observation radar is T and the chirp period of the interference radar is nT, and the cut-off frequency of LPF, f LPF is negligibly small compared with chirp frequency bandwidth, Δf. Wideband interference occurrence probability for n > 1, P 1 is given by: (10) This is based on the ratio of the number of interfered chirp duration to all chirp duration. An example of inter-radar interference scenario when n = 5 is shown in Fig. 5 a as an example of n > 1. Wideband interference broadband occurs at the part surrounded by circles in Fig. 5 a. Wideband interference occurrence probability for 0 < n < 1, P 2 is given by the following equation: (11) Fig. 5Open in figure viewerPowerPoint Examples of inter-radar interference scenarios (a) n = 5, (b) n = 0.5 An example of inter-radar interference scenario when n = 0.5 is shown in Fig. 5 b as an example of 0 < n < 1. Interference occurs at the parts surrounded by circles in Fig. 5 b. Fig. 6 shows wideband interference occurrence probability as a function of chirp period of interference radar, nT in the case of saw-tooth FMCW radar. As chirp period of interference radar increases from 1, i.e. chirp rate decreases, occurrence probability of wideband interference increases; however, interference radar signal power decreases as described before. On the other hand, as chirp period decreases from 1 to 0, occurrence probability of wideband interference increases and it reaches 1 when n < 0.5. It is noted that interference radar signal power becomes smaller as n becomes smaller than 0.5 as shown in Fig. 4. However, multiple wideband interference, i.e. pulse-like interference signal, occurs in the desired beat signal. It should be noted that no wideband interference occurs when n = 1; however, narrowband interference occurs resulting in false detection of the target. Fig. 6Open in figure viewerPowerPoint Occurrence probability of wideband interference as a function of chirp period, nT 4 Simulation of inter-radar interference 4.1 Simulation outline Computer simulations have been conducted to confirm the validity of the inter-radar interference analysis. Simulation parameters are shown in Table 1, where both the observation radar and the interference radar are fast chirp FMCW radar using sawtooth wave modulation. The observation radar has the following parameters, i.e. chirp period, T is 100 μs, cut-off frequency of LPF, f LPF is 10 MHz and chirp frequency bandwidth, Δf is 500 MHz. Chirp period of the interference radar is nT, where n = 2, 4, 8, 16 for the same chirp direction and n = −2, −4, −8, −16 for the opposite chirp direction. Table 1. Simulation parameters Transmitted signal Interference signal Parameter Value Value Unit sweep frequency 500 500 MHz sweep time 100 100·n [μs] (n = ±2,4,8,16) μs LPF pass band 10 10 MHz FFT size 1024 1024 — sampling frequency 500 500 MHz Examples of simulation scenarios of inter-radar interference are shown in Fig. 7, where n = 2 and −2. Interference signal power for CW signal P 0 is used as reference interference signal power. Fig. 7Open in figure viewerPowerPoint Simulation scenarios (a) n = 2, (b) n = −2 4.2 Simulation results Fig. 8 shows interference radar signals in time domain at the output of LPF for interference signal with n = 2, 4, 8, 16 and CW signal, i.e. n = ∞. As shown here, interference radar signal has shorter duration as the chirp period becomes longer. Fig. 8Open in figure viewerPowerPoint Interference radar signal in time domain for n > 0 Fig. 9 shows frequency spectrum in frequency domain for n > 0, which is obtained by Fourier transform of the received radar signal interfered by interference radar signal with chirp period of nT as shown in Fig. 8. Fig. 9Open in figure viewerPowerPoint Frequency spectrum in frequency domain for n > 0 Fig. 10 a shows the simulation results of the interference radar signal power as a function of chirp period, nT for n > 0. Reference, i.e. 0 dB, is the interference signal power for CW signal, and the analysis results given by (8) are also shown in Fig. 10 a. It has been confirmed that simulation results completely agree to analytical results. When n = 2, the interference radar signal power increases by about 2.5 [dB] compared with that for CW signal, and it decreases as the chirp period, nT becomes long when n > 0. When n < 0, the interference radar signal power increases as the chirp period, nT increases i.e. n decreases from −2 to −16, as shown in Fig. 10 b. It has been confirmed that simulation results agree to analytical results as well when n < 0. These results suggest that the chirp period, nT, should be longer when two fast chirp FMCW radars with the same chirp direction co-exist, and the chirp period, nT should be shorter when two fast chirp FMCW radars with the opposite chirp direction co-exist. Furthermore, it should be noted that no wideband interference occurs and only narrowband interference needs to be taken into account when multiple fast chirp FMCW radars with the same chirp period and chirp rate co-exist. Fig. 10Open in figure viewerPowerPoint Interference radar signal power as a function of chirp period, nT (a) n > 0, (b) n < 0 5 Conclusion This paper describes inter-radar interference analysis of fast chirp FMCW radars with different chirp periods to estimate SNR degradation caused by wideband interference. As the analysis results agree well to the simulation results, validity of the analysis described in this paper has been confirmed by computer simulation. Furthermore, it is revealed that inter-radar interference becomes significant when the chirp rate of interference radar is similar to that of interference radar though the inter-radar interference occurrence probability may not be very large. The inter-radar interference analysis results described here suggest that standardisation of FMCW radars is desirable to mitigate wideband interference, especially in terms of chirp rate, or chirp duration when its chirp frequency bandwidth is the same. The analytical results will be useful to consider the parameters of fast chirp radar for standardisation to mitigate inter-radar interference. They can be used to estimate the interference radar signal power in the actual inter-radar interference environments for FMCW radars co-existence study. 6 Acknowledgments This research and development work was supported by the MIC/SCOPE # 175003004. The authors also appreciate continuous support for this work by Hitachi Automotive Systems, Ltd, Japan. 7 References 1Makino Y., Nozawa T., Umehira M. et al.: ' Co-channel interference among FMCW radars using various chirp rates' ( The Institute of Electronics, Information and Communication Engineers, Japan), Technical Report, SANE2017-96, Jan. 2018 (In Japanese) Google Scholar 2Aoki K.: ' Current activities and some issues on the development of automated driving', J. Inf. Process. Inf. Manage., 2017, 60, (4), pp. 229 – 239, (In Japanese) Google Scholar 3Goppelt M., Blocher H.-L., Menzel W.: ' Analytical investigation of mutual interference between automotive FMCW radar sensors'. Proc. of the 6th German Microwave Conf., Darmstadt, Germany, 2011Google Scholar 4Luo T.-N., Wu C.-H.E., Chen Y.-J.E.: ' A 77-GHz CMOS automotive radar transceiver with anti-interference function', IEEE Trans. Circuits Syst., 2013–12, 80, (12), pp. 3247 – 3255CrossrefGoogle Scholar 5Choi J.-H., Lee H.-B., Choi J.-W. et al.: ' Mutual interference suppression using clipping and weighted-envelope normalization for automotive FMCW radar systems', IEICE Trans. Commun., 2016, E99-B, (1), pp. 280 – 287CrossrefWeb of Science®Google Scholar 6Nozawa T., Makino Y., Takaya N. et al.: ' An anti-collision automotive FMCW radar using time-domain interference detection and suppression'. paper 1a.1, Proc. of IET Radar 2017 (Int. Conf. on Radar Systems 2017), Belfast, UK, 23–26 October 2017Google Scholar 7Lutz S., Ellenrieder D., Walter T.: ' On fast chirp modulations and compressed sensing for automotive radar applications'. 2014 15th International Radar Symp., Gdansk, Poland, June 2014Google Scholar 8Skolnik M.I.: ' Introduction to radar systems' ( McGraw-Hill Book Company, New York, NY, USA, 1981), pp. 81 – 91Google Scholar Citing Literature Volume2019, Issue19October 2019Pages 5634-5638 FiguresReferencesRelatedInformation
Referência(s)