ITU-R REPORT M 2128-2008 Test results and simulations illustrating the effective duty cycle of frequency modulated pulsed radiolocation and EESS system waveforms in marine radionav.pdf

《ITU-R REPORT M 2128-2008 Test results and simulations illustrating the effective duty cycle of frequency modulated pulsed radiolocation and EESS system waveforms in marine radionav.pdf》由会员分享，可在线阅读，更多相关《ITU-R REPORT M 2128-2008 Test results and simulations illustrating the effective duty cycle of frequency modulated pulsed radiolocation and EESS system waveforms in marine radionav.pdf（18页珍藏版）》请在麦多课文档分享上搜索。

1、 Rep. ITU-R M.2128 1 REPORT ITU-R M.2128*Test results and simulations illustrating the effective duty cycle of frequency modulated pulsed radiolocation and EESS system waveforms in marine radionavigation receivers (2008) 1 Introduction This Report was developed based on measurements that were perfor

2、med with marine radionavigation radars operating at 9 410 MHz and 3 050 MHz. Typical waveforms of chirped radiolocation and earth exploration satellite service (EESS active) systems were simulated using test equipment and injected into the 9 410 MHz radar receiver to investigate how the signals duty

3、 cycle and pulse width are altered from the transmitted RF pulse, to the one that is presented to the radars detector/processor. For the 3 050 MHz radars, mathematic simulations were performed to determine the response of the radars to the EESS and radiolocation systems. 2 Background This Report was

4、 prepared to provide information on test results for possible consideration of World Radiocommunication Conference 2007 Agenda item 1.3. 3 Test signals Table 1 shows the parameters of the radiolocation waveforms. They were developed based on the characteristics of Radars A7 and A3 from Recommendatio

5、n ITU-R M.1796 Characteristics of and protection criteria for terrestrial radars operating in the radiodetermination service in the frequency band 8 500-10 500 MHz. As a baseline signal, unmodulated pulses with a width of 1s were also generated. TABLE 1 Radiolocation system waveforms System waveform

6、 No. Pulse width (s) Pulse repetition frequency (Prf)(Hz) Pulse repetition interval (Pri) (ms) Duty cycle (%) Chirp (MHz) Chirp rate (MHz/s) Radiolocation 1 10 750 1.3 0.8 10 1 Radiolocation 2 10 750 1.3 0.8 50 5 Radiolocation 3 13.6/1.65 5 000 0.20 0.8 660/80 48.5 The victims receiver IF output res

7、ponse (amplitude and pulse width) to interference from chirped pulses is a function of the rate at which the chirped frequency sweeps through the victim radar receiver passband. This rate, called chirp rate, Rc, is given by: *This Report should be brought to the attention of Radiocommunication Study

8、 Group 7. 2 Rep. ITU-R M.2128 Rc= (Bc /t) where: Rc: sweep rate (MHz/s) Bc: chirp frequency range (MHz) t: pulse duration (s). Victim radar receivers should not respond to interference on frequencies outside the 20 dB points passband of their IF circuitry, assuming that the amplitude of the interfer

9、ence is below the front-end overload threshold of the radar receiver RF front end. In some cases, the frequency sweep range of the chirp-pulse generation system used in these tests was limited by hardware to less than the full chirp range of the corresponding radar emission specified in Recommendati

10、on ITU-R M.1796. In such cases, the tests were still performed to fully and accurately replicate the response of radar receivers to the specified chirp parameters. To accomplish this goal, the chirped pulses used in the tests were swept across at least twice the 20 dB frequency response range of the

11、 victim radar receivers, at the same rate as the sometimes wider-bandwidth chirp pulses from potentially interfering sources. For example in Table 1, the 660 MHz chirp in a 13.6 s pulse (Rc= (660 MHz/13.6 s) = 48.5 MHz/s) is not possible to generate with the test equipment. An equivalent interferenc

12、e effect can be generated with an 80 MHz chirp pulse in an interval of 1.65 s (Rc= (80 MHz/1.65 s) = 48.5 MHz/s), provided that the 20 dB radar IF passband of the victim is equal to or less than 50 MHz wide. The EESS system waveform characteristics are shown below in Table 2. As in the case of the c

13、hirped waveforms from Table 1, the values are scaled to the maximum 80 MHz chirp bandwidth of the test equipment. The duty cycles are calculated using the scaled pulse widths. TABLE 2 EESS system waveform characteristics System waveform No. Pulse width (s) Scaled width (s) Prf (Hz) Pri (ms) Duty cyc

14、le (%) Chirp (MHz) Chirp rate (MHz/s) EESS 1 10 2 2000 0.5 0.4 400/80 40 EESS 2 80 16 4500 0.22 7.2 400/80 5 EESS 3 10 17.7 515 1.94 0.91 460/80 4.6 EESS 4 10 1.7 5150 1.94 0.88 460/80 46 In the tests described in this Report, the value of Rcwas always preserved and the victim radar receivers always

15、 saw the chirped interference across their full receiver IF passbands in exactly the same way as they would have if the chirped interference had been generated across wider bandwidths. That is the key element in accessing the effects of the interference and measuring the effective duty cycle. 4 Meas

16、urement technique The pulses were injected into the radar at the nominal frequency of 9 410 MHz at the low noise amplifier (LNA) input of the radar receiver. The radar was not connected to its antenna, so no other Rep. ITU-R M.2128 3 signals were able to get into the receiver. A test point was locat

17、ed on the IF circuit card and a spectrum analyser was set to zero-span mode with a resolution bandwidth consummate with the radar receiver and connected to that point to measure the response of the radar to the radiolocation and EESS systems. The radar was placed on stand-by mode so that the receive

18、r was activated, but its transmitter was not generating pulses. The analyser was used to collect data that was plotted to show the power of the pulses versus time at the fundamental frequency. The radar uses a summing multistage logarithmic amplifier. A test point was provided that is located at the

19、 output of the third amplifier. A CW signal was swept in frequency from 9 370 to 9 450 MHz to determine the response of the receiver and measure the IF bandwidth. The result is shown below in Fig. 1. The 3 dB IF bandwidth of the radar when set to short pulse mode 1, which uses a pulse width of 200 n

20、s for a maximum range of 3 NM, was measured to be about 6 MHz. Note that there is a spurious response in the receiver at 9 381 MHz 20 dB down from the peak response at 9 410 MHz. Note that these measurements were not done in a manner to verify frequency dependent rejection (FDR) values. The radioloc

21、ation and EESS system input powers and corresponding output powers at the radars IF were not calibrated to perform that measurement. These measurements were only done in a manner to show how the pulse width was reduced due to the signal chirping through the radar receiver, not the peak power. The ef

22、fective pulse width is defined in this Report as the width of the EESS and radiolocation chirped pulse that is presented to the radars target detector/processor after it has passed through the radars LNA and been convolved with the IF filter. For example, if the width of the transmitted EESS and rad

23、iolocation pulse is 10 s, but at the radars IF output it is 2 s, then the effective pulse width is 2 s. 4 Rep. ITU-R M.2128 5 Results The results of the measurements are shown in Figs. 2 through 9. The figures are annotated and show the width of the pulse after it has passed through the receivers LN

24、A and IF circuitry as it is presented to the receivers processor and detector. Note that the figures show power versus time, so for these measurements the 6 dB points are used to determine the pulse width. Fig. 2 shows that the unmodulated pulse (the baseline signal) was 1 s wide in the radars IF ba

25、ndwidth, which is the same value as the transmitted pulse. However, for the chirped radiolocation and EESS systems, the width of pulses as seen in the IF passband of the receiver are shorter than the ones that were transmitted at the RF level. Since the pulse repetition interval (pri) has not change

26、d, the effective duty cycle has been lowered as well. Figures 2 through 9 show the results using radiolocation and EESS systems that are chirped as described in 3 of this Report. Table 3 summarizes the results of the differences in pulse width between the RF transmitted pulses and the pulses that ar

27、e presented to the detector/processor of the radar receiver. The percentage difference was calculated by dividing the received pulse width by the transmitted pulse width and then multiplying by 100. Rep. ITU-R M.2128 5 6 Rep. ITU-R M.2128 Rep. ITU-R M.2128 7 8 Rep. ITU-R M.2128 TABLE 3 System wavefo

28、rm Transmitted pulse width (s) Pulse width (s) at detector/processor Percentage difference Radiolocation 1 10 5.1 51 Radiolocation 2 10 1.0 10 Radiolocation 3 1.65 0.20 12 EESS 1 2 0.20 10 EESS 2 16 1.0 6.3 EESS 3 17.7 1.2 6.8 EESS 4 1.7 0.20 11.8 6 Simulations The simulations were performed in Simu

29、linkTM, a model-based design package available as a toolbox in MATLABTM. A picture of the model is shown in Fig. 10. The left portion of the model creates the chirped signals. A repeating sequence and discrete-time voltage controlled oscillator (VCO) are used to assure coherence from pulse to pulse.

30、 The centre frequency changed for each filter, but the structure of the signal, bandwidth, and pulse width remained as described in the tables. The waveforms for each filter are otherwise identical to the ones described in Tables 1 and 2. The FDA tool block is the IF filter for a given radar, and ca

31、n be easily changed to simulate different systems. The filters used in this simulation were created from actual spectrum analyser data. The first filter tested was based on that shown in Fig. 1 (although shifted down in frequency to 60 MHz to speed up processing), to compare theoretical results with

32、 measured results. The measured IF Rep. ITU-R M.2128 9 filters from three other radars were also used in the simulation. The right-hand portion of the model is used for plotting the data on screen (in power versus time format) and saving it to a MATLABTMfile for later processing. The power versus ti

33、me format was used since this is the format of the data collected from the spectrum analyser. It must be noted that by using the discrete-time VCO, the high frequency edge of each pulse was sometimes rounded off due to lack of frequency resolution. This only caused minor errors in the results. The s

34、imulation included 4 pulses of each waveform, and the second pulse was used for comparison. This allowed the simulation to not have any zeros once noise was added (at the start or end of the simulation) and prevented large negative values once power was calculated. The simulated responses of IF filt

35、er 1 to the EESS and radiolocation systems are shown in Table 4. It must be noted that the filter was shifted down to a 60 MHz centre frequency for this simulation. Table 4 shows that the simulated 6 dB pulse width has a difference of less than 0.85 s for each waveform, and that the simulation overe

36、stimates the measured pulse width in each case. Plots for the time domain pulses are shown in Figs. 11 through 17, with power versus time. The simulated results show more detail in the time domain than the actual measurements, which are taken as zero-span spectrum analyser data in a measurement band

37、width. For example, in Fig. 15 (simulated filtered pulse), the spurious signal indicated in Fig. 1 (IF filter shape) is clearly visible, as is the overall IF filter shape. However, in the actual measurement, shown in Fig. 7, this spurious signal is not present. These results show that the simulation

38、 leads to a slightly pessimistic prediction (larger pulse width than actually measured) of how the actual IF filter would operate with the test waveforms. Therefore the simulation tool can predict a good worst-case estimate of the EESS and radiolocation system pulse width when no actual measurements

39、 are taken. 10 Rep. ITU-R M.2128 TABLE 4 Measured versus simulated 6 dB pulse widths for IF filter 1 (Fig. 1) System waveform Measured pulse width (s) at detector/processor Simulated pulse width (s) at detector/processor Difference (s) Radiolocation 1 5.1 5.95 0.85 Radiolocation 2 1.0 1.24 0.24 Radi

40、olocation 3 0.20 0.22 0.02 EESS 1 0.20 0.24 0.04 EESS 2 1.0 1.24 0.24 EESS 3 1.2 1.37 0.17 EESS 4 0.20 0.22 0.02 Rep. ITU-R M.2128 11 12 Rep. ITU-R M.2128 Rep. ITU-R M.2128 13 14 Rep. ITU-R M.2128 For simulation only with the EESS and radiolocation systems, measured IF filter responses for three oth

41、er radars were available. These IF filters, shown in Figs. 18 though 20 represent two marine radionavigation radars operating in the band 2 900-3 100 MHz one operating in the 10 GHz band, respectively. The centre frequency of each filter has been shifted down to accommodate the simulation. The resul

42、ts of the simulation for each IF filter with the EESS and radiolocation systems are summarized in Table 5, and one sample plot for each filters response is shown in Figs. 21 through 23. Rep. ITU-R M.2128 15 16 Rep. ITU-R M.2128 TABLE 5 Simulated 6 dB pulse widths for IF filters 2 through 4 System wa

43、veform IF filter 2 Simulated pulse width at detector/processor (s) IF filter 3 Simulated pulse width at detector/processor (s) IF filter 4 Simulated pulse width at detector/processor (s) Radiolocation 1 1.24 NA(1)NA(1)Radiolocation 2 0.73 0.62 0.88 Radiolocation 3 0.26 0.29 0.81 EESS 1 0.27 0.30 0.7

44、8 EESS 2 0.73 0.68 0.88 EESS 3 0.78 0.72 0.90 EESS 4 0.26 0.29 0.82 (1)This filter reduced the waveform amplitude to the noise level, and showed no pulse shape except at the beginning and end of the pulse. Therefore, no estimate of pulse width was available. In all cases, the simulated pulse width a

45、t the radar detector/processor is less than the width of the transmitted pulse. As in the case of IF filter 1, this reduces the effective duty cycle of the EESS and radiolocation systems, which allows their interference reduction circuitry/processing to better mitigate their effects. Rep. ITU-R M.21

46、28 17 18 Rep. ITU-R M.2128 7 Conclusions The measured results show that due to the chirping of the radiolocation and EESS systems and the response of the radar receiver, the effective duty cycles and pulse width of the waveforms at the radars detector/processor input have been reduced to a value muc

47、h lower than the transmitted waveforms. The simulated results show that the MATLABTMbased simulation shows good agreement with the measured results of effective pulse width. For EMC analyses, this allows the effective pulse width of a chirped signal in a radar receivers IF circuitry to be accurately predicted. The model allows for actual IF data or user-designed filters, to be used as inputs so that radar designers can evaluate the performance of their interference rejection capabilities in the presence of these types of waveforms.