1、 Rep. ITU-R M.2083 1 REPORT ITU-R M.2083 Level of unwanted emissions of mobile-satellite service feeder links operating in the bands 1 390-1 392 MHz (Earth-to-space) and 1 430-1 432 MHz (space-to-Earth) (2006) Scope This Report addresses techniques to control unwanted emission levels of mobile-satel
2、lite service (MSS) feeder links (Earth-to-space) that may operate in the band 1 390-1 392 MHz and MSS feeder links (space-to-Earth) that may operate in the band 1 430-1 432 MHz. Introduction The allocations to the fixed-satellite service (FSS) in the bands 1 390-1 392 MHz (Earth-to-space) and 1 430-
3、1 432 MHz (space-to-Earth) are limited to use by feeder links for non-geostationary-satellite networks in the MSS with service links below 1 GHz. The band 1 400-1 427 MHz is allocated to the Earth exploration-satellite service (EESS) (passive), radio astronomy and space research (passive) services o
4、n a primary basis in all Regions and that RR No. 5.340 also applies to the band 1 400-1 427 MHz. Unwanted emissions from FSS links may cause harmful interference to passive services requiring careful design of the transmission sub-system. The information provided in Annexes 1 to 3 of this Report sho
5、uld be taken into account for the control unwanted emission levels of MSS feeder links operating in the bands 1 390-1 392 MHz and 1 430-1 432 MHz. Baseband processing techniques without a specific post-amplifier filter, as described in Annex 1 for typical combinations of data rates and modulation te
6、chniques, should be used to reduce the unwanted emissions from MSS feeder links operating in the bands 1 390-1 392 MHz and 1 430-1 432 MHz into the band 1 400-1 427 MHz to the levels required for the protection of the passive services as defined in Recommendations ITU-R M.1747 and ITU-R M.1748. An a
7、dditional post-amplifier filter as described in Annex 3 should be used in cases where the baseband processing techniques are not sufficient to meet the levels required for the protection of the passive services in the band 1 400-1 427 MHz. 2 Rep. ITU-R M.2083 Annex 1 Evaluation of unwanted emissions
8、 in the 1 400-1 427 MHz band from non-GSO MSS feeder links that may operate in the 1 390-1 392 MHz and 1 430-1 432 MHz bands 1 Introduction This Annex provides power spectral-density (PSD) data generated via simulation for non-GSO MSS feeder-link transmitters that may operate in the 1 390-1 392 MHz
9、and 1 430-1 432 MHz bands. Modulation techniques considered here include offset quadrature phase shift keying (OQPSK), GMSK and 8-ary PSK (8-PSK) for channel bandwidths of 100 kHz, 300 kHz and 855 kHz. The simulation model used for this study was first validated against hardware measurement data, an
10、d then PSD data was generated via simulation for other modulation techniques and channel bandwidths. Two hardware fidelity grades were considered, low and high, in an effort to bound the expected performance of the non-GSO MSS feeder-link transmitters. 2 Technical characteristics of the MSS system T
11、he potential characteristics of non-GSO feeder links that may operate in the 1 390-1 392 MHz and 1 430-1 432 MHz bands are assumed to lie within the ranges given in Table 1. These characteristics are based primarily upon those given in Annex 2 of Recommendation ITU-R M.1184-2. TABLE 1 Potential tech
12、nical characteristics of MSS feeder links that may operate in the 1 390-1 392 MHz and 1 430-1 432 MHz bands* Link Parameter Value Band 1 390-1 392 MHz Channel bandwidth 30-855 kHz Modulation OQPSK, GMSK, 8-PSKData rate 4.8 kbit/s-3.42 Mbit/s(1) Gateway uplink (Earth-to-space) Transmit power 1-250 W
13、Band 1 430-1 432 MHz Channel bandwidth 30-855 kHz Modulation OQPSK, GMSK, 8PSKData rate 4.8 kbit/s-1.2825 Mbit/s(1) Gateway downlink (space-to-Earth) Transmit power 1-250 W * Except where noted, these characteristics are based upon the characteristics given in Annex 2 of Recommendation ITU-R M.1184-
14、2. (1)Range selected based upon data rates given in Annex 2 of Recommendation ITU-R M.1184-2 and based upon data rates which produce null-to-null bandwidths of 100 kHz, 300 kHz and 855 kHz for the modulation techniques considered in this study. Rep. ITU-R M.2083 3 3 Simulation model 3.1 Approach Fig
15、ure 1 provides the transmitter reference architecture assumed for this analysis. Figure 2 provides the power amplifier power out versus power in curve. The operating points assumed for this study are indicated on the plot. Figure 3 provides the power amplifier phase change versus power in curve. Aga
16、in, the operating points are shown. Figure 4 provides the duplexer filter attenuation characteristics assumed for this study. Also provided in Fig. 4 are the attenuation characteristics of example two-stage resonant cavity filters. The amount of duplexer attenuation was determined by assuming the MS
17、S feeder link receive band will need over 50 dB of protection from the local MSS feeder-link transmission with about 10 dB of this protection coming from isolation introduced by the return loss of the antenna and about 1 dB of loss through the circulator. Figure 5 provides a diagram of the duplexer.
18、 Table 2 describes the transmitter architecture characteristics assumed for this analysis. The use of a digital modulator with pre- or post-modulation digital filtering for the simulations is a valid assumption, as the highest modulation symbol rate considered in this study is 427.5 kbit/s (= 855/2
19、kbit/s), whereas digital modulators can be built to support modulation symbol rates up to at least 100 Mbit/s. Table 3 provides the transmitter distortion values assumed for this study. It should be noted that this study assumed that the baseband data provided to the modulator was random. If a patte
20、rn data is assumed, the results presented in this paper could change slightly. Although in general, pattern data will introduce modulation spurs in the PSD, these will likely be negligible given the large offset in frequencies between the passive service band and the proposed MSS feeder-link bands.
21、FIGURE 1 Assumed transmitter architecture 4 Rep. ITU-R M.2083 FIGURE 2 Assumed transmitter power amplifier Poutversus Pincharacteristics FIGURE 3 Assumed transmitter power amplifierchangeversus Pincharacteristics Rep. ITU-R M.2083 5 FIGURE 4 Assumed duplexer attenuation characteristics TABLE 2 Assum
22、ed transmitter architecture characteristics Value Component group Component Parameter Fidelity grade: lowFidelity grade: highComments Filter type Gaussian, raised cosine (RC) Gaussian, RC Filter can be optionally bypassed (e.g. for unfiltered case or post-modulation filtered case). Gaussian filter u
23、sed for GMSK; RC used for OQPSK; filter bypassed for 8-PSK and QAM which utilize post-modulation filtering Modulator Pre-modulation bit-shaping filter Filter implemen-tation 64-tap finite impulse response(FIR) 128-tap FIR Fewer taps can result in higher Tx output PSD levels especially beyond the fir
24、st few side lobes. Most hardware would likely use 128 taps, however, 64 taps considered here as worst case 6 Rep. ITU-R M.2083 TABLE 2 (continued) Value Component group Component Parameter Fidelity grade: lowFidelity grade: highComments Quantization 6-bit(1)12-bit(1) Fewer quantization bits can resu
25、lt in higher Tx output PSD levels especially beyond the first few side lobes. Actual hardware would likely use 8-bit or better, however, 6-bit assumed here as worst case Pre-modulation bit-shaping filter 3 dB bandwidth (one-sided) 50 kHz, 150 kHz,427.5 kHz50 kHz, 150 kHz,427.5 kHz Values respectivel
26、y apply to GMSK BTb = 0.5(2)and OQPSK BTs = 1.0(2)at 100 kbit/s, 300 kbit/s and 855 kbit/s Clock rate 1.8 MHz to22.23 MHz3.6 MHz to22.23 MHz Exact value dependent on data rate and modulation on the low end 3.6 MHz applies to 100 kbit/s OQPSK, etc. and on the high end 22.23 MHz applies to 855 kbit/s
27、GMSK Quantization 6-bit(1)12-bit(1) Fewer quantization bits can result in higher Tx output PSD levels especially beyond the first few side lobes. Actual hardware would likely use 8-bit or better, however, 6-bit assumed here as worst case LO frequency Not assumed Not assumed Modulator Output frequenc
28、y Not assumed Not assumed Simulation model uses equivalent low-pass signal representation (distortions of frequency translation modelled but not the actual frequency translation), therefore, exact definition of LO and IF frequencies not necessary Filter type RC RC Filter can be optionally bypassed (
29、e.g. for unfiltered case or pre-modulation filtered case). RC used for 8-PSK and QAM; filter bypassed for GMSK and OQPSK Modulator Post-modulation bit-shaping filter Filter implement-ation 64-tap FIR 128-tap FIR Actual hardware would likely use 128 tap or greater, however, 64 tap considered here as
30、worst case. Fewer taps can result in higher Tx output PSD levels especially beyond the first few side lobes Rep. ITU-R M.2083 7 TABLE 2 (continued) Value Component group Component Parameter Fidelity grade: lowFidelity grade: highComments Quantization 6-bit(1)12-bit(1) Fewer quantization bits can res
31、ult in higher Tx output PSD levels especially beyond the first few side lobes. Actual hardware would likely use 8-bit or better, however, 6-bit assumed here as worst case Modulator Post-modulation bit-shaping filter 3 dB bandwidth (at IF) 100 kHz,300 kHz,855 kHz 100 kHz,300 kHz,855 kHz Values respec
32、tively apply to 8-PSK 150 kbit/s, 450 kbit/s and 1.2825 Mbit/s. Values selected to produce BTs = 1.0(2)Filter type Generic Generic Filter implement-ation 64-tap FIR 128-tap FIR Fewer taps can result in a slower filter roll-off and, therefore, a higher Tx output PSD envelope Quantization 12-bit(1)16-
33、bit(1) Actual hardware would likely use 12- or 16-bit 3 dB bandwidth 500 kHz to 4.275 MHz500 kHz to 4.275 MHzInput filter or image rejection filter Roll-off 0.020 dB/kHz to 0.022 dB/kHz 0.020 dB/kHz to 0.022 dB/kHz Exact values dependent on data rate, modulation, modulator clock rate and DAC clock r
34、ate. Introduces modest filtering. Useful for limiting out-of-band emissions with little to no impact to BER Clock rate 21.6 MHz to 44.46 MHz21.6 MHz to 44.46 MHz Exact values dependent on modulation and data rate Digital-to-analogue converter Quantization 12-bit(1)16-bit(1) Actual hardware would lik
35、ely use 12-bit or better. While the PSD envelope/floor can be driven by quantization in the DAC, DAC output filter generally precludes this from occurring 3 dB bandwidth 2 MHz 2 MHz Based upon comments by MSS feeder-link hardware developers Digital-to-analogue converter (DAC) Output smoothing filter
36、 Roll-off 25 dB/MHz 25 dB/MHz 8th order Butterworth filter was used 8 Rep. ITU-R M.2083 TABLE 2 (continued) Value Component group Component Parameter Fidelity grade: lowFidelity grade: highComments IF input frequency Not assumed Not assumed LO frequency Not assumed Not assumed Simulation model uses
37、equivalent lowpass signal representation (distortions of frequency translation modelled but not the actual frequency translation), therefore, exact definition of LO and IF frequencies not necessary Up-converterRF output frequency 1 390-1 392 MHz 1 430-1 432 MHz 1 390-1 392 MHz 1 430-1 432 MHz Define
38、d by frequency bands considered for allocation to non-GSO MSS feeder links 3 dB bandwidth Bypassed Bypassed Up-converter Filter Roll-off Bypassed Bypassed Filter bypassed to ensure conservative results. Up-converter model does not produce unwanted image because baseband equivalent signal representat
39、ion used, therefore, filter not required in simulation model Amplifier Type TWTA TWTA Noise figure discussed in Table 3 3 dB bandwidth Bypassed Bypassed Pre-amplifier Filter Roll-off Bypassed Bypassed Filter bypassed to ensure conservative results. Transmitter designers must weigh the benefits of sp
40、ectral containment against the insertion loss Amplifier Type Solid-stateamplifier Solid-stateamplifier Modelled after 50 W TWTA. Noise figure discussed in Table 3 3 dB bandwidth Bypassed Bypassed Power -amplifier Filter Roll-off Bypassed Bypassed Filter bypassed to ensure conservative results. Trans
41、mitter designers must weigh the benefits of spectral containment against the insertion loss Type Resonant cavity Resonant cavity 3 dB bandwidth Bypassed Bypassed RF output filter Roll-off Bypassed Bypassed Filter bypassed to ensure conservative results. Transmitter designers must weigh the benefits
42、of spectral containment against the insertion loss Rep. ITU-R M.2083 9 TABLE 2 (end) Value Component group Component Parameter Fidelity grade: lowFidelity grade: highComments Type Resonant cavity Resonant cavity 3 dB bandwidth 2 MHz 2 MHz Duplexer (transmit path) Roll-off 1 dB/ MHz 1 dB/ MHz Duplexe
43、r required to protect the feeder-link receive band from the feeder-link transmit band. See Fig. 4 for plot of assumed duplexer attenuation characteristics. Duplexer attenuation characteristics determined by assuming the MSS feeder-link receive band will need over 50 dB of protection from the local M
44、SS feeder-link transmission with about 11 dB of this protection coming from isolation introduced by the antenna and circulator components and the rest required to come from the duplexer. Duplexer can have a large impact on the Tx PSD envelope/floor (1)Clip levels set near optimally. Similar PSDs wil
45、l result if more quantization bits are used but clip levels not set optimally. (2)Where B is the one-sided bandwidth of the filter, Tbis the bit duration when data is in a single data stream and Tsis the modulation symbol duration. TABLE 3 Assumed transmitter distortions(1) Value Parameter Low fidel
46、ity High fidelity Comments Data asymmetry 1% 0% Data asymmetry is present when one data bit polarity state is longer in duration than the other data bit polarity state it is effectively the fixed error component of data bit jitter. Data asymmetry introduces spikes in the Tx output PSD. Digital modul
47、ators typically have no data asymmetry, however, 1% is considered here as a worst-case amount. This study assumed data asymmetry was not applicable to GMSK and 8-PSK 10 Rep. ITU-R M.2083 TABLE 3 (continued) Value Parameter Low fidelity High fidelity Comments I/Q data skew 1% 0% I/Q data skew is the
48、variation from the ideal time delay between the I channel data transitions and the Q channel data transitions. I/Q data skew makes PSD more susceptible to spectral regrowth following a nonlinearity (such as the PA) as compared to ideal OQPSK. This study assumed I/Q data skew was not applicable to GM
49、SK and 8-PSK Modulator gain imbalance, peak 0.6 dB 0.6 dB Gain imbalance is present when one or more of the modulation vector magnitudes differ from the ideal value. Value dependent on modulation, bit-shaping, quantization and quantization clip level setting Modulator phase imbalance, peak 6.0 5.3 Phase imbalance is present when one or more of the modulation vector phases differ from the ideal value. Value dependent on modulation, bit-shaping, quantization an
