曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
spacing metrics, so as not to cause significant peak distortion itself and not to mask the anomalies caused by the satellite
failure. Typically, the SQM receiver must have a precorrelation bandwidth of at least 16 MHz for GPS and at least 15 MHz
for GLONASS.
8.9.4 The test metrics are smoothed using low-pass digital filters. The time constant of these filters are to be shorter
than those used jointly (and standardized at 100 seconds) by the reference receivers for deriving differential corrections and
by the aircraft receiver for smoothing pseudo-range measurements (using carrier smoothing). The smooth metrics are then
compared to thresholds. If any one of the thresholds is exceeded, an alarm is generated for that satellite.
8.9.5 The thresholds used to derive performance are defined as minimum detectable errors (MDEs) and minimum
detectable ratios (MDRs). Fault-free false detection probability and missed detection probability are used to derive MDEs
and MDRs. The noise in metrics (D) and (R), as denoted σD,test and σR,test below, is dominated by multipath errors. Note that
the metric test can also have a mean value (μtest) caused by SQM receiver filter distortion. Threshold tests must also account
for the mean values.
8.9.6 The MDE and MDR values used in the SQM performance simulations are calculated based on the following
equations:
MDE = (Kffd + Kmd) σD,test and
MDR = (Kffd + Kmd) σR,test
where
Kffd = 5.26 is a typical fault-free detection multiplier representing a false detection probability of 1.5 × 10–7 per test;
Kmd = 3.09 is a typical missed detection multiplier representing a missed detection probability of 10–3 per test;
σD,test is the standard deviation of measured values of difference test metric D; and
σR,test is the standard deviation of measured values of ratio test metric R.
8.9.7 If multiple independent SQM receivers are used to detect the failures, the sigma values can be reduced by the
square root of the number of independent monitors.
ATT D-45 23/11/06
Annex 10 — Aeronautical Communications Volume I
8.9.8 A failure is declared if
| D,test – μD,test | ≥ MDE or
| R,test – μR,test | ≥ MDR
for any of the tests performed, where μX,test is the mean value of the test X that accounts for fault-free SQM receiver filter
distortion, as well as correlation peak distortion peculiar to the specific C/A code PRN. (Not all C/A code correlation peaks
have the same slope. In a simulation environment, however, this PRN distortion can be ignored, and a perfect correlation
peak can be used, except for simulated filter distortion.)
8.10 The standard deviations of the test statistics, σD,test and σR,test can be determined via data collection on a
multicorrelator receiver in the expected operating environment. The data collection receiver utilizes a single tracking pair of
correlators and additional correlation function measurement points which are slaved to this tracking pair, as illustrated in
Figure D-9. Data is collected and smoothed for all available measurement points in order to compute the metrics. The
standard deviation of these metrics define σD,test. It is also possible to compute these one sigma test statistics if a multipath
model of the installation environment is available.
8.10.1 The resulting σD,test is highly dependent on the multipath environment in which the data are collected. The
deviation due to multipath can be an order of magnitude greater than that which would result from noise even at minimum
carrier-to-noise level. This aspect illustrates the importance of the antenna design and siting criteria which are the primary
factors in determining the level of multipath that will enter the receiver. Reducing multipath will significantly decrease the
resulting MDEs and thus improve the SQM capabilities.
8.10.2 Mean values μD,test and μR,test, on the other hand, are determined in a relatively error-free environment, such as
through the use of GPS and GLONASS signal simulator as input. These mean values model the nominal SQM receiver’s
filter distortion of the autocorrelation peak, including the effects of distortion due to adjacent minor autocorrelation peaks.
The mean values can differ for the various PRNs based on these properties.
8.11 In order for the ground monitor to protect users against the different threat models described above, it is necessary
to assume that aircraft receivers have specific characteristics. If no such constraints were assumed, the complexity of the
ground monitor would be unnecessarily high. Evolution in the technology may lead to improved detection capability in the
aircraft receiver and may alleviate the current constraints.
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：
附件10--航空电信an10_v1_6ed下(106)
