Development of a Real-Time Multipath MonitorBased on Multi-Correlator Observations*

摘要

So far, multi-correlator observations have primarily beenused for the detection of so-called “Evil Waveforms”,anomalous signals transmitted from a satellite. The firstobservation of such an evil waveform was reported in1993, after an anomalous behavior of SV19 had beendetected. The measurement of the power spectrum of thisanomalous signal revealed a large spike at the centerfrequency instead of showing a normal sinc functionshape. In 1994, the reason for this anomalous behaviorhas been identified as a failure in the satellite's signalgeneration and transmission hardware. The last years haveseen much effort to detect such signal failures with theconsequence that current satellite based augmentationsystems like LAAS or WAAS include Signal QualityMonitoring (SQM) concepts to protect airborne usersagainst such threats (e.g. [1] [4],[5] [6],[7]).The detection of satellite signal failures benefits from thefact that such failures result in distortions of the signal'scorrelation function. Therefore, one possible approach todetect such failures is to monitor the incoming signal bymeans of several correlators which are placed at distinctlocations along the correlation function. By means of thistechnique, the actual shape of the correlation function canbe compared to its “nominal” shape and deviations fromthe nominal shape can be assigned to signal failures. Inorder to be able to detect only the satellite signal failures,other influences that also distort the correlation function(mainly multipath) must be considered as well. This isusually done by means of calibration measurements at adistinct site (the future monitor station), where themultipath influences in this specific environment are“recorded” and the required monitor thresholds areadjusted accordingly. This method ensures that onlydistortions caused by signal failures are detected and thatfalse alarms caused by multipath signals are minimized.Since both satellite signal failures and multipath result ina distorted correlation function, it should also be feasibleto use multi-correlator observations for the detection ofmultipath signals. Based on this idea, this paper providesa concept to use correlation peak observations fordetecting and monitoring multipath signals in real-timeand to provide the user with instant information whetheror not a satellite signal is affected by multipath.The general idea for the implementation of this real-timemultipath monitor is to constantly monitor the correlationfunction by means of several correlators, to combine thecorrelator outputs mathematically and to compare themwith a threshold that indicates whether or not thecombined correlation peak observations exceed theircorresponding noise level. If this threshold is exceeded, itis assumed that the incoming signal is not only affectedby noise but also by other interfering signals (mainly bymultipath, less frequently by signal failures).The paper starts with a description of the generalmonitoring concept. The second section deals with thequestion how the correlator outputs can be combined in