Signal Deformation Monitoring for Anomalous Multipath Threats

Abstract

The current WAAS signal quality monitor algorithm was designed to mitigate anomalous signal distortions. This paper assesses the ability of this monitor mitigate distortions produced from satellite-induced multipath. Leveraging experience from the SVN-49 anomaly, a single-reflection signal threat model is defined and expanded to include an elevation-angle dependence, which may potentially reduce observability of distortions viewed from widely-distributed monitor receivers. Next, the range errors for both singlefrequency and dual-frequency aviation users are modeled relative to the monitor’s ability to detect them. It is shown that the existing WAAS signal deformation monitor can protect both single and dualfrequency aviation users against a wide range of SVinduced, single-reflection multipath parameters despite significant attenuation of monitor sensitivity due to elevation-angle dependence. BACKGROUND The threat of anomalous signal deformations has existed for users of high-integrity differential GNSS navigation systems for many years. Developers of SBAS and GBAS, in particular, originally analyzed that event and proposed several types of anomalous distortions that, without monitoring and detection, could pose a hazard to aviation users. Subsequently, a threat model that encompassed that thinking was proposed and later adopted as the standard by ICAO in 2000 [1]. That threat model specifically identified two classes of anomalous deformations—digital and analog—to capture the general characteristics of distortions observed on the SV-19 fault. Further, the model was ultimately expanded and proposed as representative worst case for all anomalous signal deformation faults. Signal deformation monitors were subsequently developed to mitigate any/all SVinduced distortions using that ICAO threat model for validation. SVN-49 anomaly in 2009 was caused by an internal reflection in the signal payload; it resembled multipath. That anomaly is not considered a fault, however, because the satellite was never declared healthy. No WAAS users were ever at risk of exposure. Still, the validation threat model was proposed to account for general signal distortions and anomalous multipath is a specific type of distortion against which validated monitors may be assessed. The SVN-49 anomaly was also peculiar in that it had an elevation angle-dependence. (See Figure 1.) That potentially challenges detection capabilities for networks that observe the satellite from widelyseparated locations. Figure 1. L1 C/A chip shape measured for specific elevation angles for SVN-49 (PRN-01) as measured by an 18 MHz NovAtel receiver. [2] Previous work has broadly assessed the capability of the WAAS signal quality monitor to protect singlefrequency aviation users against the multipath threat [3]. However, relatively little has been done to address users of dual-frequency WAAS where range errors due to biases are larger while error bounds are reduced. In addition, to date, nothing has been done to account for the potential elevation-angle variations exhibited by anomalous multipath of the type observed on SVN49. That effect is potentially significant for WAAS monitoring, which relies on a wide network of reference stations distributed across North America. WAAS also traditionally presumes the observation is independent of elevation angle. This paper seeks to assess the existing WAAS signal deformation monitor capability to mitigate the elevation angle-dependent multipath threat for both single and dual-frequency WAAS users. ANALYSIS Single-Reflection Threat Model A simple signal-reflection model (with a reflected signal scale parameter R and delay offset, d) first was introduced in [3] as a model of this fault model against which to evaluate the WAAS signal deformation monitor. The equation for this threat model—a (single) reflection corrupted C/A code c(t)—is given below. c(t) = c (t) + R ⋅ c (t) = c (t) + R ⋅ c (t − d) (1) Traditionally, signal deformation monitor analyses presume the transmitted code c(t) affects the ground receivers and the user receivers equivalently. However, multipath threats may cause different receivers to experience different range errors as a function of the elevation angle of their lines of sight to the satellite. This has the potential to “blind” the monitor to the distortion while a user experiences its full effects. Accordingly, this effect is modeled as a reduction in the monitor measurements relative to those of the user. This effectively adds a third parameter to the two-parameter model above. Table 1. Parameter limits for the Single-Reflection (Multipath) Threat Model Parameter Range A -0.99 to 0.99 d (m) 0 to 100 m  -10 dB to 0 dB (0.1 to 1) The parameter ranges of the reflected code cMP(t)= R ⋅ c (t − d) and the monitor attenuation factor () are provided in Table 1. User Receiver Configurations The Minimum Operational Performance Standard (MOPS) version DO-229D describes the allowed receiver configuration for L1-only aviation users of WAAS [4]. These include constraints on discriminator type, correlator spacing, bandwidth, and precorrelation filter differential group-delay. Similar constraints for dual-frequency users are not yet finalized but have been proposed. The design space for these receivers will be far more limited in order to reduce the magnitude of the potential errors due to signal deformation threats of all kinds. The discriminator spacing and bandwidth constraints for the single-frequency and dual-frequency receivers modeled in this paper are illustrated in Figures 1 through 3. (For this paper,  receivers were not modeled for single-frequency users.) More details on these configurations are provided in Table 2. Figure 1. Current WAAS user receiver configurations for single-frequency (L1 C/A-code) EML users. 0 . 1