Sigma Estimation, Inflation, and Monitoring In the LAAS Ground System

The prescribed Local Area Augmentation System (LAAS) algorithms for the generation of vertical and horizontal protection levels implicitly assume a zero-mean, normally distributed fault-free error model for the broadcast pseudorange corrections. While such a model appears to be consistent with reference receiver ranging errors due to thermal noise and diffuse multipath, remaining errors such as ground reflection multipath and systematic reference receiver/antenna errors may be non-gaussian, non-zero-mean, or both. To ensure navigation integrity, special care must therefore be taken on the ground in the definition of the broadcast correction error standard deviation ( gnd pr _ σ ). In this regard, this paper defines a new approach toward the estimation and inflation of the broadcast value of gnd pr _ σ to account for both ground reflection multipath errors as well as gaussian errors due to thermal noise and diffuse multipath. A sufficient condition for the acceptability of non-zero means is also presented. An additional requirement concerning the broadcast gnd pr _ σ is that it be possible to monitor for unexpected violations (failure events where, during operation, the true sigma exceeds gnd pr _ σ ) with acceptable residual integrity risk. The use of a ‘real-time’ sigma estimate as a monitor statistic is undesirable because the real-time estimate is slow to respond to actual sigma violations due