A quantitative measure of integrity error

Integrity testing calls for evaluation criteria far different from measures used in judging navigation accuracy. Evaluation methods now planned could be improved, especially in regard to bias detection sensitivity near the Containment Warning Level. This investigation was prompted by conflicting demands for large data ensembles and practicality of tests, exacerbated by added requirements for high confidence levels. Complexity of sole means FDI sequences adds still more demands beyond those of integrity test for supplemental navigation. A new criterion of performance is offered as a solution. Present plans for GPS integrity testing-for both sole means and supplemental navigation-are envisioned without the recommendations herein. Existing concepts are based on GO/NO-GO testing throughout and, at least for sole means, there is one external flag (for availability; alarm flags are internal). Concepts herein are offered as potential solutions to unforeseen future events (e.g., tighter protection radius demands) or to marginal coverage (which has always exerted a limiting influence on GPS integrity)