The effect on GNSS ambiguity resolution of using an approximate weight matrix.

Abstract

A theorem of Teunissen (1999) states that the integer least-squares estimator mazimizes the probability of correct integer estimation within the class of admissible integer estimators. Applied to the problem of GNSS ambiguity resolution, this implies that the largest success rate is obtained when the integer least-squares principle is used for estimating the integer carrier phase ambiguities. No other integer ambiguity estimator will produce higher success rates. In this contribution it will be shown how the success rate is affected when a too optimistic or a too pessimistic precision description of the ambiguities is used in the ambiguity estimation process.