Performance analysis and improved method for single epoch, single frequency attitude determination using GPS

The baseline constrained LAMBDA method is an effective algorithm for attitude determination using GPS signal, but the performance and success rate for single epoch, single frequency applications need to be improved. To deal with this problem, the performance of original algorithm was studied through the analysis of the factors affecting the search efficiency. A new model in which the unknown integer ambiguities are treated as noise was given for a further improvement of the search efficiency, and at the same time the success rate was also improved with the constrained value of the integer ambiguity. The testing results show that the time consumption of the proposed algorithm decreases sharply over that of the original algorithm while the success rate of GPS attitude determination increases significantly.