Fault Tolerant and Adaptive GPS Attitude Determination System

In this paper a fault tolerant platform for the GPS attitude determination is proposed. The algorithm encompasses speed, adaptability and performance as its key objectives and also deals with single hard errors (SHE) from the fault tolerance perspective. The technique is based on the ambiguity function method (AFM) but overcomes restrictions and computational overheads incurred by existing techniques such as AFM. The adaptation of the GPS architecture is done by using fine-grained or cellular parallel genetic algorithm to compute more efficiently the attitude parameters in terms of speed and performance. The algorithm also has the ability to efficiently search the complex search space imposed by the problem in addition to being immune to cycle slips compared to other conventional methods.