Fault tolerant three-dimensional cellular genetic algorithms with adaptive migration schemes

This paper presents a three-dimensional cellular genetic algorithm (3D-cGA) to achieve fault tolerance; in particular, faults occurred due to single event upsets (SEUs). Cellular genetic algorithms (cGAs) are targeted due to their high performance and inherent features, among others. Genetic diversity is the main factor in identifying and isolating faults occurred, particularly at individuals´ phenotypes. This type of failure could affect system functionality, and thus, fault tolerant design approaches at an algorithm level, such as the approach proposed, are targeted nowadays. We consider the GPS attitude determination as a case study to investigate the effectiveness of the proposed 3D-cGA. The overall results show success in handling up to 40% faults. In order to alleviate the deterioration in the performance of the algorithm, we propose a number of adaptive migration schemes. The different schemes are tested to show superior improvements in the algorithm´s performance in terms of efficiency, efficacy, and speed.