Dynamically Adjusted Genetic and Fireworks Algorithm for Deployment Optimization of Indoor Positioning Signal Sources

The performance of indoor positioning systems is affected by spatial deployment of signal sources, thus we presented a combined method of adjusted Genetic algorithm (AGA) and Fireworks algorithm (FWA) to provide the optimal deployment solution. The threshold value of fitness function is dynamically adjusted in GA, which reduces the probability of falling into local optimum and decreases the number of trivial solutions. With fireworks of fine chromosomes, the number of explosion sparks and radius of all explosion sparks are set as explosion factors. Through explosion and random generation, supplemented individuals are produced to increase population diversity and guarantee individual quality. Population evolves to the next generation by crossover and mutation, and the optimal result from evolutions is one solution of deployment with a number of signal sources and their locations. Based on a set of experiments, the AGA&FWA algorithm has been proved with good performance of convergence ability and computing expense. AGA&FWA has the best searching ability compared with the existing optimization methods, considering only the positioning error, or together with the cost of iBeacons. Our proposed AGA&FWA approach thus can provide the optimal signal sources' deployment for indoor positioning.