Location prediction optimisation in WSNs using Kriging interpolation

Many wireless sensor network (WSN) applications rely on precise location or distance information. Despite the potentials of WSNs, efficient location prediction is one of the subsisting challenges. This paper presents novel prediction algorithms based on a Kriging interpolation technique. Given that each sensor is aware of its location only, the aims of this work are to accurately predict the temperature at uncovered areas and estimate positions of heat sources. By taking few measurements within the field of interest and by using Kriging interpolation to iteratively enhance predictions of temperature and location of heat sources in uncovered regions, degree of accuracy is significantly improved. Following a range of independent Monte Carlo runs in different experiments, it is shown through a comparative analysis that proposed algorithm delivers approximately 98% prediction accuracy.