Research On Indoor Terminal Location Algorithms Based On Fruit Fly Optimization

Fruit fly is a kind of arthropod with sensitive smell and vision.Its functions of smell,vision and congenital immune response provide important biological inspirations for researchers to investigate fruit fly optimization algorithms.In addition,although fruit fly optimization is a novel way to solve indoor terminal node location problems,the related research works have been sparely reported.Therefore,for the problem of low positioning precision for unknown nodes in indoor environments,three fruit fly optimization algorithms are developed to solve ZigBee node positioning problems,relying upon the functions of fruit flies' smell,vision and immune response.Their computational complexity analyses and numerically comparative experiments are also carried out.The research works are helpful for both enhancing the positioning accuracy of indoor terminal nodes and accelerating the rapid development of fruit fly optimization.The main works and the acquired achievements are summarized as follows:A.For the problem of two-dimensional ZigBee node localization,a reported stochastic programming model on receiving signal intensity indication location is transformed into an unconstrained expected value programming one.It is solved by a new fruit fly optimization algorithm in stochastic environments which is developed based on the biological characteristics of fruit flies' smell,vision and collaborative foraging.The theoretically computational complexity analysis shows that the algorithm can execute evolution with a fast search speed.Numerically experimental results indicate that it,with high positioning precision and strong convergence,is available for RSSI location.B.For the problem of three-dimensional ZigBee node localization,a RSSI positioning model with noise is converted into a constrained expected value programming model.In order to solve such a model,a fruit fly immune coevolutionary optimization algorithm with is designed to search for the location of an unknown node in spational environments.Experimental results show that for the problem of spationally environmental node location,the algorithm is effective with high positioning precision and efficient search performance.C.In order to effectively come with the difficulty of parameter uncertainty presented in the path loss index of the logarithmic loss model for the problem of three-dimensional node positioning in stochastic environments,the average error function of signal intensity is taken as the objective function.An improved fruit fly immune optimization approach is designed to optimize the path loss index and the coordinates of the unknown node included in the logarithmic loss model.Numerical experiments illustrate that the algorithm can perform well with the merits of strong noise suppression and high precise positioning.