Research On Indoor Wifi Location Algorithm Based On Heuristic Clustering

Heuristic clustering algorithm is a research hotspot in data mining and machine learning.It is widely used in data mining,wireless sensor networks and other fields.With the development of mobile Internet,indoor WIFI location technology has been widely used,among which WIFI location technology based on fingerprint database is the most widely used.In order to obtain more accurate location,many researchers use heuristic clustering algorithm to obtain more accurate location information.The main research contents of this paper are as follows:First of all,through a large number of papers in this field,and experimental simulation,compared with the traditional WIFI positioning algorithm has been proposed,found that most of the researchers in the establishment of fingerprint database,are in a continuous area to collect fingerprints,it takes too much time.In the phase of fingerprint data collection,POI fingerprint database was established in this paper to divide the regions to be located,and each POI region was established with its own fingerprint information,which avoided some redundant fingerprint collection work and improved the positioning efficiency.Secondly,traditional heuristic clustering algorithm has the advantages of fast convergence and easy implementation,but the sensitivity of initial solution seriously affects the quality of clustering algorithm.In this paper,FSG MHC(Fireworks Search Guided multi-way Heuristic Clustering algorithm)is proposed.The classical heuristic clustering algorithm was first called P times to generate P local optimal solutions.Then,in the search space,P local optimal solutions are taken as the search starting point,and fireworks search is adopted for multipath search,and a new search starting point is obtained by convergence of fireworks algorithm.Finally,the new search starting point is used to invoke the classical heuristic clustering algorithm to obtain high-quality clustering results.The specific fireworks search step is as follows:for the given pEep path search,the fireworks selection operator is designed based on the entropy concentration to determine the search direction;Then the local optimal center point is transformed by mutation,mapping and migration operator to find a better search starting point.Finally,through experimental analysis,the clustering quality of FSG_MHC algorithm on four different data sets is significantly higher than that of the comparison clustering algorithm.In addition,the error analysis of the positioning system shows that the average positioning error is 1.730 meters,which is obviously better than the positioning result of the comparison algorithm.