Design And Implementation Of WIFI Based Virtual Indoor Positioning And Navigation System

With the appearance of all kinds of LBS applications, location-based services have been applied more and more extensively, ranging from outside areas to complex indoor environments. In terms of indoor environments, buildings'architectures limit the usage of tradition positioning technology. Some new location algorithms occurred during the last decades which have brought huge economic benefits.Virtual location and navigation systems give the access to a new interaction. Although traditional 2D plane can tell us our position and orientation, it's not distinguishable enough to involve the user in the scene. Virtual reality systems based on 2D map create a real-like scene which would give users excellent experience in environment scene by immersive interactions. However,virtual 3D localization and navigation systems are only applied in some diverted traffic roads,and there are no maturing applications which can gain high performance.At the same time, more and more intelligent systems are going to be developed to interact with customers which based on indoor environment. In this paper, a WIFI-based virtual location and navigation system has been proposed which combined with mobile location and virtual reality, and multi-clients can communicate with the others and can also do data synchronization. Experiments and results have also been illustrated in this paper.The convenience and high accuracy make WIFI-based Indoor positioning a hot topic today. Science the accuracy and stability of WIFI-based indoor positioning are not reliable enough, a new improved indoor location and navigation algorithm based on fingerprint matching was proposed according the comparison among traditional similarity algorithms. The proposed algorithm confirms the final fingerprint similarity by fuzzy inference which combined with cosine similarity, Pearson correlation coefficient and average Euclidean distance, and a visual 3D system was involved in to gain high experience. First of all, fingerprint data collection was completed in location areas before data process which could remove inferences, and a probability distribution of RSSI would be generated in a given indoor area and each labeled position has multiple fingerprint data which could gain high location accuracy. Then mobile devices would send location requests in real-time location phase. Representative similarities between request data and fingerprint data would contribute to final results, fuzzy logic is involved in to get the weights of above similarity algorithm. However, fingerprint data extraction should take latest results into consideration to reduce calculation. Finally, location results would be calculated by weighted triangulation of all candidate positions which have high similarity with requests. At the same time, the location result would be reflected in an visual 3D system integrated in mobile phones. The experimental results show that the improved matching algorithm has high representative characteristics, and the performance of the algorithm is more stable than the traditional similarity algorithms.The system indicates that the virtual localization and navigation system could improve the intelligence of interaction between human and indoor environments, and also provide great possibilities for further applications development which based on fingerprint localization and navigation.