Improved localization by mal-function Wi-Fi detection and DCT/quantization in disaster environments

In indoor environments, the KNN algorithm is useful for positioning method. The conventional KNN algorithm, however, is unsuitable in disaster scenarios because user devices receive distorted information. To solve this problem, we first recognize whether a disaster situation happens or not, and perform KNN algorithm after elimination of mal-function Wi-Fi(s). Also, we solve inherent large amount of data problem of the KNN algorithm by using DCT (Discrete Cosine Transform)/ Quantization compression method, and analyze the positioning accuracy and capacity reduction ratio of the proposed method through intensive simulations.