LCSW: A Novel Indoor Localization System Based on CNN-SVM Model with WKNN in Wi-Fi Environments

Xuhui Wang,Xin Deng,Hao Zhang,Kai Liu,Penglin Dai

DOI: https://doi.org/10.1007/978-981-19-6135-9_13

2022-01-01

Abstract:In this paper, we propose a novel indoor localization system, which fuses convolutional neural network (CNN) and support vector machine (SVM) model with an upgraded weighted K-nearest neighbor (WKNN) algorithm, called LCSW, to enhance the localization accuracy and robustness of the system. To this end, we propose a two-layer localization scheme. Specifically, in the first locating layer, we primarily partition the whole environment into certain subareas, then continuously collect the sequence data of RSSI in different time and reshape the data format as square matrix to serve as input of the modified CNN-SVM model to locate the target to a subarea. Then, in the second locating layer, the improved WKNN is used to calculate the precise location of the target in the corresponding subarea, which adopts the variance characteristic of Wi-Fi signal to assist the calculation of weights in measuring the distance and the cosine similarity to assist the assignment of weights in computing the coordinate, respectively. Finally, extensive real-world experiments are conducted to demonstrate the effectiveness of the proposed methods.