A CSI Correction Localization Method Combined with Densely Connected Networks
A technology for connecting networks and positioning methods, applied in the field of indoor positioning technology and data analysis, can solve the problems of poor positioning stability, low positioning accuracy, and large workload of fingerprint collection, so as to increase dissimilarity, enhance robustness, and improve positioning. performance effect
Active Publication Date: 2022-05-31
HANGZHOU DIANZI UNIV
View PDF4 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] Aiming at the problems of poor positioning stability, low positioning accuracy and heavy fingerprint collection workload in the prior art, the present invention proposes a CSI correction positioning method combined with a densely connected network, and introduces a CSI anomaly elimination method based on an isolated forest, a new CSI amplitude fingerprint form, improved CSI positioning densely connected network, generalized continuation interpolation method to expand fingerprint library, and improved KNN correction positioning algorithm combined with Bhattachary coefficient, etc., reduce the error of positioning algorithm and improve the overall positioning performance and stability
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0033] A densely connected network is a convolutional neural network that alleviates the vanishing gradient problem by stacking the output of the previous layer of the network with the input of the current layer in the channel dimension. On the one hand, the dense connection of the network not only improves the problem of gradient disappearance and model degradation, but also enhances the reuse of features, which is conducive to the transmission of channel state information between layers; on the other hand, the number of parameters required by the densely connected network Significantly less than traditional convolutional networks, with higher parameter efficiency. The method of using neural network to extract CSI features and using probability weighted positioning has the problems of large positioning error and poor positioning system robustness when the prediction probability is low. However, using the improved KNN correction positioning algorithm combined with the Babbitt c...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a CSI correction positioning method combined with a densely connected network, which includes using an isolated forest to eliminate abnormal CSI, constructing a CSI amplitude fingerprint form including time, frequency and antenna pair information, and training the improved CSI positioning dense Connect the convolutional network to establish the correspondence between CSI and spatial location, and use the generalized continuation interpolation method to construct the interpolation fingerprint library, use the probability weighted location estimation based on the neural network, and use the interpolation fingerprint library and the improved KNN combined with the Barthel coefficient The correction positioning method performs position correction for the prediction result with a lower maximum probability. The invention makes full use of the information contained in CSI; comprehensively excavates the potential characteristics of CSI to make the training process more efficient; combined with the improved KNN correction positioning algorithm of the Bhattachary coefficient, it effectively reduces the positioning error of the probability weighted positioning algorithm based on the neural network, and improves the efficiency of the training process. The stability of the positioning system.
Description
technical field [0001] The invention relates to the field of indoor positioning technology and data analysis technology, in particular, to a CSI correction positioning method combined with a densely connected network. Background technique [0002] A large number of indoor scenarios have conditions and requirements for location-based services. Since Wi-Fi devices are widely used indoors, some studies have used Channel State Information (CSI) that can be obtained by Wi-Fi for positioning. However, the current CSI-based indoor positioning algorithms do not maximize the potential of CSI in positioning, and there are still problems of insufficient data utilization and single positioning algorithm, which reduces the accuracy and robustness of CSI-based indoor positioning algorithms. SUMMARY OF THE INVENTION [0003] Aiming at the problems of poor positioning stability, low positioning accuracy and large workload of fingerprint collection in the prior art, the present invention p...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & AuthorityPatents(China)
IPC IPC(8): H04W4/02H04W4/33H04W4/021G06N3/04G06N3/08
CPCH04W4/023H04W4/33H04W4/021G06N3/084G06N3/045Y02D30/70
Inventor姚子扬尚俊娜施浒立
OwnerHANGZHOU DIANZI UNIV