CHISEL: Compression-Aware High-Accuracy Embedded Indoor Localization With Deep Learning

Wang, Liping; Tiku, Saideep; Pasricha, Sudeep

doi:10.1109/les.2021.3094965

Cited by 26 publications

(9 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method can lead to better results, as it is very likely that data taken at almost the same time can be in both the training and test sets. On the contrary, in [80], the authors use the validation component (1111 samples) of the dataset as the test set. The training portion of the dataset is split into training (15,950 samples) and validation (3987 samples) subsets based on an 80:20 split.…”

Section: Ujiindoor Results Analysismentioning

confidence: 99%

Comprehensive Analysis of Applied Machine Learning in Indoor Positioning Based on Wi-Fi: An Extended Systematic Review

Bellavista-Parent

Torres-Sospedra

Pérez-Navarro

2022

Sensors

View full text Add to dashboard Cite

Nowadays, there are a multitude of solutions for indoor positioning, as opposed to standards for outdoor positioning such as GPS. Among the different existing studies on indoor positioning, the use of Wi-Fi signals together with Machine Learning algorithms is one of the most important, as it takes advantage of the current deployment of Wi-Fi networks and the increase in the computing power of computers. Thanks to this, the number of articles published in recent years has been increasing. This fact makes a review necessary in order to understand the current state of this field and to classify different parameters that are very useful for future studies. What are the most widely used machine learning techniques? In what situations have they been tested? How accurate are they? Have datasets been properly used? What type of Wi-Fi signals have been used? These and other questions are answered in this analysis, in which 119 papers are analyzed in depth following PRISMA guidelines.

show abstract

Section: Ujiindoor Results Analysismentioning

confidence: 99%

Comprehensive Analysis of Applied Machine Learning in Indoor Positioning Based on Wi-Fi: An Extended Systematic Review

Bellavista-Parent

Torres-Sospedra

Pérez-Navarro

2022

Sensors

View full text Add to dashboard Cite

show abstract

“…In 2021, Qin et al propose a CDAE-CNN-based Wi-Fi fingerprint location algorithm called Ccpos, which uses the K-means algorithm to segment the dataset, and then uses the CDAE-CNN network for specific location prediction [ 36 ]. In 2022, Wang et al designed a novel DNN-based indoor positioning framework called CHISEL, which combines a convolutional encoder and a CNN classifier [ 37 ]. Jaehoon Cha et al propose a hierarchical auxiliary deep neural network called HADNN, which uses a continuous feedforward network to identify buildings and estimate the floor coordinates [ 38 ].…”

Section: Related Workmentioning

confidence: 99%

“…To further evaluate the performance of our improved DAE and enhanced DNN algorithms, we compare them with state-of-the-art deep learning indoor positioning algorithms, including SAEDNN [ 34 ], CNNLoc [ 35 ], CCpos [ 36 ], CHISEL [ 37 ], CHISEL-DA [ 37 ], and HADNN [ 38 ]. The results are shown in Table 4 and Figure 8 , respectively.…”

Section: Performance Evaluationmentioning

confidence: 99%

“…It can be seen from Table 5 that the average positioning error of machine learning methods in specific-location positioning is 6.13 m at the lowest (KNN) and 10.42 m at the highest (DT and GNB), while the DAE model proposed in this paper reduces to 6.01 m. Similarly, the state-of-the-art deep learning algorithms used for comparison include SAEDNN [ 34 ], CNNLoc [ 35 ], CCpos [ 36 ], CHISEL [ 37 ], CHISEL-DA [ 37 ], and HADNN [ 38 ]. Figure 9 shows the comparison results of DDAE and the advanced deep learning algorithms.…”

Section: Performance Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

Calibration-Free 3D Indoor Positioning Algorithms Based on DNN and DIFF

Yang

Deng

et al. 2022

Sensors

View full text Add to dashboard Cite

The heterogeneity of wireless receiving devices, co-channel interference, and multi-path effect make the received signal strength indication (RSSI) of Wi-Fi fluctuate greatly, which seriously degrades the RSSI-based positioning accuracy. Signal strength difference (DIFF), a calibration-free solution for handling the received signal strength variance between diverse devices, can effectively reduce the negative impact of signal fluctuation. However, DIFF also leads to the explosion of the RSSI data dimension, expanding the number of dimensions from m to Cm2, which reduces the positioning efficiency. To this end, we design a data hierarchical processing strategy based on a building-floor-specific location, which effectively improves the efficiency of high-dimensional data processing. Moreover, based on a deep neural network (DNN), we design three different positioning algorithms for multi-building, multi-floor, and specific-location respectively, extending the indoor positioning from the single plane to three dimensions. Specifically, in the stage of data preprocessing, we first create the original RSSI database. Next, we create the optimized RSSI database by identifying and deleting the unavailable data in the RSSI database. Finally, we perform DIFF processing on the optimized RSSI database to create the DIFF database. In the stage of positioning, firstly, we design an improved multi-building positioning algorithm based on a denoising autoencoder (DAE). Secondly, we design an enhanced DNN for multi-floor positioning. Finally, the newly deep denoising autoencoder (DDAE) used for specific location positioning is proposed. The experimental results show that the proposed algorithms have better positioning efficiency and accuracy compared with the traditional machine learning algorithms and the current advanced deep learning algorithms.

show abstract

“…It takes much effort, high cost, and sometimes has complexity or scalability issues when the applied indoor environment is enormous. There are several attempts to reduce the drawbacks, i.e., constructing the artificial grids by applying classical to machine learning-based techniques and reducing the database complexity by employing some compression algorithm [23]. The effort in signal point-of-view enhances the signal parameters using the dedicated filter to fight the signal fluctuations and implements several clustering techniques to remove the data outliers and improve localization accuracy [24,25].…”

mentioning

confidence: 99%

Fingerprint Database Enhancement by Applying Interpolation and Regression Techniques for IoT-based Indoor Localization

et al. 2022

View full text Add to dashboard Cite

Most applied indoor localization is based on distance and fingerprint techniques. The distance-based technique converts specific parameters to a distance, while the fingerprint technique stores parameters as the fingerprint database. The widely used Internet of Things (IoT) technologies, e.g., Wi-Fi and ZigBee, provide the localization parameters, i.e., received signal strength indicator (RSSI). The fingerprint technique advantages over the distance-based method as it straightforwardly uses the parameter and has better accuracy. However, the burden in database reconstruction in terms of complexity and cost is the disadvantage of this technique. Some solutions, i.e., interpolation, image-based method, machine learning (ML)-based, have been proposed to enhance the fingerprint methods. The limitations are complex and evaluated only in a single environment or simulation. This paper proposes applying classical interpolation and regression to create the synthetic fingerprint database using only a relatively sparse RSSI dataset. We use bilinear and polynomial interpolation and polynomial regression techniques to create the synthetic database and apply our methods to the 2D and 3D environments. We obtain an accuracy improvement of 0.2m for 2D and 0.13m for 3D by applying the synthetic database. Adding the synthetic database can tackle the sparsity issues, and the offline fingerprint database construction will be less burden. Doi: 10.28991/esj-2021-SP1-012 Full Text: PDF

show abstract

CHISEL: Compression-Aware High-Accuracy Embedded Indoor Localization With Deep Learning

Cited by 26 publications

References 11 publications

Comprehensive Analysis of Applied Machine Learning in Indoor Positioning Based on Wi-Fi: An Extended Systematic Review

Comprehensive Analysis of Applied Machine Learning in Indoor Positioning Based on Wi-Fi: An Extended Systematic Review

Calibration-Free 3D Indoor Positioning Algorithms Based on DNN and DIFF

Fingerprint Database Enhancement by Applying Interpolation and Regression Techniques for IoT-based Indoor Localization

Contact Info

Product

Resources

About