scholarly journals Wi-Fi Fingerprinting Based Room Level Indoor Localization Framework Using Ensemble Classifiers

2019 ◽  
Vol 38 (1) ◽  
pp. 151-174 ◽  
Author(s):  
Beenish Ayesha Akram ◽  
Ali Hammad Akbar
Keyword(s):  
Neural Network ◽  
Real World ◽  
Indoor Localization ◽  
Propagation Model ◽  
Radio Propagation ◽  
Ensemble Classifiers ◽  
Propagation Models ◽  
Nearest Neighbours ◽  
Set Up ◽  
Realistic Information

Over the past decennium, Wi-Fi fingerprinting based indoor localization has seized substantial attention. Room level indoor localization can enable numerous applications to increase their diversity by incorporating user location. Real-world commercial scale deployments have not been realized because of difficulty in capturing radio propagation models. In case of fingerprinting based approaches, radio propagation model is implicitly integrated in the gathered fingerprints providing more realistic information as compared to empirical propagation models. We propose ensemble classifiers based indoor localization using Wi-Fi fingerprints for room level prediction. The major advantages of the proposed framework are, ease of training, ease to set up framework providing high room-level accuracy with trifling response time making it viable and appropriate for real time applications. It performs well in comparison with recurrently used ANN (Artificial Neural Network) and kNN (k-Nearest Neighbours) based solutions. Experiments performed showed that on our real-world Wi-Fi fingerprint dataset, our proposed approach achieved 89% accuracy whereas neural network and kNN based best found configurations achieved 85 and 82% accuracy respectively.

scholarly journals An Expedient, Generalizable and Realistic Data-Driven Indoor Propagation Model

2021 ◽  
Author(s):  
Stefanos Sotirios Bakirtzis ◽  
Jiming Chen ◽  
Kehai Qiu ◽  
Jie Zhang ◽  
Ian Wassell
Keyword(s):  
Measurement Data ◽  
Wireless Channel ◽  
Propagation Model ◽  
Radio Propagation ◽  
Data Driven ◽  
Indoor Environments ◽  
Real World Data ◽  
Propagation Models ◽  
Frequency Bands ◽  
Propagation Modelling

Efficient and realistic indoor radio propagation modelling tools are inextricably intertwined with the design and operation of next generation wireless networks. Machine learning (ML)-based radio propagation models can be trained with simulated or real-world data to provide accurate estimates of the wireless channel characteristics in a computationally efficient way. However, most of the existing research on ML-based propagation models focuses on outdoor propagation modelling, while indoor data-driven propagation models remain site-specific with limited scalability. In this paper we present an efficient and credible ML-based radio propagation modelling framework for indoor environments. Specifically, we demonstrate how a convolutional encoder-decoder can be trained to replicate the results of a ray-tracer, by encoding physics-based information of an indoor environment, such as the permittivity of the walls, and decode it as the path-loss (PL) heatmap for an environment of interest. Our model is trained over multiple indoor geometries and frequency bands, and it can eventually predict the PL for unknown indoor geometries and frequency bands within a few milliseconds. Additionally, we illustrate how the concept of transfer learning can be leveraged to calibrate our model by adjusting its pre-estimate weights, allowing it to make predictions that are consistent with measurement data. <br>

scholarly journals An Expedient, Generalizable and Realistic Data-Driven Indoor Propagation Model

2021 ◽  
Author(s):  
Stefanos Sotirios Bakirtzis ◽  
Jiming Chen ◽  
Kehai Qiu ◽  
Jie Zhang ◽  
Ian Wassell
Keyword(s):  
Measurement Data ◽  
Wireless Channel ◽  
Propagation Model ◽  
Radio Propagation ◽  
Data Driven ◽  
Indoor Environments ◽  
Real World Data ◽  
Propagation Models ◽  
Frequency Bands ◽  
Propagation Modelling

Efficient and realistic indoor radio propagation modelling tools are inextricably intertwined with the design and operation of next generation wireless networks. Machine learning (ML)-based radio propagation models can be trained with simulated or real-world data to provide accurate estimates of the wireless channel characteristics in a computationally efficient way. However, most of the existing research on ML-based propagation models focuses on outdoor propagation modelling, while indoor data-driven propagation models remain site-specific with limited scalability. In this paper we present an efficient and credible ML-based radio propagation modelling framework for indoor environments. Specifically, we demonstrate how a convolutional encoder-decoder can be trained to replicate the results of a ray-tracer, by encoding physics-based information of an indoor environment, such as the permittivity of the walls, and decode it as the path-loss (PL) heatmap for an environment of interest. Our model is trained over multiple indoor geometries and frequency bands, and it can eventually predict the PL for unknown indoor geometries and frequency bands within a few milliseconds. Additionally, we illustrate how the concept of transfer learning can be leveraged to calibrate our model by adjusting its pre-estimate weights, allowing it to make predictions that are consistent with measurement data. <br>

scholarly journals CEnsLoc: Infrastructure-Less Indoor Localization Methodology Using GMM Clustering-Based Classification Ensembles

2018 ◽  
Vol 2018 ◽  
pp. 1-11 ◽  
Author(s):  
Beenish Ayesha Akram ◽  
Ali Hammad Akbar ◽  
Ki-Hyung Kim
Keyword(s):  
Machine Learning ◽  
Indoor Localization ◽  
Principal Component ◽  
Nearest Neighbors ◽  
Signal Propagation ◽  
Location Based Services ◽  
Radio Propagation ◽  
Propagation Models ◽  
Related Information ◽  
Real Time Applications

Indoor localization has continued to garner interest over the last decade or so, due to the fact that its realization remains a challenge. Fingerprinting-based systems are exciting because these embody signal propagation-related information intrinsically as compared to radio propagation models. Wi-Fi (an RF technology) is best suited for indoor localization because it is so widely deployed that literally, no additional infrastructure is required. Since location-based services depend on the fingerprints acquired through the underlying technology, smart mechanisms such as machine learning are increasingly being incorporated to extract intelligible information. We propose CEnsLoc, a new easy to train-and-deploy Wi-Fi localization methodology established on GMM clustering and Random Forest Ensembles (RFEs). Principal component analysis was applied for dimension reduction of raw data. Conducted experimentation demonstrates that it provides 97% accuracy for room prediction. However, artificial neural networks, k-nearest neighbors, K∗, FURIA, and DeepLearning4J-based localization solutions provided mean 85%, 91%, 90%, 92%, and 73% accuracy on our collected real-world dataset, respectively. It delivers high room-level accuracy with negligible response time, making it viable and befitted for real-time applications.

scholarly journals A Neural Network Propagation Model for LoRaWAN and Critical Analysis with Real-World Measurements

2017 ◽  
Vol 1 (1) ◽  
pp. 7 ◽  
Author(s):  
Salaheddin Hosseinzadeh ◽  
Mahmood Almoathen ◽  
Hadi Larijani ◽  
Krystyna Curtis
Keyword(s):  
Neural Network ◽  
Real World ◽  
Critical Analysis ◽  
Propagation Model ◽  
Network Propagation

Survey of Radio Propagation Models for Acoustic Applications in Underwater Wireless Sensor Network

2019 ◽  
Vol 09 ◽  
Author(s):  
Preeti Saini ◽  
Rishi Pal Singh ◽  
Adwitiya Sinha
Keyword(s):  
Acoustic Waves ◽  
Path Loss ◽  
Signal Strength ◽  
Propagation Model ◽  
Propagation Loss ◽  
Radio Propagation ◽  
Propagation Models ◽  
Loss Model ◽  
Path Loss Model ◽  
Radio Propagation Model

Background: Acoustic waves have a large range of applications in UWSNs from underwater monitoring to disaster management, military surveillance to assisted navigation. Acoustic waves are primarily used for wireless communication in water. But radio waves are more suitable than acoustic waves for many underwater applications (e.g. real-time applications, shallow water applications). Objectives: A propagation model is required to effectively design a radio wave based UWSN. Propagation model predicts the average received signal strength at a given distance from the transmitter and the variability of the signal strength in close spatial proximity to a particular location. Various radio propagation models are developed for air. Methods: The performance of RF-EM waves underwater is not the same as that in the air. Many parameters which have real-value in the air becomes complex valued in seawater. Thus, propagation models for air cannot be directly used to calculate propagation loss underwater. Various radio propagation models are developed for water by Al-Shamaa’a et al., Uribe and Grote, Jiang et al., Elrashidi et al., Hattab et al. Each model has some merits and demerits. Path loss model developed by Al-Shamma’a et al. is a simple model based on attenuation only. Results: Uribe and Grote have introduced distance-dependent attenuation coefficient in path loss calculation. Path loss model by Jiang et al. calculates path loss for freshwater. Model by Hattab et al. is specifically designed for UWSN. According to the authors, it is the first path loss model developed for UWSN. Elrashidi et al. have calculated path loss for freshwater and seawater at 2.4 GHz. The model includes the effect of the reflected signals on the received signal by the receiver node. Conclusion: The paper presents a comparative analysis of these various radio propagation models developed for underwater. Among these models, the radio propagation model by Hattab et al. is more realistic and covers both propagation loss and interface loss. According to the authors, it is the first radio propagation model developed for UWSNs.

Calculation of Load-Carrying Capacity of Square Concrete Filled Tube Columns Based on Neural Network

2011 ◽  
Vol 71-78 ◽  
pp. 847-850
Author(s):  
Hua Guo Gao
Keyword(s):  
Neural Network ◽  
Experimental Data ◽  
Back Propagation ◽  
Structure Design ◽  
Propagation Model ◽  
The Neural Network ◽  
Load Carrying ◽  
Predicted Values ◽  
Set Up ◽  
Concrete Filled Steel Tubes

Concrete filled steel tubes of square columns under axial load are in complicated stress, the influence of every factor on mechanics performance is difficult to ascertain accurately. Neural network performs well obtaining the relationship between input and output variables by self-studying, self-organizing, self-adapting and nonlinear mapping. In this paper a three-layer back-propagation model of network is successfully trained and set up according to experimental data of square CFT columns under load. Ten groups of experimental data were verified by the model, the results show the predicted values are in accord with test values, precision in calculation is good enough for structure design. So the neural network model can be used as an auxiliary method to calculate the capacity of square concrete filled tube columns in the project. With the increase of experimental data, the neural network precision of prediction will be improved in the future.

Calculation of Load-Carrying Capacity of Square Concrete Filled Tube Columns Based on Neural Network

2013 ◽  
Vol 351-352 ◽  
pp. 713-716
Author(s):  
Hua Guo Gao ◽  
Hang Cheng ◽  
Xiao Feng Cui
Keyword(s):  
Neural Network ◽  
Experimental Data ◽  
Axial Load ◽  
Back Propagation ◽  
Steel Tube ◽  
Structure Design ◽  
Propagation Model ◽  
Load Carrying ◽  
Predicted Values ◽  
Set Up

Steel tube and filled concrete of square CFT columns under axial load are in complicated stress condition, the influence of every kind of factors on mechanics performance is difficult to ascertain accurately. On the other hand, neural network is good at obtaining the relationship between input and output variables by self-studying, self-organizing, self-adapting and nonlinear mapping. Therefore, it is suitable that use neural network to calculating the bearing capacity of square CFT columns. In this paper a three-layer back-propagation model of network is trained according to experimental data of square CFT columns under axial load, a neural network model for axial loaded square CFT columns is set up. The model is verified by six groups of experimental data, the results show the predicted values are in good agreement with test values, precision in calculation is good enough to be used as an auxiliary method for structure design.

Bearing Capacity of Concrete Filled Square Steel Tubular Columns Based on Neural Network

2012 ◽  
Vol 502 ◽  
pp. 193-197 ◽  
Author(s):  
Hai Jun Wang ◽  
Hua Bei Zhu ◽  
Hua Wei
Keyword(s):  
Neural Network ◽  
Experimental Data ◽  
Bearing Capacity ◽  
Back Propagation ◽  
Steel Tube ◽  
Structure Design ◽  
Propagation Model ◽  
Eccentric Load ◽  
Tubular Columns ◽  
Set Up

Steel tube and filled concrete of square CFT (concrete filled steel tubular structures) columns under eccentric load are in complicated stress condition, the influence of every kind of factors on mechanics performance is difficult to ascertain accurately. On the other hand, neural network is good at obtaining the relationship between input and output variables by self-studying, self-organizing, self-adapting and nonlinear mapping. Therefore, it is suitable that use neural network to calculating the bearing capacity of square CFT columns. In this paper a four-layer back-propagation model of network is trained according to experimental data of square CFT columns under eccentric load, a neural network model for eccentrically loaded square CFT columns is set up. The model is verified by six groups of experimental data, the results show the predicted values are in good agreement with test values, precision in calculation is good enough to be used as an auxiliary method for structure design.

Detection and localization of helipad in autonomous UAV landing: a coupled visual-inertial approach with artificial intelligence

2020 ◽  
Vol 71 (7) ◽  
pp. 828-839
Author(s):  
Thinh Hoang Dinh ◽  
Hieu Le Thi Hong
Keyword(s):  
Neural Network ◽  
Unmanned Aerial Vehicles ◽  
Fiducial Marker ◽  
Localization Algorithm ◽  
Challenging Problem ◽  
Autonomous Landing ◽  
Aerial Vehicles ◽  
Set Up ◽  
Detection And Localization ◽  
Rotary Wing

Autonomous landing of rotary wing type unmanned aerial vehicles is a challenging problem and key to autonomous aerial fleet operation. We propose a method for localizing the UAV around the helipad, that is to estimate the relative position of the helipad with respect to the UAV. This data is highly desirable to design controllers that have robust and consistent control characteristics and can find applications in search – rescue operations. AI-based neural network is set up for helipad detection, followed by optimization by the localization algorithm. The performance of this approach is compared against fiducial marker approach, demonstrating good consensus between two estimations

Sign in / Sign up

Export Citation Format

Share Document