A wide range of applications, including sports and healthcare, use human activity recognition (HAR). The Internet of Things (IoT), using cloud systems, offers enormous resources but produces high delays and huge amounts of traffic. This study proposes a distributed intelligence and dynamic HAR architecture using smart IoT devices, edge devices, and cloud computing. These systems were used to train models, store results, and process real-time predictions. Wearable sensors and smartphones were deployed on the human body to detect activities from three positions; accelerometer and gyroscope parameters were utilized to recognize activities. A dynamic selection of models was used, depending on the availability of the data and the mobility of the users. The results showed that this system could handle different scenarios dynamically according to the available features; its prediction accuracy was 99.23% using the LightGBM algorithm during the training stage, when 18 features were used. The prediction time was around 6.4 milliseconds per prediction on the smart end device and 1.6 milliseconds on the Raspberry Pi edge, which can serve more than 30 end devices simultaneously and reduce the need for the cloud. The cloud was used for storing users’ profiles and can be used for real-time prediction in 391 milliseconds per request.
包括体育和医疗在内的各种应用都在使用人类活动识别(HAR)。使用云系统的物联网(IoT)提供了巨大的资源,但也产生了高延迟和巨大的流量。本研究利用智能物联网设备、边缘设备和云计算提出了分布式智能和动态 HAR 架构。这些系统用于训练模型、存储结果和处理实时预测。在人体上部署了可穿戴传感器和智能手机,从三个位置检测活动;利用加速计和陀螺仪参数识别活动。根据数据的可用性和用户的移动性,对模型进行了动态选择。结果表明,该系统可以根据可用特征动态处理不同的场景;在训练阶段,使用 LightGBM 算法,当使用 18 个特征时,其预测准确率为 99.23%。智能终端设备的每次预测时间约为 6.4 毫秒,Raspberry Pi 边缘设备的每次预测时间约为 1.6 毫秒,可同时为 30 多台终端设备提供服务,减少了对云的需求。云用于存储用户配置文件,每次请求的实时预测时间为 391 毫秒。
{"title":"Dynamic and Distributed Intelligence over Smart Devices, Internet of Things Edges, and Cloud Computing for Human Activity Recognition Using Wearable Sensors","authors":"Ayman Wazwaz, Khalid Amin, Noura Semary, Tamer Ghanem","doi":"10.3390/jsan13010005","DOIUrl":"https://doi.org/10.3390/jsan13010005","url":null,"abstract":"A wide range of applications, including sports and healthcare, use human activity recognition (HAR). The Internet of Things (IoT), using cloud systems, offers enormous resources but produces high delays and huge amounts of traffic. This study proposes a distributed intelligence and dynamic HAR architecture using smart IoT devices, edge devices, and cloud computing. These systems were used to train models, store results, and process real-time predictions. Wearable sensors and smartphones were deployed on the human body to detect activities from three positions; accelerometer and gyroscope parameters were utilized to recognize activities. A dynamic selection of models was used, depending on the availability of the data and the mobility of the users. The results showed that this system could handle different scenarios dynamically according to the available features; its prediction accuracy was 99.23% using the LightGBM algorithm during the training stage, when 18 features were used. The prediction time was around 6.4 milliseconds per prediction on the smart end device and 1.6 milliseconds on the Raspberry Pi edge, which can serve more than 30 end devices simultaneously and reduce the need for the cloud. The cloud was used for storing users’ profiles and can be used for real-time prediction in 391 milliseconds per request.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"65 12","pages":""},"PeriodicalIF":3.5,"publicationDate":"2024-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139390065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We analyse the output stream from a packet buffer governed by the policy that incoming packets are dropped with a probability related to the buffer occupancy. The results include formulas for the number of packets departing the buffer in a specific time, for the time-dependent output rate and for the steady-state output rate. The latter is the key performance measure of the buffering mechanism, as it reflects its ability to process a specific number of packets in a time unit. To ensure broad applicability of the results in various networks and traffic types, a powerful and versatile model of the input stream is used, i.e., a BMAP. Numeric examples are provided, with several parameterisations of the BMAP, dropping probabilities and loads of the system.
{"title":"Output Stream from the AQM Queue with BMAP Arrivals","authors":"A. Chydzinski","doi":"10.3390/jsan13010004","DOIUrl":"https://doi.org/10.3390/jsan13010004","url":null,"abstract":"We analyse the output stream from a packet buffer governed by the policy that incoming packets are dropped with a probability related to the buffer occupancy. The results include formulas for the number of packets departing the buffer in a specific time, for the time-dependent output rate and for the steady-state output rate. The latter is the key performance measure of the buffering mechanism, as it reflects its ability to process a specific number of packets in a time unit. To ensure broad applicability of the results in various networks and traffic types, a powerful and versatile model of the input stream is used, i.e., a BMAP. Numeric examples are provided, with several parameterisations of the BMAP, dropping probabilities and loads of the system.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"96 5","pages":""},"PeriodicalIF":3.5,"publicationDate":"2024-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139390437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
LoRaWANs play a critical role in various applications such as smart farming, industrial IoT, and smart cities. The strategic placement of gateways significantly influences network performance optimization. This study presents a comprehensive analysis of the tradeoffs between system costs and bitrate maximization for selecting optimal gateway locations in LoRaWANs. To address this challenge, a rigorous mathematical model is formulated to incorporate essential factors and constraints related to gateway selection. Furthermore, we propose an innovative metaheuristic algorithm known as the M-VaNSAS algorithm, which effectively explores the solution space and identifies favorable gateway locations. The Pareto front and TOPSIS methods are employed to evaluate and rank the generated solutions, providing a robust assessment framework. Our research findings highlight the suitability of a network model comprising 144 gateways tailored for the Ubon Ratchathani province. Among the evaluated algorithms, the M-VaNSAS method demonstrates exceptional efficiency in gateway location selection, outperforming the PSO, DE, and GA methods.
{"title":"Multi-Objective Optimization of Gateway Location Selection in Long-Range Wide Area Networks: A Tradeoff Analysis between System Costs and Bitrate Maximization","authors":"Charuay Savithi, Chutchai Kaewta","doi":"10.3390/jsan13010003","DOIUrl":"https://doi.org/10.3390/jsan13010003","url":null,"abstract":"LoRaWANs play a critical role in various applications such as smart farming, industrial IoT, and smart cities. The strategic placement of gateways significantly influences network performance optimization. This study presents a comprehensive analysis of the tradeoffs between system costs and bitrate maximization for selecting optimal gateway locations in LoRaWANs. To address this challenge, a rigorous mathematical model is formulated to incorporate essential factors and constraints related to gateway selection. Furthermore, we propose an innovative metaheuristic algorithm known as the M-VaNSAS algorithm, which effectively explores the solution space and identifies favorable gateway locations. The Pareto front and TOPSIS methods are employed to evaluate and rank the generated solutions, providing a robust assessment framework. Our research findings highlight the suitability of a network model comprising 144 gateways tailored for the Ubon Ratchathani province. Among the evaluated algorithms, the M-VaNSAS method demonstrates exceptional efficiency in gateway location selection, outperforming the PSO, DE, and GA methods.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"106 6","pages":""},"PeriodicalIF":3.5,"publicationDate":"2024-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139391262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurate localization holds paramount importance across many applications within the context of smart cities, particularly in vehicular communication systems, the Internet of Things, and Integrated Sensing and Communication (ISAC) technologies. Nonetheless, achieving precise localization remains a persistent challenge, primarily attributed to the prevalence of non-line-of-sight (NLOS) conditions and the presence of uncertainties surrounding key wireless transmission parameters. This paper presents a comprehensive framework tailored to address the intricate task of localizing multiple nodes within ISAC systems significantly impacted by pervasive NLOS conditions and the ambiguity of transmission parameters. The proposed methodology integrates received signal strength (RSS) and time-of-arrival (TOA) measurements as a strategic response to effectively overcome these substantial challenges, even in situations where the precise values of transmitting power and temporal information remain elusive. An approximation approach is judiciously employed to facilitate the inherent non-convex and NP-hard nature of the original estimation problem, resulting in a notable transformation, rendering the problem amenable to a convex optimization paradigm. The comprehensive array of simulations conducted within this study corroborates the efficacy of the proposed hybrid cooperative localization method by underscoring its superior performance relative to conventional approaches relying solely on RSS or TOA measurements. This enhancement in localization accuracy in ISAC systems holds promise in the intricate urban landscape of smart cities, offering substantial contributions to infrastructure optimization and service efficiency.
在智慧城市的许多应用中,尤其是在车载通信系统、物联网和综合传感与通信(ISAC)技术中,精确定位至关重要。然而,实现精确定位仍然是一个长期存在的挑战,这主要归因于非视线(NLOS)条件的普遍存在以及关键无线传输参数的不确定性。受普遍存在的非视距条件和传输参数不确定性的严重影响,ISAC 系统中的多个节点的定位任务错综复杂,本文提出了一个专门用于解决这一问题的综合框架。所提出的方法整合了接收信号强度(RSS)和到达时间(TOA)测量,作为有效克服这些重大挑战的战略对策,即使在发射功率和时间信息的精确值仍然难以捉摸的情况下也是如此。为了解决原始估计问题固有的非凸和 NP-困难性质,本研究明智地采用了近似方法,从而实现了显著的转变,使问题适合于凸优化范例。本研究中进行的一系列综合模拟证实了所提出的混合合作定位方法的有效性,强调了其相对于仅依赖 RSS 或 TOA 测量的传统方法的卓越性能。在智能城市错综复杂的城市景观中,ISAC 系统定位精度的提高为基础设施优化和服务效率做出了巨大贡献。
{"title":"Robust ISAC Localization in Smart Cities: A Hybrid Network Approach for NLOS Challenges with Uncertain Parameters","authors":"Turke Althobaiti, R. A. Khalil, Nasir Saeed","doi":"10.3390/jsan13010002","DOIUrl":"https://doi.org/10.3390/jsan13010002","url":null,"abstract":"Accurate localization holds paramount importance across many applications within the context of smart cities, particularly in vehicular communication systems, the Internet of Things, and Integrated Sensing and Communication (ISAC) technologies. Nonetheless, achieving precise localization remains a persistent challenge, primarily attributed to the prevalence of non-line-of-sight (NLOS) conditions and the presence of uncertainties surrounding key wireless transmission parameters. This paper presents a comprehensive framework tailored to address the intricate task of localizing multiple nodes within ISAC systems significantly impacted by pervasive NLOS conditions and the ambiguity of transmission parameters. The proposed methodology integrates received signal strength (RSS) and time-of-arrival (TOA) measurements as a strategic response to effectively overcome these substantial challenges, even in situations where the precise values of transmitting power and temporal information remain elusive. An approximation approach is judiciously employed to facilitate the inherent non-convex and NP-hard nature of the original estimation problem, resulting in a notable transformation, rendering the problem amenable to a convex optimization paradigm. The comprehensive array of simulations conducted within this study corroborates the efficacy of the proposed hybrid cooperative localization method by underscoring its superior performance relative to conventional approaches relying solely on RSS or TOA measurements. This enhancement in localization accuracy in ISAC systems holds promise in the intricate urban landscape of smart cities, offering substantial contributions to infrastructure optimization and service efficiency.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"131 5","pages":""},"PeriodicalIF":3.5,"publicationDate":"2023-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139146418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The approach of federated learning (FL) addresses significant challenges, including access rights, privacy, security, and the availability of diverse data. However, edge devices produce and collect data in a non-independent and identically distributed (non-IID) manner. Therefore, it is possible that the number of data samples may vary among the edge devices. This study elucidates an approach for implementing FL to achieve a balance between training accuracy and imbalanced data. This approach entails the implementation of data augmentation in data distribution by utilizing class estimation and by balancing on the client side during local training. Secondly, simple linear regression is utilized for model training at the client side to manage the optimal computation cost to achieve a reduction in computation cost. To validate the proposed approach, the technique was applied to a stock market dataset comprising stocks (AAL, ADBE, ASDK, and BSX) to predict the day-to-day values of stocks. The proposed approach has demonstrated favorable results, exhibiting a strong fit of 0.95 and above with a low error rate. The R-squared values, predominantly ranging from 0.97 to 0.98, indicate the model’s effectiveness in capturing variations in stock prices. Strong fits are observed within 75 to 80 iterations for stocks displaying consistently high R-squared values, signifying accuracy. On the 100th iteration, the declining MSE, MAE, and RMSE (AAL at 122.03, 4.89, 11.04, respectively; ADBE at 457.35, 17.79, and 21.38, respectively; ASDK at 182.78, 5.81, 13.51, respectively; and BSX at 34.50, 4.87, 5.87, respectively) values corroborated the positive results of the proposed approach with minimal data loss.
{"title":"Reduction in Data Imbalance for Client-Side Training in Federated Learning for the Prediction of Stock Market Prices","authors":"Momina Shaheen, M. Farooq, Tariq Umer","doi":"10.3390/jsan13010001","DOIUrl":"https://doi.org/10.3390/jsan13010001","url":null,"abstract":"The approach of federated learning (FL) addresses significant challenges, including access rights, privacy, security, and the availability of diverse data. However, edge devices produce and collect data in a non-independent and identically distributed (non-IID) manner. Therefore, it is possible that the number of data samples may vary among the edge devices. This study elucidates an approach for implementing FL to achieve a balance between training accuracy and imbalanced data. This approach entails the implementation of data augmentation in data distribution by utilizing class estimation and by balancing on the client side during local training. Secondly, simple linear regression is utilized for model training at the client side to manage the optimal computation cost to achieve a reduction in computation cost. To validate the proposed approach, the technique was applied to a stock market dataset comprising stocks (AAL, ADBE, ASDK, and BSX) to predict the day-to-day values of stocks. The proposed approach has demonstrated favorable results, exhibiting a strong fit of 0.95 and above with a low error rate. The R-squared values, predominantly ranging from 0.97 to 0.98, indicate the model’s effectiveness in capturing variations in stock prices. Strong fits are observed within 75 to 80 iterations for stocks displaying consistently high R-squared values, signifying accuracy. On the 100th iteration, the declining MSE, MAE, and RMSE (AAL at 122.03, 4.89, 11.04, respectively; ADBE at 457.35, 17.79, and 21.38, respectively; ASDK at 182.78, 5.81, 13.51, respectively; and BSX at 34.50, 4.87, 5.87, respectively) values corroborated the positive results of the proposed approach with minimal data loss.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"77 6","pages":""},"PeriodicalIF":3.5,"publicationDate":"2023-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138951408","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
LoRa technology is being integrated into industrial applications as part of Industry 4.0 owing to its longer range and low power consumption. However, noise, interference, and the fading effect all have a negative impact on LoRa performance in an industrial environment, necessitating solutions to ensure reliable communication. This paper evaluates and compares LoRa’s performance in terms of packet error rate (PER) with and without forward error correction (FEC) in an industrial environment. The impact of integrating an infinite impulse response (IIR) or finite impulse response (FIR) filter into the LoRa architecture is also evaluated. Simulations are carried out in MATLAB at 868 MHz with a bandwidth of 125 kHz and two spreading factors of 7 and 12. Many-to-one and one-to-many communication modes are considered, as are line of sight (LOS) and non-line of Sight (NLOS) conditions. Simulation results show that, compared to an environment with additive white Gaussian noise (AWGN), LoRa technology suffers a significant degradation of its PER performance in industrial environments. Nevertheless, the use of forward error correction (FEC) contributes positively to offsetting this decline. Depending on the configuration and architecture examined, the gain in signal-to-noise ratio (SNR) using a 4/8 coding ratio ranges from 7 dB to 11 dB. Integrating IIR or FIR filters also boosts performance, with additional SNR gains ranging from 2 dB to 6 dB, depending on the simulation parameters.
作为工业 4.0 的一部分,LoRa 技术因其较远的传输距离和较低的功耗而被集成到工业应用中。然而,噪声、干扰和衰减效应都会对 LoRa 在工业环境中的性能产生负面影响,因此需要解决方案来确保通信的可靠性。本文评估并比较了 LoRa 在工业环境中使用和不使用前向纠错(FEC)时的数据包错误率(PER)性能。此外,还评估了在 LoRa 架构中集成无限脉冲响应(IIR)或有限脉冲响应(FIR)滤波器的影响。仿真在 MATLAB 中进行,频率为 868 MHz,带宽为 125 kHz,两个扩展因子分别为 7 和 12。考虑了多对一和一对多通信模式,以及视线(LOS)和非视线(NLOS)条件。仿真结果表明,与加性白高斯噪声(AWGN)环境相比,LoRa 技术在工业环境中的 PER 性能明显下降。不过,使用前向纠错(FEC)可积极抵消这种下降。根据所研究的配置和架构,使用 4/8 编码比的信噪比 (SNR) 增益范围在 7 dB 到 11 dB 之间。集成 IIR 或 FIR 滤波器也能提高性能,额外的信噪比增益从 2 dB 到 6 dB 不等,具体取决于模拟参数。
{"title":"Performance Evaluation of LoRa Communications in Harsh Industrial Environments","authors":"L. Aarif, Mohamed Tabaa, Hanaa Hachimi","doi":"10.3390/jsan12060080","DOIUrl":"https://doi.org/10.3390/jsan12060080","url":null,"abstract":"LoRa technology is being integrated into industrial applications as part of Industry 4.0 owing to its longer range and low power consumption. However, noise, interference, and the fading effect all have a negative impact on LoRa performance in an industrial environment, necessitating solutions to ensure reliable communication. This paper evaluates and compares LoRa’s performance in terms of packet error rate (PER) with and without forward error correction (FEC) in an industrial environment. The impact of integrating an infinite impulse response (IIR) or finite impulse response (FIR) filter into the LoRa architecture is also evaluated. Simulations are carried out in MATLAB at 868 MHz with a bandwidth of 125 kHz and two spreading factors of 7 and 12. Many-to-one and one-to-many communication modes are considered, as are line of sight (LOS) and non-line of Sight (NLOS) conditions. Simulation results show that, compared to an environment with additive white Gaussian noise (AWGN), LoRa technology suffers a significant degradation of its PER performance in industrial environments. Nevertheless, the use of forward error correction (FEC) contributes positively to offsetting this decline. Depending on the configuration and architecture examined, the gain in signal-to-noise ratio (SNR) using a 4/8 coding ratio ranges from 7 dB to 11 dB. Integrating IIR or FIR filters also boosts performance, with additional SNR gains ranging from 2 dB to 6 dB, depending on the simulation parameters.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"49 1","pages":""},"PeriodicalIF":3.5,"publicationDate":"2023-11-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139220590","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This research proposes a unique platform for energy management optimization in smart grids, based on 6G technologies. The proposed platform, applied on a virtual power plant, includes algorithms that take into account different profiles of loads and fairly schedules energy according to loads priorities and compensates for the intermittent nature of renewable energy sources. Moreover, we develop a bidirectional energy transition mechanism towards a fleet of intelligent vehicles by adopting vehicle-to-grid technology and peak clipping. Performance analysis shows that the proposed energy provides fairness to electrical vehicles, satisfies urgent loads, and optimizes smart grids energy.
{"title":"Electric Vehicles Energy Management for Vehicle-to-Grid 6G-Based Smart Grid Networks","authors":"Rola Naja, Aakash Soni, Circe Carletti","doi":"10.3390/jsan12060079","DOIUrl":"https://doi.org/10.3390/jsan12060079","url":null,"abstract":"This research proposes a unique platform for energy management optimization in smart grids, based on 6G technologies. The proposed platform, applied on a virtual power plant, includes algorithms that take into account different profiles of loads and fairly schedules energy according to loads priorities and compensates for the intermittent nature of renewable energy sources. Moreover, we develop a bidirectional energy transition mechanism towards a fleet of intelligent vehicles by adopting vehicle-to-grid technology and peak clipping. Performance analysis shows that the proposed energy provides fairness to electrical vehicles, satisfies urgent loads, and optimizes smart grids energy.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"21 1","pages":""},"PeriodicalIF":3.5,"publicationDate":"2023-11-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139232947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A result of the pandemic is an urgent need for data collaborations that empower the clinical and scientific communities in responding to rapidly evolving global challenges. The ICU4Covid project joined research institutions, medical centers, and hospitals all around Europe in a telemedicine network for sharing capabilities, knowledge, and expertise distributed within the network. However, healthcare data sharing has ethical, regulatory, and legal complexities that pose several restrictions on their access and use. To mitigate this issue, the ICU4Covid project integrates a federated learning architecture, allowing distributed machine learning within a cross-institutional healthcare system without the data being transported or exposed outside their original location. This paper presents the federated learning approach to support the decision-making process for ICU patients in a European telemedicine network. The proposed approach was applied to the early identification of high-risk hypertensive patients. Experimental results show how the knowledge of every single node is spread within the federation, improving the ability of each node to make an early prediction of high-risk hypertensive patients. Moreover, a performance evaluation shows an accuracy and precision of over 90%, confirming a good performance of the FL approach as a prediction test. The FL approach can significantly support the decision-making process for ICU patients in distributed networks of federated healthcare organizations.
{"title":"A Federated Learning Approach to Support the Decision-Making Process for ICU Patients in a European Telemedicine Network","authors":"Giovanni Paragliola, Patrizia Ribino, Zaib Ullah","doi":"10.3390/jsan12060078","DOIUrl":"https://doi.org/10.3390/jsan12060078","url":null,"abstract":"A result of the pandemic is an urgent need for data collaborations that empower the clinical and scientific communities in responding to rapidly evolving global challenges. The ICU4Covid project joined research institutions, medical centers, and hospitals all around Europe in a telemedicine network for sharing capabilities, knowledge, and expertise distributed within the network. However, healthcare data sharing has ethical, regulatory, and legal complexities that pose several restrictions on their access and use. To mitigate this issue, the ICU4Covid project integrates a federated learning architecture, allowing distributed machine learning within a cross-institutional healthcare system without the data being transported or exposed outside their original location. This paper presents the federated learning approach to support the decision-making process for ICU patients in a European telemedicine network. The proposed approach was applied to the early identification of high-risk hypertensive patients. Experimental results show how the knowledge of every single node is spread within the federation, improving the ability of each node to make an early prediction of high-risk hypertensive patients. Moreover, a performance evaluation shows an accuracy and precision of over 90%, confirming a good performance of the FL approach as a prediction test. The FL approach can significantly support the decision-making process for ICU patients in distributed networks of federated healthcare organizations.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"41 5","pages":""},"PeriodicalIF":3.5,"publicationDate":"2023-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139256521","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This research presents a machine learning modeling process for detecting mental fatigue using three physiological signals: electrodermal activity, electrocardiogram, and respiration. It follows the conventional machine learning modeling pipeline, while emphasizing the significant contribution of the feature selection process, resulting in, not only a high-performance model, but also a relevant one. The employed feature selection process considers both statistical and contextual aspects of feature relevance. Statistical relevance was assessed through variance and correlation analyses between independent features and the dependent variable (fatigue state). A contextual analysis was based on insights derived from the experimental design and feature characteristics. Additionally, feature sequencing and set conversion techniques were employed to incorporate the temporal aspects of physiological signals into the training of machine learning models based on random forest, decision tree, support vector machine, k-nearest neighbors, and gradient boosting. An evaluation was conducted using a dataset acquired from a wearable electronic system (in third-party research) with physiological data from three subjects undergoing a series of tests and fatigue stages. A total of 18 tests were performed by the 3 subjects in 3 mental fatigue states. Fatigue assessment was based on subjective measures and reaction time tests, and fatigue induction was performed through mental arithmetic operations. The results showed the highest performance when using random forest, achieving an average accuracy and F1-score of 96% in classifying three levels of mental fatigue.
{"title":"Enhancing Mental Fatigue Detection through Physiological Signals and Machine Learning Using Contextual Insights and Efficient Modelling","authors":"Carole-Anne Cos, Alexandre Lambert, Aakash Soni, Haifa Jeridi, Coralie Thieulin, Amine Jaouadi","doi":"10.3390/jsan12060077","DOIUrl":"https://doi.org/10.3390/jsan12060077","url":null,"abstract":"This research presents a machine learning modeling process for detecting mental fatigue using three physiological signals: electrodermal activity, electrocardiogram, and respiration. It follows the conventional machine learning modeling pipeline, while emphasizing the significant contribution of the feature selection process, resulting in, not only a high-performance model, but also a relevant one. The employed feature selection process considers both statistical and contextual aspects of feature relevance. Statistical relevance was assessed through variance and correlation analyses between independent features and the dependent variable (fatigue state). A contextual analysis was based on insights derived from the experimental design and feature characteristics. Additionally, feature sequencing and set conversion techniques were employed to incorporate the temporal aspects of physiological signals into the training of machine learning models based on random forest, decision tree, support vector machine, k-nearest neighbors, and gradient boosting. An evaluation was conducted using a dataset acquired from a wearable electronic system (in third-party research) with physiological data from three subjects undergoing a series of tests and fatigue stages. A total of 18 tests were performed by the 3 subjects in 3 mental fatigue states. Fatigue assessment was based on subjective measures and reaction time tests, and fatigue induction was performed through mental arithmetic operations. The results showed the highest performance when using random forest, achieving an average accuracy and F1-score of 96% in classifying three levels of mental fatigue.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"34 15","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135868234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Most IoT systems designed for the implementation of mission-critical systems are multi-layered. Much of the computing is done in the service and gateway layers. The gateway layer connects the internal section of the IoT to the cloud through the Internet. The failure of any node between the servers and the gateways will isolate the entire network, leading to zero tolerance. The service and gateway layers must be connected using networking topologies to yield 100% fault tolerance. The empirical formulation of the model chosen to connect the service’s servers to the gateways through routers is required to compute the fault tolerance of the network. A rectangular and interstitial mesh have been proposed in this paper to connect the service servers to the gateways through the servers, which yields 0.999 fault tolerance of the IoT network. Also provided is an empirical approach to computing the IoT network’s fault tolerance. A rectangular and interstitial mesh have been implemented in the network’s gateway layer, increasing the IoT network’s ability to tolerate faults by 11%.
{"title":"Enhancing the Fault Tolerance of a Multi-Layered IoT Network through Rectangular and Interstitial Mesh in the Gateway Layer","authors":"Sastry Kodanda Rama Jammalamadaka, Bhupati Chokara, Sasi Bhanu Jammalamadaka, Balakrishna Kamesh Duvvuri, Rajarao Budaraju","doi":"10.3390/jsan12050076","DOIUrl":"https://doi.org/10.3390/jsan12050076","url":null,"abstract":"Most IoT systems designed for the implementation of mission-critical systems are multi-layered. Much of the computing is done in the service and gateway layers. The gateway layer connects the internal section of the IoT to the cloud through the Internet. The failure of any node between the servers and the gateways will isolate the entire network, leading to zero tolerance. The service and gateway layers must be connected using networking topologies to yield 100% fault tolerance. The empirical formulation of the model chosen to connect the service’s servers to the gateways through routers is required to compute the fault tolerance of the network. A rectangular and interstitial mesh have been proposed in this paper to connect the service servers to the gateways through the servers, which yields 0.999 fault tolerance of the IoT network. Also provided is an empirical approach to computing the IoT network’s fault tolerance. A rectangular and interstitial mesh have been implemented in the network’s gateway layer, increasing the IoT network’s ability to tolerate faults by 11%.","PeriodicalId":37584,"journal":{"name":"Journal of Sensor and Actuator Networks","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136112985","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}