High-accuracy and low-latency indoor place prediction for mobile users can enable a wide range of applications for domains such as assisted living and smart homes. In this paper, we propose GoPlaces, a practical indoor place prediction system that works on mobile devices without requiring any new infrastructure. GoPlaces does not rely on servers or specialized localization infrastructure, except for a single cheap off-the-shelf WiFi access point that supports ranging with Round Trip Time (RTT) protocol. GoPlaces enables personalized place naming and prediction, and it protects users’ location privacy. It fuses inertial sensor data with distances estimated using the WiFi-RTT protocol to predict the indoor places a user will visit. GoPlaces employs an attention-based BiLSTM model to detect user’s current trajectory, which is then used together with historical information stored in a prediction tree to infer user’s future places. We implemented GoPlaces in Android and evaluated it in several indoor spaces. The experimental results demonstrate prediction accuracy as high as 86%. Furthermore, they show GoPlaces is feasible in real life because it has low latency and low resource consumption on the phones.
{"title":"On-device indoor place prediction using WiFi-RTT and inertial sensors","authors":"Pritam Sen , Xiaopeng Jiang , Qiong Wu , Manoop Talasila , Wen-Ling Hsu , Cristian Borcea","doi":"10.1016/j.pmcj.2025.102118","DOIUrl":"10.1016/j.pmcj.2025.102118","url":null,"abstract":"<div><div>High-accuracy and low-latency indoor place prediction for mobile users can enable a wide range of applications for domains such as assisted living and smart homes. In this paper, we propose GoPlaces, a practical indoor place prediction system that works on mobile devices without requiring any new infrastructure. GoPlaces does not rely on servers or specialized localization infrastructure, except for a single cheap off-the-shelf WiFi access point that supports ranging with Round Trip Time (RTT) protocol. GoPlaces enables personalized place naming and prediction, and it protects users’ location privacy. It fuses inertial sensor data with distances estimated using the WiFi-RTT protocol to predict the indoor places a user will visit. GoPlaces employs an attention-based BiLSTM model to detect user’s current trajectory, which is then used together with historical information stored in a prediction tree to infer user’s future places. We implemented GoPlaces in Android and evaluated it in several indoor spaces. The experimental results demonstrate prediction accuracy as high as 86%. Furthermore, they show GoPlaces is feasible in real life because it has low latency and low resource consumption on the phones.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"114 ","pages":"Article 102118"},"PeriodicalIF":3.5,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145158997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01Epub Date: 2025-10-04DOI: 10.1016/j.pmcj.2025.102121
Yunlong Gu , Meng Xu , Jiguang Li , Qilei Li , Zhao Huang , Mengshan Li , Lixin Guan , Mikko Valkama
With the growing demand for location-based services, WiFi localization plays a critical role in indoor environments. While most existing methods rely on Multi-Layer Perceptrons (MLPs), these models often suffer from limited accuracy and poor generalization across diverse deployment conditions. Kolmogorov–Arnold Networks (KANs), with their B-spline-based basis functions, better capture complex nonlinear relationships while reducing overfitting risks. However, original KANs still incur high computational costs. To address this, we propose WiKAN(WiFi KAN), a lightweight KAN-based model for indoor WiFi localization. WiKAN reduces computational complexity by simplifying the network structure to just two KANLinear layers and replacing parameter-intensive operations with optimized matrix multiplications using reconstructed basis functions. Compared to conventional computation of basis coefficients, matrix operations enable faster inference on modern hardware and improve scalability. Furthermore, WiKAN integrates SiLU and B-spline activations through a learnable linear combination, balancing smooth approximation and nonlinear representation. Experiments on three benchmark datasets (UJIIndoorLoc, Tampere, and JARIL) demonstrate that WiKAN achieves superior performance to both MLP and standard KAN models: over 99.9% building accuracy, up to 100% floor classification, and average positioning error reduced to 5.91 meters. Additionally, runtime analysis and parameter count comparisons confirm the model’s computational efficiency. Code is publicly available at: https://github.com/gyl555666/WiKAN.
{"title":"WiKAN: Lightweight Kolmogorov–Arnold Networks for accurate indoor WiFi localization","authors":"Yunlong Gu , Meng Xu , Jiguang Li , Qilei Li , Zhao Huang , Mengshan Li , Lixin Guan , Mikko Valkama","doi":"10.1016/j.pmcj.2025.102121","DOIUrl":"10.1016/j.pmcj.2025.102121","url":null,"abstract":"<div><div>With the growing demand for location-based services, WiFi localization plays a critical role in indoor environments. While most existing methods rely on Multi-Layer Perceptrons (MLPs), these models often suffer from limited accuracy and poor generalization across diverse deployment conditions. Kolmogorov–Arnold Networks (KANs), with their B-spline-based basis functions, better capture complex nonlinear relationships while reducing overfitting risks. However, original KANs still incur high computational costs. To address this, we propose WiKAN(WiFi KAN), a lightweight KAN-based model for indoor WiFi localization. WiKAN reduces computational complexity by simplifying the network structure to just two KANLinear layers and replacing parameter-intensive operations with optimized matrix multiplications using reconstructed basis functions. Compared to conventional computation of basis coefficients, matrix operations enable faster inference on modern hardware and improve scalability. Furthermore, WiKAN integrates SiLU and B-spline activations through a learnable linear combination, balancing smooth approximation and nonlinear representation. Experiments on three benchmark datasets (UJIIndoorLoc, Tampere, and JARIL) demonstrate that WiKAN achieves superior performance to both MLP and standard KAN models: over 99.9% building accuracy, up to 100% floor classification, and average positioning error reduced to 5.91 meters. Additionally, runtime analysis and parameter count comparisons confirm the model’s computational efficiency. Code is publicly available at: <span><span>https://github.com/gyl555666/WiKAN</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"114 ","pages":"Article 102121"},"PeriodicalIF":3.5,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145268085","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01Epub Date: 2025-08-17DOI: 10.1016/j.pmcj.2025.102107
Armir Bujari , Mirko Franco , Claudio E. Palazzi , Davide Quaglio , Anna Maria Vegni
Pervasive and mobile computing can play a crucial role in the prevention, detection and management of natural and human-caused disasters. In this context, the Internet of Vehicles (IoV) is particularly noteworthy due to its recent technological advancements and increasing prevalence. In fact, IoV can be leveraged to improve various applications, including those aimed at reducing the millions of fatalities that occur every year. The effectiveness of these applications often relies on the rapid dissemination of emergency messages through position-based forwarding protocols, which can unfortunately be vulnerable to adversarial attacks. Without loss of generality, we focus on the specific case study of road safety to provide a realistic example and discuss two potential attacks based on fake position claims that malicious nodes could easily execute to compromise the performance of the position-based forwarding protocol. We also propose and analyze a validation system based on machine learning (ML) techniques designed to detect malicious nodes, discard false information, and protect against these attacks.
{"title":"Position claim verification for emergency message propagation in Vehicular Ad-Hoc Networks","authors":"Armir Bujari , Mirko Franco , Claudio E. Palazzi , Davide Quaglio , Anna Maria Vegni","doi":"10.1016/j.pmcj.2025.102107","DOIUrl":"10.1016/j.pmcj.2025.102107","url":null,"abstract":"<div><div>Pervasive and mobile computing can play a crucial role in the prevention, detection and management of natural and human-caused disasters. In this context, the Internet of Vehicles (IoV) is particularly noteworthy due to its recent technological advancements and increasing prevalence. In fact, IoV can be leveraged to improve various applications, including those aimed at reducing the millions of fatalities that occur every year. The effectiveness of these applications often relies on the rapid dissemination of emergency messages through position-based forwarding protocols, which can unfortunately be vulnerable to adversarial attacks. Without loss of generality, we focus on the specific case study of road safety to provide a realistic example and discuss two potential attacks based on fake position claims that malicious nodes could easily execute to compromise the performance of the position-based forwarding protocol. We also propose and analyze a validation system based on machine learning (ML) techniques designed to detect malicious nodes, discard false information, and protect against these attacks.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"114 ","pages":"Article 102107"},"PeriodicalIF":3.5,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145027667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01Epub Date: 2025-10-14DOI: 10.1016/j.pmcj.2025.102122
Xiaoping Wu , Xiang Wang , Lingfang Kong, Keqi Zhou
Received Signal Strength (RSS) measurements are widely applied in wireless localization. In this paper, standard form of Alternating Direction Method of Multipliers (ADMM) is designed for source localization using RSS. The Maximum Likelihood (ML) estimation problem of RSS-based localization is equivalent to the standard ADMM form by defining the intermediate variables. Following this, we develop the solutions to the subproblems in the ADMM structure. The convergence of the proposed ADMM solution is discussed based on the convexity analysis of the subproblems, providing the evidence for its stable performance. The simulated results show that the ADMM solution performs efficiently, especially with a small number of sensors or in the presence of high noise levels. In addition, we also verify the bias performance in the source position estimation.
{"title":"Asymptotically efficient ADMM solutions for source localization using RSS measurements","authors":"Xiaoping Wu , Xiang Wang , Lingfang Kong, Keqi Zhou","doi":"10.1016/j.pmcj.2025.102122","DOIUrl":"10.1016/j.pmcj.2025.102122","url":null,"abstract":"<div><div>Received Signal Strength (RSS) measurements are widely applied in wireless localization. In this paper, standard form of Alternating Direction Method of Multipliers (ADMM) is designed for source localization using RSS. The Maximum Likelihood (ML) estimation problem of RSS-based localization is equivalent to the standard ADMM form by defining the intermediate variables. Following this, we develop the solutions to the subproblems in the ADMM structure. The convergence of the proposed ADMM solution is discussed based on the convexity analysis of the subproblems, providing the evidence for its stable performance. The simulated results show that the ADMM solution performs efficiently, especially with a small number of sensors or in the presence of high noise levels. In addition, we also verify the bias performance in the source position estimation.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"114 ","pages":"Article 102122"},"PeriodicalIF":3.5,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145333339","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01Epub Date: 2025-08-23DOI: 10.1016/j.pmcj.2025.102104
Zhiqi Li, Qing Wei, Wenle Bai
Unmanned aerial vehicles (UAVs) are viewed as a potential technology for handling user offloading duties as edge nodes. With their adaptable qualities, UAVs may be quickly deployed to useful locations and service consumers. However, the inability of UAVs to operate continuously for an extended time is a challenge for the current UAV-assisted mobile edge computing solutions. We put forth an optimization problem that involves the dynamic division of computational windows for UAVs, the optimization of user grouping and user transmission power, and the optimization of UAV deployment locations to save energy. We design a Communication-Computing Resource Scheduling with Dynamic computational Window allocation (CCRS-DW) algorithm to realize the problem decomposition and optimization. Specifically, the -means clustering technique and the bisection search are used to tackle this problem. Simulation results show that the energy consumption of the proposed CCRS-DW scheme is significantly lower than that of other benchmark schemes.
{"title":"Minimizing communication-computing energy consumption for UAV assisted collaborative computing offloading","authors":"Zhiqi Li, Qing Wei, Wenle Bai","doi":"10.1016/j.pmcj.2025.102104","DOIUrl":"10.1016/j.pmcj.2025.102104","url":null,"abstract":"<div><div>Unmanned aerial vehicles (UAVs) are viewed as a potential technology for handling user offloading duties as edge nodes. With their adaptable qualities, UAVs may be quickly deployed to useful locations and service consumers. However, the inability of UAVs to operate continuously for an extended time is a challenge for the current UAV-assisted mobile edge computing solutions. We put forth an optimization problem that involves the dynamic division of computational windows for UAVs, the optimization of user grouping and user transmission power, and the optimization of UAV deployment locations to save energy. We design a Communication-Computing Resource Scheduling with Dynamic computational Window allocation (CCRS-DW) algorithm to realize the problem decomposition and optimization. Specifically, the <span><math><mi>K</mi></math></span>-means clustering technique and the bisection search are used to tackle this problem. Simulation results show that the energy consumption of the proposed CCRS-DW scheme is significantly lower than that of other benchmark schemes.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"113 ","pages":"Article 102104"},"PeriodicalIF":3.5,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144913694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-01Epub Date: 2025-08-19DOI: 10.1016/j.pmcj.2025.102106
Annalisa Socievole, Clara Pizzuti
Community detection plays a critical role in disaster recovery and pervasive computing, where identifying cohesive social groups enables more effective communication, coordination, and resource allocation. In mobile and resource-constrained environments such as emergency response systems or mobile opportunistic networks, community detection methods must balance accuracy with computational efficiency. In this work, we propose a novel approach that uncovers community structures from a sparse representation of the original graph, addressing the need for lightweight and scalable algorithms in pervasive and mobile systems. Specifically, we apply Spielman–Srivastava spectral sparsification as a preprocessing step to reduce the number of edges while preserving the key spectral properties that underpin community structure. We then apply a modularity-optimizing genetic algorithm on the sparsified graph. Our experiments, conducted on both synthetic benchmarks and real-world networks, demonstrate that the proposed method, namely SSGA, achieves competitive or superior accuracy compared to state-of-the-art baselines, even under aggressive sparsification. We also analyze the cumulative computational complexity of the approach and provide an optimized implementation based on truncated spectral decomposition and parallel genetic operations. The results confirm that SSGA is not only accurate and robust but also computationally efficient, making it particularly well-suited for pervasive and mobile scenarios where time, energy, and connectivity are limited.
{"title":"Efficient community detection in disaster networks using spectral sparsification","authors":"Annalisa Socievole, Clara Pizzuti","doi":"10.1016/j.pmcj.2025.102106","DOIUrl":"10.1016/j.pmcj.2025.102106","url":null,"abstract":"<div><div>Community detection plays a critical role in disaster recovery and pervasive computing, where identifying cohesive social groups enables more effective communication, coordination, and resource allocation. In mobile and resource-constrained environments such as emergency response systems or mobile opportunistic networks, community detection methods must balance accuracy with computational efficiency. In this work, we propose a novel approach that uncovers community structures from a sparse representation of the original graph, addressing the need for lightweight and scalable algorithms in pervasive and mobile systems. Specifically, we apply Spielman–Srivastava spectral sparsification as a preprocessing step to reduce the number of edges while preserving the key spectral properties that underpin community structure. We then apply a modularity-optimizing genetic algorithm on the sparsified graph. Our experiments, conducted on both synthetic benchmarks and real-world networks, demonstrate that the proposed method, namely SSGA, achieves competitive or superior accuracy compared to state-of-the-art baselines, even under aggressive sparsification. We also analyze the cumulative computational complexity of the approach and provide an optimized implementation based on truncated spectral decomposition and parallel genetic operations. The results confirm that SSGA is not only accurate and robust but also computationally efficient, making it particularly well-suited for pervasive and mobile scenarios where time, energy, and connectivity are limited.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"113 ","pages":"Article 102106"},"PeriodicalIF":3.5,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144865802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-08-01Epub Date: 2025-06-05DOI: 10.1016/j.pmcj.2025.102077
Liliana Martirano , Lucio La Cava , Andrea Tagarelli
Social media have widely been recognized as a valuable proxy for investigating users’ opinions by echoing virtual venues where individuals engage in daily discussions on a wide range of topics. Among them, climate change is gaining momentum due to its large-scale impact, tangible consequences for society, and enduring nature. In this work, we investigate the social debate surrounding climate emergency, aiming to uncover the fundamental patterns that underlie the climate debate, thus providing valuable support for strategic and operational decision-making. To this purpose, we leverage Graph Mining and NLP techniques to analyze a large corpus of tweets spanning seven years pertaining to the Conference of the Parties (COP), the leading global forum for multilateral discussion on climate-related matters, based on our proposed framework, named NATMAC, which consists of three main modules designed to perform network analysis, topic modeling and affective computing tasks. Our contribution in this work is manifold: (i) we provide insights into the key social actors involved in the climate debate and their relationships, (ii) we unveil the main topics discussed during COPs within the social landscape, (iii) we assess the evolution of users’ sentiment and emotions across time, and (iv) we identify users’ communities based on multiple dimensions. Furthermore, our proposed approach exhibits the potential to scale up to other emergency issues, highlighting its versatility and potential for broader use in analyzing and understanding the increasingly debated emergent phenomena.
{"title":"Unveiling user dynamics in the evolving social debate on climate crisis during the conferences of the parties","authors":"Liliana Martirano , Lucio La Cava , Andrea Tagarelli","doi":"10.1016/j.pmcj.2025.102077","DOIUrl":"10.1016/j.pmcj.2025.102077","url":null,"abstract":"<div><div>Social media have widely been recognized as a valuable proxy for investigating users’ opinions by echoing virtual venues where individuals engage in daily discussions on a wide range of topics. Among them, climate change is gaining momentum due to its large-scale impact, tangible consequences for society, and enduring nature. In this work, we investigate the social debate surrounding climate emergency, aiming to uncover the fundamental patterns that underlie the climate debate, thus providing valuable support for strategic and operational decision-making. To this purpose, we leverage Graph Mining and NLP techniques to analyze a large corpus of tweets spanning seven years pertaining to the Conference of the Parties (COP), the leading global forum for multilateral discussion on climate-related matters, based on our proposed framework, named NATMAC, which consists of three main modules designed to perform network analysis, topic modeling and affective computing tasks. Our contribution in this work is manifold: (i) we provide insights into the key social actors involved in the climate debate and their relationships, (ii) we unveil the main topics discussed during COPs within the social landscape, (iii) we assess the evolution of users’ sentiment and emotions across time, and (iv) we identify users’ communities based on multiple dimensions. Furthermore, our proposed approach exhibits the potential to scale up to other emergency issues, highlighting its versatility and potential for broader use in analyzing and understanding the increasingly debated emergent phenomena.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"112 ","pages":"Article 102077"},"PeriodicalIF":3.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144364551","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-08-01Epub Date: 2025-06-23DOI: 10.1016/j.pmcj.2025.102080
Juan M. Núñez V., Sebastián López Flórez, Juan M. Corchado, Fernando De la Prieta
Climate change represents a critical threat to global food security, affecting agricultural production and exacerbating the food crisis projected by the FAO for 2050. Soil recovery and the adoption of sustainable agricultural practices, such as organic farming, are essential to address this challenge. Smart organic farming improves soil quality, crop productivity, and water retention capacity. In this context, vermiculture, which utilizes Eisenia Foetida (red worms), plays a fundamental role. This article highlights how humus production through vermiculture has been significantly optimized through an Edge AIoT platform that integrates an agricultural recommendation system based on bio-inspired algorithms, an LSTM network for predicting humus and worm populations, and a control system to regulate variables such as temperature, humidity, and pH. The results show an increase in humus production from 37.58% to 87.88% and in the worm population from 35.5% to 83%. Vermicompost, obtained through the non-thermophilic biodegradation of organic waste by worms, acts as a crucial biofertilizer that sustainably increases crop yields and helps farmers adapt to environmental stresses, contributing to the Sustainable Development Goals (SDGs). Finally, seven experiments were conducted in which the Edge AIoT-based agricultural recommendation platform optimized the vermicomposting process, improving efficiency and productivity in humus production. This technological approach not only mitigates the impact of climate change but also supports the recovery of degraded soils and promotes sustainable agricultural practices essential for ensuring future food security.
{"title":"Edge AIoT-based agricultural recommendation platform to improve humus productivity in vermicomposting processes","authors":"Juan M. Núñez V., Sebastián López Flórez, Juan M. Corchado, Fernando De la Prieta","doi":"10.1016/j.pmcj.2025.102080","DOIUrl":"10.1016/j.pmcj.2025.102080","url":null,"abstract":"<div><div>Climate change represents a critical threat to global food security, affecting agricultural production and exacerbating the food crisis projected by the FAO for 2050. Soil recovery and the adoption of sustainable agricultural practices, such as organic farming, are essential to address this challenge. Smart organic farming improves soil quality, crop productivity, and water retention capacity. In this context, vermiculture, which utilizes Eisenia Foetida (red worms), plays a fundamental role. This article highlights how humus production through vermiculture has been significantly optimized through an Edge AIoT platform that integrates an agricultural recommendation system based on bio-inspired algorithms, an LSTM network for predicting humus and worm populations, and a control system to regulate variables such as temperature, humidity, and pH. The results show an increase in humus production from 37.58% to 87.88% and in the worm population from 35.5% to 83%. Vermicompost, obtained through the non-thermophilic biodegradation of organic waste by worms, acts as a crucial biofertilizer that sustainably increases crop yields and helps farmers adapt to environmental stresses, contributing to the Sustainable Development Goals (SDGs). Finally, seven experiments were conducted in which the Edge AIoT-based agricultural recommendation platform optimized the vermicomposting process, improving efficiency and productivity in humus production. This technological approach not only mitigates the impact of climate change but also supports the recovery of degraded soils and promotes sustainable agricultural practices essential for ensuring future food security.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"112 ","pages":"Article 102080"},"PeriodicalIF":3.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144472127","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-08-01Epub Date: 2025-07-18DOI: 10.1016/j.pmcj.2025.102091
Yang Han , Tasiu Muazu , Omaji Samuel , Shiyu Miao
Machine learning algorithms are powerful tools for analyzing data with several observations approximately equal to the number of predictors. However, the privacy of data owners may be revealed during the processes of analysis and mining in a distributed scenario. Today, federated learning is employed as the best paradigm for collaborative model training without disclosing the privacy of the data owners. Unfortunately, efficient client selection and incentive mechanisms need to provide for encouraging data sharing and analysis for constrained and non-constrained devices. Furthermore, trust in the system must be considered. To this end, this study proposes a federated blockchain-based incentive and selection mechanism for a federated learning system. Clients are selected using support vector machines (SVM), while the accuracy of SVM is improved by recursive feature elimination (RFE). A real-time incentive is provided to clients for collaborative learning using deep Q reinforcement learning, and an optimal incentive allocation policy is derived using the Markov decision process (MDP) framework. For miners’ selection, a proof of utility consensus is proposed using a sixteen-round addition game. Extensive simulations are conducted to evaluate the efficiency of the proposed system model. The performance of the proposed system is determined by its optimal statistical utility, system utility, and client utility, respectively. From the experimental results, the proposed SVM-RFE model outperform the existing algorithms. Additionally, security analysis is performed, which shows that the proposed system is safe against background knowledge attacks.
{"title":"A federated learning-based selection and incentive system using blockchain technology","authors":"Yang Han , Tasiu Muazu , Omaji Samuel , Shiyu Miao","doi":"10.1016/j.pmcj.2025.102091","DOIUrl":"10.1016/j.pmcj.2025.102091","url":null,"abstract":"<div><div>Machine learning algorithms are powerful tools for analyzing data with several observations approximately equal to the number of predictors. However, the privacy of data owners may be revealed during the processes of analysis and mining in a distributed scenario. Today, federated learning is employed as the best paradigm for collaborative model training without disclosing the privacy of the data owners. Unfortunately, efficient client selection and incentive mechanisms need to provide for encouraging data sharing and analysis for constrained and non-constrained devices. Furthermore, trust in the system must be considered. To this end, this study proposes a federated blockchain-based incentive and selection mechanism for a federated learning system. Clients are selected using support vector machines (SVM), while the accuracy of SVM is improved by recursive feature elimination (RFE). A real-time incentive is provided to clients for collaborative learning using deep Q reinforcement learning, and an optimal incentive allocation policy is derived using the Markov decision process (MDP) framework. For miners’ selection, a proof of utility consensus is proposed using a sixteen-round addition game. Extensive simulations are conducted to evaluate the efficiency of the proposed system model. The performance of the proposed system is determined by its optimal statistical utility, system utility, and client utility, respectively. From the experimental results, the proposed SVM-RFE model outperform the existing algorithms. Additionally, security analysis is performed, which shows that the proposed system is safe against background knowledge attacks.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"112 ","pages":"Article 102091"},"PeriodicalIF":3.5,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144722416","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-08-01Epub Date: 2025-05-23DOI: 10.1016/j.pmcj.2025.102079
Dr M. Anugraha , Dr S. Selvin Ebenezer , Dr S. Maheswari
A Mobile Ad-Hoc Network (MANET) represents a set of wireless networks that create the network without requiring centralized control. Moreover, the MANET serves as an effectual communication network but is impacted by security issues. MANET intrusion detection constantly monitors network traffic for potential intrusions. Still, it requires network nodes for analyzing, and processing the data, which leads to the highest processing charge. For solving such difficulties, the EIK Herd Anaconda Optimization (EHAO)-based routing, and EHAO-trained Deep Kronecker Network (EHAO-DKN) for intrusion detection is devised in this paper. The MANET simulation is the prime step for attaining the routing. The proposed EHGAO with the fitness factors are considered in the routing. The intrusion presence in the MANET is detected at the Base Station (BS), where the Z-score normalization is applied to normalize the log data. The Wave Hedges metric effectively selects the relevant features, and the EHAO-DKN detects the intrusion. Furthermore, the EHAO-based routing obtained the optimal trust, energy, and delay of 85.30, 2.905 J, and 0.608 mS as well as the accuracy, sensitivity, and specificity of 92.40 %, 91.50 %, and 91.50 % are achieved by the EHAO-DKN-based intrusion detection.
{"title":"Hybrid elk herd green anaconda-based multipath routing and deep learning-based intrusion detection In MANET","authors":"Dr M. Anugraha , Dr S. Selvin Ebenezer , Dr S. Maheswari","doi":"10.1016/j.pmcj.2025.102079","DOIUrl":"10.1016/j.pmcj.2025.102079","url":null,"abstract":"<div><div>A Mobile Ad-Hoc Network (MANET) represents a set of wireless networks that create the network without requiring centralized control. Moreover, the MANET serves as an effectual communication network but is impacted by security issues. MANET intrusion detection constantly monitors network traffic for potential intrusions. Still, it requires network nodes for analyzing, and processing the data, which leads to the highest processing charge. For solving such difficulties, the EIK Herd Anaconda Optimization (EHAO)-based routing, and EHAO-trained Deep Kronecker Network (EHAO-DKN) for intrusion detection is devised in this paper. The MANET simulation is the prime step for attaining the routing. The proposed EHGAO with the fitness factors are considered in the routing. The intrusion presence in the MANET is detected at the Base Station (BS), where the Z-score normalization is applied to normalize the log data. The Wave Hedges metric effectively selects the relevant features, and the EHAO-DKN detects the intrusion. Furthermore, the EHAO-based routing obtained the optimal trust, energy, and delay of 85.30, 2.905 J, and 0.608 mS as well as the accuracy, sensitivity, and specificity of 92.40 %, 91.50 %, and 91.50 % are achieved by the EHAO-DKN-based intrusion detection.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"112 ","pages":"Article 102079"},"PeriodicalIF":3.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144241479","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号