Time series datasets often exhibit complex and diverse characteristics, including seasonal fluctuations, long-term trends, and irregular variations, which makes it difficult for a single, fixed machine learning model to achieve consistently accurate forecasts across different domains. To address this challenge, we introduce a neuroevolutionary framework for time series forecasting that integrates Evolutionary Algorithms with a Transformer Encoder designed for time series forecasting. This hybrid approach leverages the Transformer’s ability to capture short- and long-term dependencies while using evolutionary search to adjust the model’s hyperparameters to the unique dynamics of each dataset. In this work, we adopt an encoder-only Transformer architecture as a design choice motivated by computational efficiency and the nature of time series forecasting tasks, where the prediction does not require a full encoder–decoder structure. Its optimization is guided by evolutionary algorithms, namely Genetic Algorithm and one of the Estimation of Distribution Algorithms; the Population-Based Incremental Learning. The effectiveness of the approach is validated by benchmarking against ten publicly available univariate time series datasets that cover various patterns and structures. The results demonstrate the notable performance of the proposed model in terms of forecast precision and robustness, highlighting its ability to generalize across various time series scenarios.

As the number of EVs on the road increases, new and innovative charging solutions are needed to support their energy demand, especially in areas lacking infrastructure, in order to alleviate range anxiety in EV owners. Mobile Charging Stations (MCSs) have emerged as a viable solution, offering on-demand charging by traveling to areas of need. Researchers have explored several allocation mechanisms for MCS dispatch, including energy auctions, which perform well under the assumption that all participants adhere to protocol. However, customer no-shows—a common challenge in retail services—can significantly disrupt MCS operations. A high rate of EV no-shows leads to stagnation and resource wastage, ultimately reducing the efficiency of charging service delivery. To address this, we propose a reputation management system that enables MCS fleets to prioritize well-behaved customers, ensuring better resource utilization. Additionally, EVs are incentivized to form coalitions with their neighbors, submitting aggregate requests backed by a deposit to promote responsible behavior. The proposed coalition- and reputation-based energy auction system is evaluated across various quality-of-service metrics to assess its effectiveness.

Mobile crowdsensing has become a key paradigm in IoT technology. It allows utilization of built-in sensors of participants’ mobile devices to provide sensing as a service. However, ensuring data reliability remains a critical challenge due to the presence of erroneous, or malicious data, which can compromise the integrity and accuracy of crowdsensed information. This paper addresses this challenge by introducing anchor nodes to enhance the quality of data fusion in mobile crowdsensing. Anchor nodes are identified through a systematic process incorporating a dynamic reputation adjustment mechanism and weighted data fusion. Validation using a real sensor dataset demonstrates the effectiveness of the proposed approach. It shows an improvement in data fusion quality. The results highlight the importance of anchor-based trust management (TM) in enhancing data fusion quality in crowdsensing.

Modern edge cloud platforms must efficiently deploy and route containerized microservice DAGs under strict latency and cost constraints, while adapting to rapidly changing workloads and infrastructure states. Deep Reinforcement Learning (DRL) schedulers adapt well to dynamics but often lack semantic awareness of service intent and task dependencies, resulting in suboptimal decisions in unseen scenarios. To overcome these limitations, we introduce a Retrieval-Augmented Generation-assisted DRL (RAG-DRL) framework that integrates a lightweight DRL agent with a graph-based RAG module powered by a partially frozen LLM. A dynamic memory graph encodes contextual information such as node resources, network latencies, and SLA feedback. The LLM retrieves relevant historical deployments and current service intents to generate soft placement plans and reward estimates, which guide the DRL agent. These priors accelerate convergence, improve generalization across diverse conditions, and ensure real-time responsiveness. Evaluations on a realistic urban-scale edge cloud testbed confirm that RAG-DRL significantly reduces SLA violations, end-to-end latency, and resource imbalance, outperforming modern container-based schedulers. Our framework converges faster, maintains latency below 65 ms on scale, limits SLA violations to 12% under heavy load, and achieves 90 % resource utilization with balanced distribution.

In edge computing environments, efficient computation resource management is crucial for optimizing service allocation to hosts in the form of containers. These environments experience dynamic user demands and high mobility, making traditional static and heuristic-based methods inadequate for handling such complexity and variability. Deep Reinforcement Learning (DRL) offers a more adaptable solution, capable of responding to these dynamic conditions. However, existing DRL methods face challenges such as high reward variability, slow convergence, and difficulties in incorporating user mobility and rapidly changing environmental configurations. To overcome these challenges, we propose a novel DRL framework for computation resource optimization at the edge layer. This framework leverages a customized Markov Decision Process (MDP) and Proximal Policy Optimization (PPO), integrating a Graph Convolutional Transformer (GCT). By combining Graph Convolutional Networks (GCN) with Transformer encoders, the GCT introduces a spatio-temporal reward-shaping mechanism that enhances the agent's ability to select hosts and assign services efficiently in real time while minimizing the overload. Our approach significantly enhances the speed and accuracy of resource allocation, achieving, on average across two datasets, a 30% reduction in convergence time, a 25% increase in total accumulated rewards, and a 35% improvement in service allocation efficiency compared to standard DRL methods and existing reward-shaping techniques. Our method was validated using two real-world datasets, MOBILE DATA CHALLENGE (MDC) and Shanghai Telecom, and was compared against standard DRL models, reward-shaping baselines, and heuristic methods. • Proposing a DRL framework that integrates reward shaping for resource management. • Introducing a novel MDP design that considers the dynamic nature of the users. • Presenting a novel reward shaping mechanism, incorporating GCN and transformers.

This paper addresses the challenges of Last Mile Delivery (LMD) in crowdsourced platforms under time and budget constraints. LMD service providers face a continuous increase in demand with limited resources, such as workers and budgets. With tasks that vary in urgency, limited resources often lead to task failures. Furthermore, the increasing number of requesters and workers complicates the governance of the LMD platform, making it difficult to maintain credibility, integrity, and transparency. Current LMD solutions generally focus on route optimization and service acceleration, overlooking the challenge of combined time-critical and budget-constrained tasks. This creates the need for an agile, accelerated, and decentralized LMD framework that can 1) handle tasks instantly upon submission, 2) motivate successful completion based on urgency, and 3) compensate for tasks with deficient budgets to fortify long-term success and avoid failure due to constrained resources. For this purpose, this paper introduces the first Blockchain Hybrid Crowdsourced Auction-based LMD framework (BHCA-LMD), which is designed to effectively manage time and budget constraints. Built on-chain for transparency and decentralization, the framework uses drones and ground vehicles as a hybrid delivery mode for their capabilities in speeding up deliveries. BHCA-LMD incorporates an auctioning system to alleviate failures due to limited budgets, with a feedback mechanism to prevent entities from abusing the platform’s profit for their benefit, ensuring the credibility and reliability of the framework. The evaluation shows an on-time task completion rate of up to 74% with the presence of malicious behavior and a 15% lower gas consumption compared to the closest benchmark.

During a Remote Robotic Surgery (RRS) session, multimodal data traffic with different requirements is initiated. In order to achieve a cost-effective deployment of such a system, it is crucial to tailor resource allocation policies based on the different quality of service (QoS) requirements of each data traffic. In this paper, we focus on resource allocation in a 5G-enabled tactile Internet RRS system using network function virtualization (NFV). In particular, we investigate the joint placement and scheduling of Virtualized Network Functions (VNFs) in a RRS system under both deterministic and dynamic settings. An integer linear program (ILP) is used to formulate the problem. Due to its high computational complexity, we first propose an efficient greedy algorithm to solve the ILP under deterministic settings. Simulation results show that our proposed algorithm achieves near-optimal performance and outperforms the benchmark solutions in terms of cost and admission rate. It can reduce cost by up to 37% and improve admission rate by up to 34% while satisfying both latency and reliability constraints. Furthermore, our results show that modeling the multimodal data traffic by multiple VNF Forwarding Graphs (VNF-FGs) with different QoS requirements achieves a significant gain in terms of cost and acceptance rate compared to modeling it by a single VNF-FG with the most stringent requirements. We then considered a dynamic environment where latency variations and traffic arrivals may occur over time. Using the principles of optimal stopping theory, we propose an adaptive dynamic scheduler that is capable of triggering recalculations of the existing optimal solution based on the observed cumulative number of traffic arrivals and latency violations without the need for predictions. Our proposed optimal scheduler minimizes the migration cost compared to other schedulers.

Artificial intelligence (AI) has the potential to revolutionize healthcare by automating the detection and classification of events and anomalies. In the scope of this work, events and anomalies are abnormalities in the patient’s data, where the former are due to a medical condition, such as a seizure or a fall, and the latter are erroneous data due to faults or malicious attacks. AI-based event and anomaly detection (EAD) and their classification can improve patient outcomes by identifying problems earlier, enabling more timely interventions while minimizing false alarms caused by anomalies. Moreover, the advancement of Medical Internet of Things (MIoT), or wearable devices, and their high processing capabilities facilitated the gathering, AI-based processing, and transmission of data, which enabled remote patient monitoring, and personalized and predictive healthcare. However, it is fundamental in healthcare to ensure the explainability of AI systems, meaning that they can provide understandable and transparent reasoning for their decisions. This article proposes an online EAD approach using a lightweight autoencoder (AE) on the MIoT. The detected abnormality is explained using KernelSHAP, an explainable AI (XAI) technique, where the explanation of the abnormality is used, by an artificial neural network (ANN), to classify it into an event or anomaly. Intensive simulations are conducted using the Medical Information Mart for Intensive Care (MIMIC) data set for various physiological data. Results showed the robustness of the proposed approach in the detection and classification of events, regardless of the percentage of the present anomalies.

Multiagent deep reinforcement learning (MDRL) is a promising research area in which agents learn complex behaviors in cooperative or competitive environments. However, MDRL comes with several challenges that hinder its usability, including sample efficiency, curse of dimensionality, and environment exploration. Recent works proposing federated reinforcement learning (FRL) to tackle these issues suffer from problems related to model restrictions and maliciousness. Other proposals using reward shaping (RS) require considerable engineering and could lead to local optima. In this article, we propose a novel Blockchain-assisted multiexpert demonstration cloning (MEDC) framework for MDRL. The proposed method utilizes expert demonstrations in guiding the learning of new MDRL agents, by suggesting exploration actions in the environment. A model sharing framework on Blockchain is designed to allow users to share their trained models, which can be allocated as expert models to requesting users to aid in training MDRL systems. A Consortium Blockchain is adopted to enable traceable and autonomous execution without the need for a single trusted entity. Smart Contracts are designed to manage users and models allocation, which are shared using IPFS. The proposed framework is tested on several applications and is benchmarked against existing methods in FRL, RS, and imitation learning-assisted RL. The results show the outperformance of the proposed framework in terms of learning speed and resiliency to faulty and malicious models.

This article addresses the problem of multiple source localization (MSL) using the Internet of Things (IoT) sensors. MSL entails determining the locations of multiple unknown sources by fusing sensory data within a designated Area of Interest (AoI). Existing solutions suffer from limitations, such as increased algorithmic complexity, as the number of sources increases and degraded performance in sparse sensor placement scenarios. This article proposes a novel source-independent approach resilient to sparse sensor placements based on conditional generative adversarial networks (cGANs) and image processing-based peak finding with subpixel peak refinement to address the MSL problem. The proposed approach formulates the MSL problem into two subproblems: 1) image-to-image translation and 2) 2-D peak finding. The cGAN translates the raw measurement data to an intensity field through image-to-image translation. Then, a peak-finding algorithm based on persistent homology with subpixel peak refinement is applied to localize the unknown sources accurately. The proposed approach is tested through radioactive source localization experiments, benchmark comparisons, and adaptability evaluation in unseen environments.

Deep Reinforcement Learning (DRL) has emerged as a powerful paradigm for solving complex problems. However, its full potential remains inaccessible to a broader audience due to its complexity, which requires expertise in training and designing DRL solutions, high computational capabilities, and sometimes access to pre-trained models. This necessitates the need for hassle-free services that increase the availability of DRL solutions to a variety of users. To enhance the accessibility to DRL services, this paper proposes a novel blockchain-based crowdsourced DRL as a Service (DRLaaS) framework. The framework provides DRL-related services to users, covering two types of tasks: DRL training and model sharing. Through crowdsourcing, users could benefit from the expertise and computational capabilities of workers to train DRL solutions. Model sharing could help users gain access to pre-trained models, shared by workers in return for incentives, which can help train new DRL solutions using methods in knowledge transfer. The DRLaaS framework is built on top of a Consortium Blockchain to enable traceable and autonomous execution. Smart Contracts are designed to manage worker and model allocation, which are stored using the InterPlanetary File System (IPFS) to ensure tamper-proof data distribution. The framework is tested on several DRL applications, proving its efficacy.

Electric vehicle (EV) charging infrastructure development is one of the key aspects of the electrification of transportation systems. However, incorporating EV public charging facilities has practical and financial concerns, especially in developing economies. Recently, mobile public charging stations (MPCS) have been considered an alternate, practical, commercially scalable solution. Present research on MPCS operation algorithms assumes centralized algorithm execution and does not offer any features related to consumer data security. Moreover, they do not exploit the multi-port charging facility of MPCS, which is more economical for scheduling services. In this paper, a comprehensive framework for the decentralized execution of the MPCS operation algorithm for EV charging service using a hybrid cloud-edge server architecture is proposed. The decentralized execution enhances EV data security and reduces load on the cloud server, data transmission, and storage charges. Four vehicle routing algorithms, including exact integer programming (IP), Greedy, Greedy+2Opt heuristic, and meta-heuristic Tabu search algorithms, are applied to test the efficacy of the proposed framework. From the studies, the proposed framework with Greedy+2Opt performed well concerning scalability. Compared to one-to-one MPCS-EV charging, the proposed one-to-many charging framework offered a significant reduction in the transportation cost and the required number of MPCSs for a given energy delivered. Further, for a given MPCS battery capacity and service time window, the framework delivered higher energy compared to one-to-one service model.

Reducing the battery charging time of an electric vehicle (EV) is one of the key factors to boost the widespread adoption of EVs. The commercial, off-board high power, dc fast charging station need high initial investment and maintenance cost. On the other hand, the standard on-board type-1 and type-2 ac chargers with$3.3~kW$to$19~kW$need long time to charge. This paper proposes a combinatory ac and dc charging approach to increase the charging rate of EV batteries. The proposed combinatory charging approach provides a technique to charge EV battery from the on-board type-2 ac charger and drivetrain integrated dc charger. For drivetrain integrated dc charging, a dc input port$(N (+),O(-))$is formed using the neutral of the EV motor winding$(N)$and negative rail of the drivetrain inverter$(O)$. Through this dc input port, power from the renewable energy source-based dc microgrids, solar rooftops and other EV battery can be accepted for charging. The EV drivetrain inverter is controlled as an integrated interleaved dc-dc converter (IDC) to receive power from dc sources with EV motor windings reutilized as filter inductors. The control scheme for regulating the voltage across common dc-link accepting power from type-2 ac charger and integrated interleaved dc charger is presented. The performance analysis of EV motor and drivetrain integrated DC charger is validated through Finiet Element methods (FEM) co-simulation using Ansys Maxwell and Simplorer. A scaled experimental prototype is developed to validate the proposed combined ac and dc charging approach.

This paper tackles the problem of frequent disconnections in the Quality-of-Service Optimized Link State Routing (QoS-OLSR) protocol due to mobility in mesh networks. Different routing protocols are proposed for mesh networks, such as Better Approach to Mobile Ad-hoc Networking (BATMAN), Optimized Link State Routing (OLSR), and QoS-OLSR 1.0. QoS-OLSR 1.0 improves the performance compared to the other protocols by incorporating cluster head formation and a Quality-of-Service metric used in cluster head selection. However, it still does not mitigate the impact of mobility in the formulated QoS metric. In this work, we propose and implement a cluster-based QoS-OLSR protocol for mesh networks that accounts for mobility. The work proposes a global QoS metric for nodes, calculated as the average quality of all neighbors’ links. Also, a relative QoS metric is proposed and calculated based on the direct link with the neighbor and its global QoS. Based on the relative QoS, nodes in the network select cluster heads. In case of disconnection from the selected cluster head, the nodes join an existing cluster head to recover from disconnection. Cluster heads in the proposed protocol are responsible for determining Multi-point relays (MPRs) to 2-hop and 3-hop away cluster heads to enhance connectivity. The proposed protocol is implemented as a Linux-compatible protocol for emulations using Mininet-WiFi and compared to BATMAN and QoS-OLSR 1.0. The proposed protocol outperforms the benchmark protocols in terms of throughput, Packet Delivery Ratio, and Round Trip Time with low and high mobility nodes part of the network.

To cope with the increasing growth of last-mile delivery, crowdsourcing last-mile delivery has been adopted as a flexible and cost-efficient way to deliver parcels quickly and efficiently. However, some potential downsides to crowdsourcing last-mile delivery include concerns about safety, reliability, and transparency. Therefore, blockchain has been adopted to promote transparency in the last-mile delivery process. Despite the impact of a worker’s reputation on task completion, existing works do not offer a traceable and transparent reputation metric for lastmile delivery workers. This paper proposes a blockchain-based framework for reputation management in crowdsourced last-mile delivery. The proposed framework is designed as smart contracts that maintain and update crowdsourced workers’ information, mainly reputation, in a transparent and traceable manner. In addition, the framework allows requesters to create their delivery tasks and workers to get allocated available tasks. The proposed framework uses Solidity to interact with smart contracts for requesters and workers. The cost analysis demonstrates the proposed framework’s feasibility and cost efficiency.

Mobile Crowdsensing (MCS) is a sensing paradigm where individuals collectively perform a sensing task using their smart devices . Sensing tasks can be classified as one-time or continuous. In the former, only one-time readings from the devices of the recruited workers are needed. However, in continuous sensing tasks, collecting information continuously during a specific period is required. Due to workers’ mobility, ensuring a satisfactory level of the quality of information (QoI) of the sensing data is challenging since workers may leave the Area of Interest (AoI) before the task is over, causing low area coverage. Current existing recruitment systems for continuous sensing rely on historical mobility traces to recruit the group of workers. However, since workers’ mobility patterns are dynamic in nature, thus, a real-time prediction of their locations in the AoI needs to be considered to ensure that the required value of QoI is achieved. Hence, in this work (1) machine learning is employed to predict users’ location during the sensing period and (2) a novel recruitment system is proposed for continuous sensing tasks. The simulation results, using a real-life trajectories dataset, show the efficacy of the proposed solution when compared to benchmark.

In target localization applications, readings from multiple sensing agents are processed to identify a target location. The localization systems using stationary sensors use data fusion methods to estimate the target location, whereas other systems use mobile sensing agents (UAVs, robots) to search the area for the target. However, such methods are designed for specific environments, and hence are deemed infeasible if the environment changes. For instance, the presence of walls increases the environment’s complexity and affects the collected readings and the mobility of the agents. Recent works explored deep reinforcement learning (DRL) as an efficient and adaptable approach to tackle the target search problem. However, such methods are either designed for single-agent systems or for noncomplex environments. This work proposes two novel multiagent DRL models for target localization through search in complex environments. The first model utilizes proximal policy optimization, convolutional neural networks, Convolutional AutoEncoders to create embeddings, and a shaped reward function using breadth first search to obtain cooperative agents that achieve fast localization at low cost. The second model improves the first model in terms of computational complexity by replacing the shaped reward with a simple sparse reward, subject to the availability of Expert Demonstrations. Expert demonstrations are used in Demonstration Cloning, a novel method that utilizes demonstrations to guide the learning of new agents. The proposed models are tested on a scenario of radioactive target localization, and benchmarked with existing methods, showing efficacy in terms of localization time and cost, in addition to learning speed and stability.

Workers recruitment remains a significant issue in Mobile Crowdsourcing (MCS), where the aim is to recruit a group of workers that maximizes the expected Quality of Service (QoS). Current recruitment systems assume that a pre-defined pool of workers is available. However, this assumption is not always true, especially in cold-start situations, where a new MCS task has just been released. Additionally, studies show that up to 96% of the available candidates are usually not willing to perform the assigned tasks. To tackle these issues, recent works use Online Social Networks (OSNs) and Influence Maximization (IM) to advertise about the desired MCS tasks through influencers, aiming to build larger pools. However, these works suffer from several limitations, such as 1) the lack of group-based selection methods when choosing influencers, 2) the lack of a well-defined worker recruitment process following IM, 3) and the non-dynamicity of the recruitment process, where the workers who refuse to perform the task are not substituted. In this paper, an Influence- and Interest-based Worker Recruitment System (IIWRS), using OSNs, is proposed. The proposed system has two main components: 1) an MCS-, group-, and interest-based IM approach, using a Genetic Algorithm, to select a set of influencers from the network to advertise about the MCS tasks, and 2) a dynamic worker recruitment process which considers the social attributes of workers, and is able to substitute those who do not accept to perform the assigned tasks. Empirical studies are performed using real-life datasets, while comparingIIWRSwith existing benchmarks.

Driven by privacy concerns and the promise of Deep Learning, researchers have devoted significant effort to exploring the applicability of Machine Learning (ML). In the domains of communication, network, and service management, ML-based decision-making solutions are eagerly sought to replace traditional model-driven approaches, addressing the growing complexity and heterogeneity of modern systems. In this context, Federated Learning (FL) has gained increasing interest as a decentralized approach that overcomes the limitations of centralized systems for data analysis.

The “black-box” nature of artificial intelligence (AI) models has been the source of many concerns in their use for critical applications. Explainable Artificial Intelligence (XAI) is a rapidly growing research field that aims to create machine learning models that can provide clear and interpretable explanations for their decisions and actions. In the field of cybersecurity, XAI has the potential to revolutionize the way we approach network and system security by enabling us to better understand the behavior of cyber threats and to design more effective defenses. In this survey, we review the state of the art in XAI for cybersecurity and explore the various approaches that have been proposed to address this important problem. The review follows a systematic classification of cybersecurity threats and issues in networks and digital systems. We discuss the challenges and limitations of current XAI methods in the context of cybersecurity and outline promising directions for future research.

Fog computing empowers the internet of vehicles (IoV) paradigm by offering computational resources near the end users. In this dynamic paradigm, users tend to move in and out of the range of fog nodes which has implications for the quality of service of the vehicular applications. To cope with these limitations, scholars addressed forming federations of fog providers for task offloading purposes. Nonetheless, a few challenges remain a burden for the formation of the federations. The formation mechanisms used to structure the federations of providers are still not fully stable. This causes a problem because a structureless federation can lead to an underperforming infrastructure. Furthermore, most of the literature ignored the honesty metrics of the providers and how trustworthy they are in allocating the agreed-upon resources for processing the tasks. Moreover, adopting a central reputation mechanism is questionable in terms of reliability due to many complications including the lack of consensus. In this work, we develop a Blockchain-based reputation mechanism for assisting the formation of fog federations for IoV applications. Our mechanism comprises on-chain smart contracts for storing and manipulating the providers’ reputations, and an off-chain Hedonic-based formation process that considers the parameters extracted from the chain to build the federations. We develop smart contracts using Solidity and deploy them on the Ethereum Blockchain. We test our mechanism using the EUA dataset as a proof of concept and compare it to other works in the literature. The results obtained show that our approach is able to enhance the overall payoff and quality of service in the IoV paradigm.

Remote robotic surgery is one of the most interesting Tactile Internet (TI) applications. It has a huge potential to deliver healthcare services to remote locations. Moreover, it provides better precision and accuracy to diagnose and operate on patients. Remote robotic surgery requires ultra-low latency and ultra-high reliability. The aforementioned stringent requirements do not apply for all the multimodal data traffic (i.e., audio, video, and haptic) triggered during a surgery session. Hence, customizing resource allocation policies according to the different quality-of-service (QoS) requirements is crucial in order to achieve a cost-effective deployment of such system. In this paper, we focus on resource allocation in a softwarized 5G-enabled TI remote robotic surgery system through the use of Network Functions Virtualization (NFV). Specifically, this work is devoted to the joint placement and scheduling of application components in an NFV-based remote robotic surgery system, while considering haptic and video data. The problem is formulated as an integer linear program (ILP). Due to its complexity, we propose a greedy algorithm to solve the developed ILP in a computationally efficient manner. The simulation results show that our proposed algorithm is close to optimal and outperforms the benchmark solutions in terms of cost and admission rate. Furthermore, our results demonstrate that splitting application traffic to multiple VNF-forwarding graphs (VNF-FGs) with different QoS requirements achieves a significant gain in terms of cost and admission rate compared to modeling the whole application traffic with one VNF-FG having the most stringent requirements.

This paper tackles the problem of cluster head and Multi-Point Relay (MPR) selection for the Quality-of-Service Optimized Link State Routing (QoS-OLSR) protocol in mesh networks. Mesh networks emerged to extend the connectivity of users and to enable the exchange of messages in an adhoc manner. Mesh networks apply routing protocols such as Better Approach to Mobile Adhoc Networking (BATMAN) and Optimized Link State Routing (OLSR) where the latter was deemed more suitable for high mobility networks. Research efforts have proposed extending the OLSR protocol to a QoS-OLSR protocol that takes into consideration metrics such as bandwidth, connectivity, remaining energy, and velocity in QoS computation. The proposed QoS is used for cluster head and MultiPoint Relays (MPRs) selection. In this work, we propose and implement a cluster-based QoS-OLSR protocol for mesh networks. A QoS metric that aggregates the average estimated transmission count and the reachability of a node is defined. The proposed protocol is implemented by modifying the Linux OLSR distribution to perform emulations using meshnet-lab under different mobility models. In a static network, the proposed protocol outperforms BATMAN and OLSR in terms of Packet Delivery Ratio (PDR) and Round Trip Time (RTT). In addition, the proposed protocol outperforms BATMAN in terms of RTT and provides comparable results to OLSR in networks with low and high mobility nodes.

In this paper, a blockchain-based management platform, PackChain, for last-mile delivery is proposed. The growing popularity of online shopping has put immense pressure on the supply-chain industry, especially on last-mile delivery. The available solutions suffer from high cost, and lack of transparency. Therefore, assuring traceability of users' actions has become a critical requirement to establish trust between parties. The proposed PackChain framework uses the Ethereum blockchain to offer a crowdsourcing platform for last-mile delivery with autonomous and transparent processes. PackChain provides all the core functions needed for the delivery framework to operate through smart contracts such as managing user information, accepting offers, verifying transactions, and handling payments. In addition, the framework relies on proofs of delivery as an arbitration mechanism between users to release or hold funds. The proposed framework is implemented using Solidity and Web3.js to interact between clients and carriers. The emulation results demonstrate the feasibility and cost-efficiency of the proposed solution11The full code of the smart contract and the related logic is also made publicly available on Github..

Today, online media applications are an important source of information. People are creating and sharing more information than ever before around the world. Being provided by unreliable sources, some news can be misleading. In fact, the assessment of the correctness of the news can be region related. In other words, news can be true in a specific region while fake in another. Existing proposed solutions for fake news detection developed in centralized platforms are not considering the location from where the news gets announced, but they are focused more on the news content. In this paper, a region-based distributed fake news detection framework is proposed. The framework is deployed in a mobile crowdsensing (MCS) environment where a set of workers responsible for collecting news are selected based on their availability in a specific region. The selected workers share the news to the nearest edge node, where the pre-processing and detection of fake news are executed locally. The detection process uses a pre-trained BERT model where it achieved an accuracy of 91 %.

The increasing number of Internet of Things (IoT) devices and low-cost sensors have facilitated developments in large-scale monitoring applications. However, the accuracy of low-cost sensors remains questionable. Monitoring applications, such as environmental monitoring, try to detect “interesting” data points or patterns, known as anomalies, that do not conform to the norm. These include erroneous data caused by hardware failures or malicious attacks, and nonerroneous data due to unexpected phenomenon, caused by events, such as unexpected high traffic volume. Traditionally, IoT devices collect raw data and periodically upload them to the cloud for processing, which includes anomaly detection. However, the increasing processing capabilities of IoT devices have made the on-device anomaly detection possible in an online and real-time manner. In this article, multivariate long short-term memory (LSTM) autoencoder is proposed for anomaly and event detection in IoT devices. In addition, the proposed approach integrates smart inference, based on a game-theoretical approach, which dynamically changes the period of detection based on the stability of the data, aiming to optimize power consumption and elongate the lifetime of the device. The proposed anomaly and event detection model was simulated and implemented on an STM32H743 Nucleo board, and results show the robustness of the model regardless of the number of anomalies and events present.

The proliferation of electric vehicles (EVs), owing to their advantages over internal combustion engine vehicles, has introduced many challenges, due to the lack of charging infrastructure that can handle such a large surge of EVs. Therefore, alternative feasible charging solutions, such as EV-to-Grid (V2G) and EV-to-EV (V2V) charging have gained prominence thanks to the bidirectional charger. However, there are several challenges hindering the adoption of V2V energy sharing solutions. The existing frameworks that detail the main aspects in energy management protocols emphasize solely on the EV integration with the grid, with the assumption of the grid availability and capability of supporting V2V energy sharing. In this article, a novel holistic energy management framework for an efficient V2V energy sharing is proposed. The proposed framework offers a complete overview of the different stages in the V2V charging problem and introduces possible solutions for each stage and its integration with other stages to form a comprehensive V2V solution, that is not only cost-effective but also maximizes user satisfaction and social welfare, while simultaneously fulfills the highest number of energy demands.

Electric Vehicles (EVs) are being widely adopted to completely replace vehicles with internal combustion engines. However, the development of charging infrastructure to support the growing number of EVs has been lacking primarily due to the high cost of installation. Additionally, the immature and non-uniform deployment of charging stations has led to the absence of charging infrastructure in areas such as highways and rural areas. As a result, emerging concepts such as, EV-to-Home (V2H), EV-to-Grid (V2G) and EV-to-EV (V2V) charging have gained prominence. In this paper, the V2V energy sharing concept is exploited, where an intelligent and comprehensive framework to manage and allocate energy between EVs is proposed. This work presents a realistic modeling of the V2V energy sharing problem and proposes a two-layer matching approach that can efficiently match the EVs. The proposed approach not only optimizes the cost, but also the time, system energy efficiency, user satisfaction, and social welfare; hence making the approach more realistic and inclusive. Gale-Shapley is utilized to produce stable matchings, in the first layer, whereas a user-satisfaction model is devised to ensure realistic matchings, in the second layer. Additionally, a real-life dataset is developed from commercially available EVs and the performance of the system is evaluated using this dataset, in addition to realistic parameters derived from real-life data. The proposed approach is compared with a benchmark, where the results show that efficient, realistic, and effective V2V matches are achievable, paving the way for the adoption of such framework.

We argue in this paper that the role of the cloud should be reshaped from being a passive virtual market to become an active platform for monetizing data. The objective is to enable the cloud to be an active platform that can help data providers reach a wider set of data consumers. This will allow these consumers to be exposed to a larger variety of data that benefits data analytic applications. To achieve this vision, we propose a novel game theoretical model, which consists of a mix of cooperative and competitive strategies. The players of the game are the data providers, cloud platform, and cloud users. The strategies of the players are modeled using the two-sided market theory that takes into consideration the network effects (externalities) among the players. Simulations conducted using Amazon and google clustered data show that the proposed model improves the total surplus of involved parties in terms of cloud resources provision and monetary profits compared to the current merchant model.

With the advent of mobile crowd sourcing (MCS) systems and its applications, theselectionof the right crowd is gaining utmost importance. The increasing variability in the context of MCS tasks makes the selection of not only the capable but also the willing workers crucial for a high task completion rate. Most of the existing MCS selection frameworks rely primarily on reputation-based feedback mechanisms to assess the level of commitment of potential workers. Such frameworks select workers having high reputation scores but without any contextual awareness of the workers, at the time of selection, or the task. This may lead to an unfair selection of workers who will not perform the task. Hence, reputation on its own only gives an approximation of workers’ behaviors since it assumes that workers always behave consistently regardless of the situational context. However, following the concept of cross-situational consistency, where people tend to show similar behavior in similar situations and behave differently in disparate ones, this work proposes a novel recruitment system in MCS based on behavioral profiling. The proposed approach uses machine learning to predict the probability of the workers performing a given task, based on their learned behavioral models. Subsequently, a group-based selection mechanism, based on the genetic algorithm, uses these behavioral models in complementation with a reputation-based model to recruit a group of workers that maximizes the quality of recruitment of the tasks. Simulations based on a real-life dataset show that considering human behavior in varying situations improves the quality of recruitment achieved by the tasks and their completion confidence when compared with a benchmark that relies solely on reputation.

Summary Communities that gather functionally identical or complementary cloud services aim to provide better visibility, efficiency, and market share. This paper investigates the issue of forming these communities in distributed decision‐making settings under incomplete information. By incomplete information, we mean only partial information about the individual performance of cloud services within communities and about how they will behave within these communities is available. Forming communities in these particular settings is still an open problem. Most of the existing models require real‐time global knowledge about the services and high computational complexity, which makes the community formation extremely hard and time‐consuming. In this paper, we propose a strategic Distributed Decision‐making Mechanism (DDM) that regulates the cloud services decision‐making process. DDM first generates an initial set of data based on information obtained from existing cloud services regarding their single and cooperative efficiency. By analyzing this set and on the basis of a distance function, the decision‐making mechanism with regard to which community to form is implemented as a decision profile of strategies and their expected utility computed in terms of computational efficiency. DDM efficiently and systematically helps 1) communities find appropriate cloud services to invite as new members and 2) single services find suitable communities to join. To evaluate the proposed mechanism, we performed experiments using real data including 142 users and 4,000 cloud services obtained from the CloudArmor, CloudHarmony, and WS‐DREAM datasets. The experimental results show that our algorithms outperform the existing solutions.

Thanks to the capabilities of the built-in sensors of smart devices, mobile crowd-sensing (MCS) has become a promising technique for massive data collection. In this paradigm, the service provider recruits workers (i.e., common people with smart devices) to perform sensing tasks requested by the consumers. To efficiently handle workers' recruitment and task allocation, several factors have to be considered such as the quality of the sensed data that the workers can deliver and the different tasks locations. This allocation becomes even more challenging when the MCS tries to efficiently allocate multiple tasks under limited budget, time constraints, and the uncertainty that selected workers will not be able to perform the tasks. In this paper, we propose a service computing framework for time constrained-task allocation in location based crowd-sensing systems. This framework relies on (1) a recruitment algorithm that implements a multi-objective task allocation algorithm based on Particle Swarm Optimization, (2) queuing schemes to handle efficiently the incoming sensing tasks in the server side and at the end-user side, (3) a task delegation mechanism to avoid delaying or declining the sensing requests due to unforeseen user context, and (4) a reputation management component to manage the reputation of users based on their sensing activities and task delegation. The platform goal is to efficiently determine the most appropriate set of workers to assign to each incoming task so that high quality results are returned within the requested response time. Simulations are conducted using real datasets from Foursquare1and Enron email social network.2Simulation results show that the proposed framework maximizes the aggregated quality of information, reduces the budget and response time to perform a task and increases the average recommenders' reputation and their payment.

With the tremendous advances in ubiquitous computing, and more specifically with mobile phones, mobile crowd sensing (MCS) has become an appealing part of IoT. However, MCS is vulnerable to multiple types of attacks which could be caused by external or internal adversaries. While external attacks, such as spoofing and jamming are addressed using network's security measures, internal attacks such as maliciously degrading the Quality of Service (QoS) of the tasks by intentionally submitting false reports, should be handled by the MCS systems. The current solutions aim to maximize the completion of tasks in selection based on the reputation or the credibility of the workers, but without consideration of internal attacks threat. This paper focuses on internal misbehaving act, similar to Sybil attacks, where users impersonate multiple identities to change the majority-voting result of the task or to selfishly maximize their profit with minimal costs, using multiple devices. These workers take advantage of the anonymity of MCS for privacy protection to pose the attack. This paper studies the need for a resilient approach which detects and eliminates misbehaving devices during workers' selection in MCS.

We consider community policing on the road with pervasive recording technologies such as dashcams and smartphones where citizens are actively volunteering to capture and report various threats to traffic safety to the police via mobile apps. This kind of novel community policing has recently gained significant popularity in Korea and India. In this work, we identify people's general attitude and concerns toward community policing on the road through an online survey. We then address the major concerns by building a mobile app that supports easy event capture/access, context tagging, and privacy preservation. Our two-week user study (n = 23) showed Roadwatch effectively supported community policing activities on the road. Further, we found that the critical factors for reporting are personal involvement and seriousness of risks, and participants were mainly motivated by their contribution to traffic safety. Finally, we discuss several practical design implications to facilitate community policing on the road.

Dashcams support continuous recording of external views that provide evidence in case of unexpected traffic-related accidents and incidents. Recently, sharing of dashcam videos has gained significant traction for accident investigation and entertainment purposes. Furthermore, there is a growing awareness that dashcam video sharing will greatly extend urban surveillance. Our work aims to identify the major motives and concerns behind the sharing of dashcam videos for urban surveillance. We conducted two survey studies (n=108, n=373) in Korea. Our results show that reciprocal altruism/social justice and monetary reward were the major motives and that participants were strongly motivated by altruism and social justice. Our studies have also identified major privacy concerns and found that groups with greater privacy concerns had lower altruism and justice motive, but had higher monetary motive. Our main findings have significant implications on the design of a dashcam video-sharing service.

With the high penetration of smart phones in our daily life and the increasing capabilities of these smart devices, a new type of applications has emerged. Lately, Opportunistic Mobile Social Networks (OMSN), which allow users to directly update their social networks while moving have seen the day. Typically, users require the Internet to share information on their social network and this produces various limitations. However, OMSN exploit opportunistic links between mobiles' of users with social similarities to share contents and updates, bypassing the need for an Internet connection. In this paper, a survey on OMSN applications and existing architectures is provided. Key challenges to the development of these networks are elaborated along with suggested solutions. An architecture for collaborative learning application using OMSN that tries to answer these challenges is presented. A prototype is implemented and deployed to validate the proposed solution. Clearly, OMSN can be an addition to the smart campus initiative.

Web services are loosely-coupled business applications willing to cooperate in distributed settings within different groups called communities. Communities aim to provide web services with better visibility, efficiency, market share and total payoff. A number of mechanisms and models have been recently proposed to aggregate web services and make them cooperate within their communities. However, forming optimal and stable communities as coalitions to maximize individual and group efficiency and income is yet to be addressed. In this paper, we propose an efficient community formation mechanism using cooperative game-theoretic techniques, particularly Shapley value, core, ε-core and convex games. We propose a mechanism for community membership requests and selections of web services under two scenarios: (1) there is only one community and many web services aim to join it; and (2) web services can join different established communities. The ultimate objective is to develop a mechanism for web services to form stable groups allowing them to maximize their efficiency and generate near-optimal (welfare-maximizing) communities where a taxation-based solution is being proposed. The theoretical and extensive experimental results show that our algorithms provide web services and community owners, in real-world-like environments, with applicable and near-optimal decision making mechanisms.

Cloud computing is an emerging paradigm that provides scalable computing and storage capabilities where resources are accessed on a pay-as-you-go basis. Software as a Service (SaaS) applications are hosted in the cloud and made available as services for tenants' organizations over a network. To achieve reusability in cloud computing, software and hardware resources are shared among multiple tenants. Conventional multitenant SaaS applications provide the same set of services for all tenants thus resulting in one-size-fits-all applications. However, as tenants may have different requirements, customizable SaaS solutions are needed. To accommodate evolving tenants' requirements, the SaaS instance should evolve systematically. In this paper, we present a multitenant single instance SaaS evolution platform based on Software Product Lines (SPLs). The platform specifies a set of evolution rules, based on feature modeling, that govern evolution decisions. We also present the early implementation phases of the proposed approach based on SPLs and Model Driven Architecture (MDA) concepts.

Nowadays, mobile applications/devices have become the trends, especially, when they were gradually shifted from basic communication services to supporting more sophisticated service provisioning. Mobile applications are usually very light, are nowadays likely to be often connected to the Internet, and can be used quite easily. However, these applications exhibit some challenges related to limited resources they have access to, including limited processing power, memory, storage size, battery power, and intermittent network connection. In fact, these considerations have to be taken seriously into consideration when developing mobile applications especially if those applications will be used for critical services, for example, to collect and report vital health data over a long period of time. In this paper, we study the use of mobile applications for monitoring patient's vital. Mobile devices, through an application, are connected to body-strapped biosensors to collect and synchronize these parameters with information systems. This synchronization should be done in such a way that the cost of synchronization is kept low and urgent readings are delivered as soon as possible. To optimize the synchronization process and reduce its cost, we propose and validate cost-oriented algorithms. A case study is developed to illustrate the applicability and effectiveness of our innovative techniques in making continuous monitoring an efficient process.

Cloud computing has emerged as a model for utility computing that promotes on-demand scalability, flexible application deployment and reuse. Software product lines (SPL) promote reusable application development for product families. As any computing system, cloud-based systems evolve to respond to changing clients' requirements. Cloud-based applications can be modeled as Software-as-a-Service (SaaS) families similar to the SPL products. As SPL development techniques rely on feature models to describe the commonality and variability of family member applications, such techniques can be used to model variability in SaaS. In this paper, we describe a unified and systematic framework for modeling cloud services in a vendor-neutral manner. In addition, we demonstrate the applicability of the variability framework for building and customizing SaaS multitenant applications. Our approach is based on a meta-model that formalizes the multiple views of service-oriented SaaS applications. A proof of concept tool that automatically generates multitenant applications (to adapt to changing requirements of tenants) is presented. Our approach facilitates development of cloud SaaS families in a systematic, consistent, and platform independent way.

During the last decade, social applications have witnessed a rapid growth in their use. Millions of people are utilising them on a daily basis in order to share their experience, information and to communicate with their family members and friends. Lately, these technologies have been used to foster

Despite the rapid growth of the mobile technology, mobile devices are still considered as resource constrained with limited battery. Same computations are awkward to be undertaken on these devices with limited processing capabilities. Other processes are costly in terms of battery consumption. Ideally, mobile applications will have the possibility to decide either to do a computation locally or remotely depending on the current device capabilities status. Making such decision is very challenging as many interrelated factors are to be considered (e.g. network connection, battery level, and processing capabilities). In this paper, we propose a framework that supports developers in implementing such smartness fitness within their mobile applications. This solution provides approaches in form of algorithms to instrument code of mobile applications to behave in smart way. Incorporating these algorithms will allow for on-the-fly decision of local versus remote computation using a calculated cost function. We conducted some experimental scenarios to evaluate the usability and effectiveness of our decision-based algorithms. The results we have obtained prove that for the same computation, depending on the size of data, the network status and the device status, the decision of the engine may differ.

User interface (UI) development methods are poorly integrated with standard software engineering (SE) practices. Despite current efforts of closing the conceptual gap between these two disciplines, there is still a lack of methodologies supporting a collaborative and synchronized development approach. To address this shortcoming, we propose an integrated development methodology for use cases and task models. Use cases have become the standard to model functional requirements, whereas task models are used to specify UI requirements. Based on this understanding we propose using CTT task models to incrementally enrich the UI-related steps in the use case model, thus achieving a clear separation of concerns and avoiding potential inconsistencies between the two artifacts.

Complete and precise software requirements description is critical in successful development of software systems. This description specifies both functional requirements that define the different functionalities the system should perform, and non-functional requirements that define how the system should perform these functional requirements. Valuable software should meet both its functional (FRs) and non-functional requirements (NFRs). In this paper, we show the possibility of associating NFRs to behavioral models. We propose a framework where NFRs are defined as a set of non functional attribute goals which derive the composition of behavioral models towards the construction of a behavioral model of the overall system.