The Internet of Things (IoT) has emerged as a rapid technological domain in recent years, with predictions indicating its utilization to surpass one billion interconnected devices on a global scale by the year 2030. In the pursuit of Exploiting the full potential of this pervasive and interconnected network of devices, the cultivation of trust and the establishment of reputational frameworks among IoT entities have become imperative. Although numerous trust management models have been proposed within the IoT paradigm, these approaches have often considered frameworks that utilize direct trust along with indirect trust which refers to the concept of trust that is established through a network of intermediaries rather than through direct interaction or verification between IoT devices to evaluate the trustworthiness of a node in IoT network, this latter usually presents many challenges such as consuming much energy and involves more computational power, this paper aims to study, improve and adapt a distributed trust model already proposed for IoT in the context of filtering packets by malicious nodes. This adaptation involves using behavioral trust along with direct trust to achieve network security and prevent malicious actors from participating in the network operations.

The Internet of Things (IoT) has become increasingly susceptible to cyber attacks, making it crucial to have strong detection mechanisms. This paper looks at using adversarial machine learning to boost IoT security. We implemented and evaluated Feedforward Neural Networks (FNN) and Long Short-Term Memory (LSTM) models on the Bot-IoT dataset for binary and multi-class classification tasks. We then evaluated how these models perform under Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD) adversarial attacks. Our results show high accuracy in detecting attacks under normal conditions. However, the models showed major weaknesses when faced with adversarial examples. This study highlights the urgent need for building adversarially robust machine learning models for IoT security. It also gives insights into how different model architectures perform against various attack intensities.

This paper designs a federated learning scheme that combines collaborative game theory to improve the decentralized management of multiple dams. Dams working together can use the FedAS algorithm integrated with cooperative coalitions to solve the issues of water levels while also solving data heterogeneity and resource inconsistencies within the coalition. This helps in resolving issues of predictability and equity of water distribution through optimal resource-sharing goals and adapting to local data patterns. Test results demonstrate better results compared with non-cooperative scenarios, such as a more stable water level and a better balance of water distributed with greater frequency of cooperation which shows great potential for resource management.

Urban congestion often delays emergency services, thus endangering lives in critical situations. In this work, we propose an innovative system that combines Convolutional Neural Networks (CNNs) with Genetic Algorithms (GAs) to improve the prioritization of emergency vehicles in smart urban environments. Based on real-time traffic data from the Internet of Things (IoT), the system detects emergency vehicles (EVs) using visual cues like flashing beacons and vehicle shapes. The model was trained on varied global datasets and tested specifically in Moroccan urban environments, with an accuracy rate of 82.6% in emergency vehicle detection in varied environments. Meanwhile, the GA-based optimization also improves traffic signal control, reducing average delays at intersections by 37% in simulated tests while maintaining balanced prioritization between emergency and normal traffic. Designed for adaptability, the framework operates effectively in cities with heterogeneous infrastructures by dynamically adjusting to local traffic patterns. The results demonstrate that our approach significantly improves emergency response times and enhances traffic efficiency in complex urban settings. By applying this model in real-world contexts in Morocco, where infrastructure variability poses unique challenges, we can contribute to ongoing efforts toward harmonizing technological progress with public safety, developing a scalable model for the build-out of smarter and more resilient cities.

International audience

In this paper, we presents a novel hierarchical federated learning architecture specifically designed for smart agricultural production systems and crop yield prediction. Our approach introduces a seasonal subscription mechanism where farms join crop-specific clusters at the beginning of each agricultural season. The proposed three-layer architecture consists of individual smart farms at the client level, crop-specific aggregators at the middle layer, and a global model aggregator at the top level. Within each crop cluster, clients collaboratively train specialized models tailored to specific crop types, which are then aggregated to produce a higher-level global model that integrates knowledge across multiple crops. This hierarchical design enables both local specialization for individual crop types and global generalization across diverse agricultural contexts while preserving data privacy and reducing communication overhead. Experiments demonstrate the effectiveness of the proposed system, showing that local and crop-layer models closely follow actual yield patterns with consistent alignment, significantly outperforming standard machine learning models. The results validate the advantages of hierarchical federated learning in the agricultural context, particularly for scenarios involving heterogeneous farming environments and privacy-sensitive agricultural data.

The main objective of this contribution is to provide an in-depth analysis of the vulnerabilities present in the various layers of the SDN environment as well as to propose a novel solution to detect distributed denial-of-service (DDoS) attacks, as they pose serious threat to the stability and availability of softwaredefined networks. In addition, a federated learning framework is introduced to identify DDoS attacks in SDN environments, which simultaneously protects privacy while maintaining high detection accuracy. Our proposed solution reduces the risk of data breaches and protects the confidentiality of sensitive data by training models locally. We have used FL to train three classifiers: Deep neural networks (DNN), convolutional neural networks (CNN) and (LSTM) Long Short-term memory to classify two categories of DDoS attacks, namely: UDP Flood, TCP SYN. Achieving 99.99% accuracy and a 99.99% F1-score on TCP SYN floods, alongside 99.94% and 99.97% on UDP floods, our federated CNN not only exceeds the most robust centralized benchmarks but also outperforms our own federated DNN and LSTM models, establishing a new benchmark for SDN DDoS detection while ensuring complete privacy of raw traffic within each domain.

Modern agriculture faces unprecedented challenges requiring innovative solutions for enhanced productivity and sustainability. Digital Twins (DTs), coupled with Artificial Intelligence (AI) and the Internet of Things (IoT), offer a transformative paradigm for creating dynamic virtual replicas of agricultural systems. This paper introduces an advanced interactive web-based Digital Twin framework designed to bridge the gap between virtual sensor data and intelligent, actionable insights for smart farming. The framework integrates a high-fidelity simulation core, encompassing sophisticated mathematical models for environmental dynamics and plant biophysiology, with a robust AI engine capable of complex analysis, prediction, and decision support. Simulated data from a virtual sensor network feeds the DT, enabling the AI to perform real-time assessments, identify anomalies, predict future states, and recommend optimized management actions. The entire system is exposed through an intuitive and visually rich web interface that facilitates interactive exploration, scenario analysis, and direct visualization of the DT's evolution and the AI's reasoning. We demonstrate the framework's architecture, the formal underpinnings of its key components, and its utility as a powerful tool for research, development, and operational management in next-generation agriculture, showcasing a seamless pathway from data acquisition to intelligent intervention.

The rapid growth of large language models (LLMs) has created a demand for reliable, interpretable, and scalable ranking systems capable of comparing models based on human preferences. Traditional benchmarks often fail to capture qualitative nuances in open-domain dialog, where subjective judgments play a central role. In this work, we conduct a comparative study of six prominent ranking algorithms Elo, Glicko, TrueSkill, Bradley-Terry, Markov Chain-based ranking, and a novel HawasRank algorithm applied to the Chatbot Arena dataset containing 244,978 pairwise human preference comparisons. HawasRank is a divergence-based, game-theoretic ranking method inspired by Bregman optimization frameworks, designed to improve convergence speed, transitivity preservation, and computational efficiency in human-in-the-loop LLM evaluations. In this study, we carefully compare these algorithms based on several important criteria, such as predictive accuracy, the occurrence of transitivity violations, sensitivity to hyperparameters, convergence behavior, and CPU resource consumption. The findings of our analysis reveal the trade-offs that exist between different ranking approaches and offer practical guidance for choosing appropriate algorithms, especially in large-scale and evolving LLM evaluation environments.

Cyberbullying is an increasing threat on social media, with serious consequences for mental health, particularly among young people. Despite growing global efforts to address this problem, developing accurate detection systems remains challenging for low-resource languages in both text and audio modalities, such as Arabic and its dialects. This paper presents a binary-labeled dataset for cyberbullying detection in Moroccan Darija. The dataset merges over 4,000 newly collected YouTube comments with the OMCD (Offensive Moroccan Comments Dataset) corpus of 8,024 comments, originally labeled for offensive language. To better reflect the nature of cyberbullying, the entire corpus was reannotated from scratch, clearly distinguishing between bullying and non-bullying content, including subtle forms like sarcasm, shaming, and indirect aggression. Several Arabic transformer models were fine-tuned and evaluated using standard classification metrics. The results show that dialectspecific models, particularly DarijaBERT, achieve the best performance, underlining the importance of context-aware annotation and in-domain pre-training for cyberbullying detection in lowresource settings.

The rapid adoption of the Internet of Things (IoT) has significantly increased security risks, exposing networks to advanced cyber threats. In this paper, a structured analysis of the IoT architecture has been provided to identify vulnerabilities and attack vectors for each layer. To address these challenges, we developed a deep learning-based intrusion detection system, evaluated on the ToNIoT dataset. The experimental results confirm its effectiveness in detecting intrusions with high precision, demonstrating its potential to improve IoT security. This work highlights the potential of deep learning to improve IoT network security and resilience, providing a robust framework for future research and practical applications.

The emergence of Software-defined Networks (SDN) has played a significant role in shaping the future of networking technologies. SDN aims to improve the flexibility, efficiency, and scalability of traditional networks by centralizing network control, allowing network administrators to manage and control the network through software applications. However, the flexibility provided by SDN architecture unveils numerous emerging network security concerns that require more attention to enhance SDN network security. So, in this paper, we propose a innovative intrusion detection system (IDS) for SDN using a hybrid model that combines CNN-BiLSTM and Transformer Encoder. The proposed approach was tested using the NSL-KDD and CICIDS2017 datasets, achieving accuracy rates of 99.7% and 99.68% respectively. The obtained results demonstrate that our proposed deep learning-based approach provides a robust security solution for detecting intrusions in SDN environments.

The integration of solar energy into smart grids introduces challenges for accurate power prediction due to the variability of solar resources and the distributed nature of generation systems. Federated learning offers a promising solution for collaborative model training without centralized data collection. However, its application in solar power prediction faces unique challenges, including data heterogeneity across different geographical regions and communication constraints in edge devices. In this work, we propose a federated learning framework that incorporates data normalization techniques and adaptive model aggregation strategies to address these challenges. We evaluate our approach using real-world solar power datasets and demonstrate its effectiveness in improving prediction accuracy while ensuring data privacy and reducing communication overhead.

In the era of the Internet of Things (IoT), decentralized paradigms for machine learning are gaining prominence. In this paper, we introduce a federated learning model that capitalizes on the Euclidean distance between device model weights to assess their similarity and disparity. This is foundational for our system, directing the formation of coalitions among devices based on the closeness of their model weights. Furthermore, the concept of a barycenter, representing the average of model weights, helps in the aggregation of updates from multiple devices. We evaluate our approach using homogeneous and heterogeneous data distribution, comparing it against traditional federated learning averaging algorithm. Numerical results demonstrate its potential in offering structured, outperformed and communication-efficient model for IoT-based machine learning.

The Internet of Things (IoT) has transformed various industries by enabling seamless connectivity and data exchange between physical devices. However, optimizing the energy consumption of IoT platforms remains a significant challenge due to their complex and heterogeneous nature. In this paper, we propose a modular digital twin framework tailored for IoT platforms, aimed at evaluating data collection policies to optimize energy consumption and prolong device operational life. Our framework integrates hardware abstraction, simulation/emulation, update recommendations, and real-world synchronization to create a functional replica of IoT devices in a digital environment. We categorize IoT device modules into sensing units, communication units, and processing units, each with distinct operational profiles. To demonstrate the efficacy of our framework, we present a greenhouse monitoring IoT application, where temperature and humidity data are collected and transmitted using Zigbee protocol. Through a theoretical exposition, we illustrate the utility of our framework. Our approach lays the foundation for further research in optimizing IoT platforms for energy efficiency, fostering longer deployments, and promoting sustainability in IoT applications.

Recently, intelligent traffic management systems (ITMS) have gained attention from the scientific community due to advancements in Artificial Intelligence (AI), Internet of Things (IoT), and Big Data technologies, as well as the significant impact of road networks on individuals and industries. Although previous studies proposed solutions to traffic congestion, road safety was only a side effect and predominantly focused on highways and high-capacity roads. Traditional safety systems in cities rely heavily on human intervention. This paper presents a versatile and comprehensive system that utilizes IoT to enable real-time traffic control and improve road safety. The system incorporates Big Data technology algorithms to collect, process, and store traffic data in real-time. The prototype uses Kafka, a cutting-edge Big Data technology, and Spark Streaming backed up by Machine Learning algorithms for real-time data processing. The objective is to detect and report traffic violations, specifically pedestrians passageway violations in cities, to ensure surveillance and provide real-time updates and classification on traffic violations.

In the face of growing urban populations and the escalating number of vehicles on the roads, managing transportation efficiently and ensuring safety have become critical challenges. To tackle these issues, the development of intelligent control systems for vehicles is paramount. This paper presents a comprehensive study on vehicle control for collision avoidance, leveraging the power of Federated Deep Reinforcement Learning (FDRL) techniques. Our main goal is to minimize travel delays and enhance the average speed of vehicles while prioritizing safety and preserving data privacy. To accomplish this, we conducted a comparative analysis between the local model, Deep Deterministic Policy Gradient (DDPG), and the global model, Federated Deep Deterministic Policy Gradient (FDDPG), to determine their effectiveness in optimizing vehicle control for collision avoidance. The results obtained indicate that the FDDPG algorithm outperforms DDPG in terms of effectively controlling vehicles and preventing collisions. Significantly, the FDDPG-based algorithm demonstrates substantial reductions in travel delays and notable improvements in average speed compared to the DDPG algorithm.

Water consumption remains a major concern among the world's future challenges. For applications like load monitoring and demand response, deep learning models are trained using enormous volumes of consumption data in smart cities. On the one hand, the information used is private. For instance, the precise information gathered by a smart meter that is a part of the system's IoT architecture at a consumer's residence may give details about the appliances and, consequently, the consumer's behavior at home. On the other hand, enormous data volumes with sufficient variation are needed for the deep learning models to be trained properly. This paper introduces a novel model for water consumption prediction in smart cities while preserving privacy regarding monthly consumption. The proposed approach leverages federated learning (FL) as a machine learning paradigm designed to train a machine learning model in a distributed manner while avoiding sharing the users data with a central training facility. In addition, this approach is promising to reduce the overhead utilization through decreasing the frequency of data transmission between the users and the central entity. Extensive simulation illustrate that the proposed approach shows an enhancement in predicting water consumption for different households.

The Internet of Things (IoT) is changing today's world, and Machine Learning (ML) is a major contributor to this revolution in terms of data exchange to mature connected objects. In this context, federated learning (FL) is emerging, a new ML paradigm that drives a model on decentralized data, which can be distributed across many IoT devices. FL has grown considerably in recent years, both in academia and industry. However, the majority of FL algorithms assume that all client nodes are honest and willing to participate in cooperative model learning. Thus, each node is able to provide reliable local models to the central server. However, in real-life scenarios, nodes may be corrupt, malicious, or both, and may not cooperate fairly during training phases. In this paper, we address the above challenge by proposing a new FL algorithm called DistFL. The main objective of DistFL is to prevent biased training by identifying malicious nodes during the training phase. We evaluate the effectiveness of our technique and demonstrate it through a concrete implementation, comparing DistFL with conventional FL. Even with up to 50% malicious nodes, the runtime cost of the DistFL model is still better than that of the conventional FL model, and its final accuracy reaches 97% with a loss function convergence rate twice that of the conventional FL model. For this study, we used urban water data to deal with leakage and distribution faults in the water network among different end users in this area.

In 2020 only, 36 billion records have been leaked, 95% of those attacks have been caused by human error. Therefore organizations have been looking for multiple technologies to secure there system. One of the modern techniques is using machine learning for traffic classification which help us detect attacks based on monitoring data flow in our network or our workstation. In this paper we presents an implementation of new proposed model based on convolutional neural network (CNN) and long short term memory (LSTM). It is evident from the investigations that different machine learning methods can be used for intrusion detection. Further, the results demonstrated that the usage of machine learning techniques produce positive impact on improving the overall performance of the intrusion detection system in terms of accuracy and lowering false negatives. We using two of most known Deep Learning algorithms CNN and GRU (Convolutional Neural Network, Gated Recurrent Unit) to have a base ground regarding the performance of our proposed model. we are getting encouraging results with implementation of two of famous data-set (NSL-KDD and CIC-IDS2018) We tested our models on binary and multi-class classification for further observation.

The challenge of drinking water optimization refers to the need to effectively manage and distribute drinking water in a manner that is sustainable, efficient, and equitable. To address this challenge, governments, communities, and other stakeholders need to work together to develop and implement effective water management strategies, including the use of technology like IoT (Internet of Things) and innovative approaches to improve water efficiency, conserve water resources, and ensure access to safe drinking water for all. In this paper, we propose a new water distribution network architecture conception that can be implemented in new buildings in the context of smart homes. This architecture will allow us to better optimize the management of drinking water at the level of new Moroccan households domain in smart cities while reusing used water at the household level in toilet flushing, using an intelligent system with connected tanks. To preserve water resources and minimize the cost of water distribution.

Global population growth, climate change, and plant diseases are causing food shortages around the world, making it challenging to produce sufficient food to meet demand. To address this issue, a transition from traditional agriculture to knowledge-based agriculture using modern technologies such as precision farming is imperative.The approach taken in this article aims to protect lettuce plants against diseases and detect any infections. The use of modern technologies, notably the Internet of Things (IoT), has provided solutions to significant problems related to lettuce diseases in high-temperature and low-humidity conditions. Temperature and humidity sensors can be installed in plants to bring them back to standard conditions in case of failure. Additionally, affected plants can be automatically identified using a camera that recognizes diseases and reacts promptly to prevent their spread.The combination of IoT and machine learning for big data analysis enabled us to achieve our objectives. The results of this study indicate that the Convolutional Neural Network (CNN) model performs the best for this task, with an accuracy of 94%. This high accuracy is attributed to its ability to extract relevant features from images, a crucial aspect in image classification tasks. However, the results for K-Nearest Neighbors (KNN) and Multi-Layer Perceptron (MLP) models are also promising, with respective accuracies of 93%. These models, while slightly less accurate than concolutional neural network (CNN), offer potential utility in other types of classification tasks based on data characteristics.

With the rapid increase in network intrusions, applying an active network intrusion defense is more important than ever before. Different learning algorithms have been combined to achieve better performance. To improve the accuracy and efficiency of the network intrusion detection system, a new hybrid algorithm is designed, which combines the Random forest, Decision tree and Multilayer Perceptron (MLP) algorithms. The experimental results show that the hyprid model has a better true Positive Rate (TPR) for attack activities, rapidly data preprocessing speed, and shorter training time. In particular, the accuracy of multi-class classification can reach as high as 99.7% in the NSL-KDD dataset, 77.99% in the UNSW-NB15 dataset and 84.89% in the CIC-IDS-2017.

In this paper, we investigate Federated Learning (FL), a paradigm of machine learning that allows for decentralized model training on devices without sharing raw data, thereby preserving data privacy. In particular, we compare two strategies within this paradigm: Federated Averaging (FedAvg) and Personalized Federated Averaging (Per-FedAvg), focusing on their performance with Non-Identically and Independently Distributed (Non-IID) data. Our analysis shows that the level of data heterogeneity, modeled using a Dirichlet distribution, significantly affects the performance of both strategies, with Per-FedAvg showing superior robustness in conditions of high heterogeneity. Our results provide insights into the development of more effective and efficient machine learning strategies in a decentralized setting.

The Industrial Internet of Things (IIoT) network is composed of devices that contain sensitive data, which makes them vulnerable to various security threats. Digital Certificates can be used to reinforce the security of the IIoT network, however, their management remains a major issue. Hence, in this article, we rely on trust management to deal with the whole certificate management process in IIoT networks, from revocation to verification. For this purpose, we organize the IIoT network into a clustering architecture where each cluster head (CH) hosts an agent, called CH-UR agent, that renews/revokes the certificates of its cluster member nodes. We apply signaling game theory to build a Certificate Revocation Game modeling the interactions between a member IIoT node and the CH-UR agent. Thus, upon the belief on the member node, updated by using the Bayesian rules, the best response strategy for the CH-UR agent can be obtained. Further, we propose a new efficient certificate verification scheme based on short-lived certificates (SLCs) and suitable for IIoT network requirements. The performance evaluation of our framework proves, first, the accuracy and convergence speed of our revocation mechanism to detect untrusted devices and on-off attacks. Second, the effectiveness of our clustering architecture to reduce the resource consumption resulting from the management of SLCs to 60% even with the increase of network density. Third, the effectiveness of the proposed certificate verification scheme to reduce the time needed to obtain the revocation information as well as the resulting storage and communication overhead to achieve this purpose.

Variations in breath rates can be used as an indicator for important changes in the patient's physiological status. Although the breath rate is an important vital sign, it is often overlooked due to the difficulty of monitoring it. With the rise of the Internet of Things (IoT), connected low-cost sensors can be used as an alternative to traditional breath rate monitoring methods. This paper presents a light-weight, yet efficient, breath rate estimation technique based on continuously monitoring the variations in nasal air temperature using a low-cost temperature sensor. Sophisticated sampling mechanisms are proposed to reduce energy consumption which is an important feature for low-resources IoT devices. Results show that the proposed technique is able to reliably and accurately monitor the breath rate while reducing the processed data volume by up to 94%.

Nowadays, people's demand for water is growing as well as in public, commercial, and industrial sectors. However, the limited character for water resources presents a crucial obstacle to satisfying needs for continued human development. The control and provision of potable water are therefore the most vital challenges for the water supply system. It must use water resources in an efficient manner, and fulfill both quality and quantity demands. For this purpose, the current water supply system uses smart infrastructure that gathers, processes, stores, and delivers water from water sources to users. It is done in a very complex environment with ever-increasing demand, and often conflicting services to deliver. In this work, we developed a Federated Learning (FL) model, which is come in the form of a machine learning setting where clients collaboratively train under the orchestration of a central server while keeping the training data decentralized. The FL embodies the principles of focused data collection and minimization, it can mitigate many of the systemic privacy risks and costs resulting from traditional, centralized machine learning and data science approaches. Implementing on the central server, the FL will iterate on the gateway's learning models using the federated averaging algorithm. Our approach's capacity is shown by addressing reel instances in a real-world issue in which the proposed method finds much better results and a significant improvement in performance compared to the standard approaches while achieving our goal and suggesting a new interesting direction for research.

Federated Learning is an artificial intelligence (AI) method that brings a new approach to ensure a high level of confiden-tiality during model creation. In federated Learning, instead of sending the local private data by the nodes to a central server, the server will initiate the process by sending the initial global model to the participants nodes that will perform the training phase based on their data locally and send back the results (Weights of the model) to the server for the aggregation, finally the server updates the global model based on this aggregation to make the model more efficient and send it to the nodes. Sometimes, it can be some selfish nodes injected into the partic-ipants, these selfish nodes send faulty results to the server, if the number of these nodes is minimum, the impact can be negligible, but if there are many of these selfish nodes, it deceives and pushes the whole global model to make false results. Hence, that will impact the other honest node's performances, since the server sends the model updates to them. In this research, we will focus on this problem and we will demonstrate how these selfish nodes can impact the performance of the global model completely, also we will propose a system model that can detect and eliminate them from the participants list.

In this paper, we address the resource provisioning problem for service function chaining (SFC) in terms of the placement and chaining of virtual network functions (VNFs) within a multi-access edge computing (MEC) infrastructure to reduce service delay. We consider the VNFs as the main entities of the system and propose a mean-field game (MFG) framework to model their behavior for their placement and chaining. Then, to achieve the optimal resource provisioning policy without considering the system control parameters, we reduce the proposed MFG to a Markov decision process (MDP). In this way, we leverage reinforcement learning with an actor-critic approach for MEC nodes to learn complex placement and chaining policies. Simulation results show that our proposed approach outperforms benchmark state-of-the-art approaches.

Sampling is a core process in IoT systems. It deter-mines the data volume circulating within the network as well as the energy consumption on the IoT devices. Adaptive sampling aims to control the volume of generated data to reduce energy and bandwidth consumption without undermining data quality. Within this context, we propose two new adaptive sampling techniques: a light-weight adaptive sampling algorithm and an optimized uniform sampling method. We tested our methods using various real data-sets and compared their performances against state-of-the-art adaptive sampling algorithms in terms of data quality and data volume. The results show that the proposed methods are consistently among the best with a noticeable reduction in computational load.

With the increasing use of wireless communication technologies, it is important to monitor electromagnetic exposure, ideally with high temporal and spatial resolutions. This paper presents our low-cost electro-smog measurement process, covering hardware selection, RF power measurement, and RF power correction. Then, a time series analysis is performed on the electromagnetic exposure data collected in the city of Sala Al Jadida - Morocco for seven days. The results show that the electro-smog exposure has a strong predictable pattern and a preliminary time series model is derived.

Air pollution continues to draw global attention, and still causes adverse environmental issues and health effects. Questions about how air pollution evolves spatially are also still unsolved. Many air quality monitoring stations are deployed in several countries to give insight about air quality. However, it is quite frequent for these stations to go out of order during their long life time. These incidents may lead to a significant loss of pollution data. To mitigate this issue, we propose to leverage spatial correlation between pollution monitoring stations to predict the lost data. To reduce the complexity of the prediction model, for each monitoring station, we identify the best set of other stations to use in the prediction model. Correlation-based station selection is shown to outperform distance-based station selection and provides an R2above 0.8 when applied to the Airparif datasets.

Multi-access edge computing (MEC) is a key enabler to reduce the latency of vehicular network. Due to the vehicles mobility, their requested services (e.g., infotainment services) should frequently be migrated across different MEC servers to guarantee their stringent quality of service requirements. In this paper, we study the problem of service migration in a MEC-enabled vehicular network in order to minimize the total service latency and migration cost. This problem is formulated as a nonlinear integer program and is linearized to help obtaining the optimal solution using off-the-shelf solvers. Then, to obtain an efficient solution, it is modeled as a multi-agent Markov decision process and solved by leveraging deep Q learning (DQL) algorithm. The proposed DQL scheme performs a proactive services migration while ensuring their continuity under high mobility constraints. Finally, simulations results show that the proposed DQL scheme achieves close-to-optimal performance.

Low-latency IoT applications, such as autonomous vehicles, augmented/virtual reality devices, and security applications, require high computation resources to make decisions on the fly. However, these kinds of applications cannot tolerate offloading their tasks to be processed on a cloud infrastructure due to the experienced latency. Therefore, edge computing (EC) is introduced to enable low latency by moving the tasks processing closer to the users at the edge of the network. The edge of the network is characterized by the heterogeneity of edge devices (EDs) forming it; thus, it is crucial to devise novel solutions that take into account the different physical resources of each ED. In this article, we propose a resource representation scheme, allowing each ED to expose its resource information to the supervisor of the edge node through the mobile EC application programming interfaces proposed by the European Telecommunications Standards Institute. The information about the ED resource is exposed to the supervisor of the edge node each time a resource allocation is required. To this end, we leverage a Lyapunov optimization framework to dynamically allocate resources at the EDs. To test our proposed model, we performed intensive theoretical and experimental simulations on a testbed to validate the proposed scheme and its impact on different system's parameters. The simulations have shown that our proposed approach outperforms other benchmark approaches and provides low latency and optimal resource consumption.

Industrial IoT network (IIoT) is the result of integrating Internet of Things into the industrial processes to make manufacturers more efficient and reactive. This integration brings with it new security and privacy challenges because of the sensitive information exchanged between devices and the infrastructure. In order to prevent malicious nodes from abusing the IIoT system, this paper investigates a new distributed certificate revocation protocol. Hence, we propose cluster-based certificate revocation mechanism for IIoT network by using game theory. Indeed, our proposed signaling game allows to make a decision on a targeted node after evaluating its behaving. In this approach, we assume an IIoT network organized as a set of clusters called communities. Each community consists of a set of member nodes and a community leader who uses a multi-stage game to renew the certificate of well behaving member nodes and revoke the certificate of malicious member nodes. Based on the Perfect Bayesian Equilibrium, we describe a reactive algorithm implementing the certificates update mechanism. By simulations, we evaluate the convergence speed, the dynamic and the accuracy of the leader's posterior belief on nodes with different behaviors.

Network slicing (NS) will have an essential role to enhance the isolation and the flexibility of the future generation of cellular networks (5G) with heterogeneous capabilities, notably in ultra-dense (UD) urban zones. Thus, it may be the principal component necessary to respond to 5G UD heterogeneous networks (UD-HetNets) technical requirements. The aim of this paper is to propose a new NS architecture to resolve user equipment (UE)-association problem in 5G UD-HetNets. Accordingly, we formulated the problem as a one-to-many matching game based on matching theory, while exploiting the isolation character of slicing to eliminate interferences between pico-cells and those among macro-cell and pico-cells. Next, we proposed the UE-slice association algorithm (U-S.AA) to find the stable matching among user equipments (UEs) and different network slices. Numerical simulation results validate our theoretical model, and prove the efficient of the proposed user-slice association solution to enhance the global network performance, respecting the UEs quality of service (QoS), as well as improving the energy efficiency (EE) of UEs.

Unmanned aerial vehicles (UAVs) are a promising technology employed as moving aggregators and wireless power transmitters for IoT networks. In this paper, we consider an UAV-IoT wireless energy and data transmission system and the decision-making problem is investigated. We aim at optimizing the nodes' utilities by defining a good packet delivery and energy transfer policy according to the system state. We formulate the problem as a Markov Decision Process (MDP) to tackle the successive decision issues. As the MDP formalism achieves its limits when the neighbors' interactions are considered, we formulate the problem as a Graph-based MDP (GMDP). We then propose a Mean-Field Approximation (MFA) algorithm to find a solution. The simulation results show that our framework achieves a good analysis of the system behavior.

Vehicular Ad-hoc Networks (VANETs) have many promising applications such as improving vehicle driver's safety, and so, decreasing car deaths. In VANETs, the main communication way is broadcast. Such Broadcast needs to be done efficiently especially in congested areas to avoid the broadcast storm problem. A crashed vehicle disseminates a message about its incident to all nodes in the network. The vehicle nodes of the network have to disseminate in their turn the alert message carefully to maximize both reachability and saved rebroadcast (by reducing congestion) while minimizing the delay. This concern is demonstrated and modeled by decision hypothesis. In this paper, at first, we are keen on the Markov decision process (MDP), which is utilized to model and tackle such successive decision issues. We aim to optimize a utility for each vehicle relying upon an irregular domain and choices made by it. As the MDP formalism achieves its cutoff points when it is important to consider the cooperation between the several vehicles. We will begin utilizing the Graph-based MDP where the state and activity spaces are factorizable by factors. We then demonstrate that the transition functions and rewards are deteriorated into nearby functions, and the reliance relations between the vehicles are spoken to by a graph. To Figure the optimal strategy, we use Mean Field Approximation (MFA) for tackling GMDP issue.

We propose a deep learning based technique for power quality disturbances (PQD) detection and identification that aims at mimicking the reasoning of human field experts. This technique consists of processing small-size images containing superimposed time and frequency representations of the electric signal. The classification of PQD is performed with a convolutional neural network (CNN) trained with synthetic signals containing various single and multiple PQDs. Simulation results show that our technique is able to detect and identify with a high accuracy, in addition to pure sinusoidal, eight single PQDs and 20 of their combinations (up to four PQDs in the same signal) even in the presence of noise. Features such as lower computational load and simplicity while maintaining high performance sets the proposed technique apart from previous ones.

This paper presents an IoT platform designed for air pollution monitoring. It aiming to facilitate the testing of different data collection strategies, to simplify the air quality monitoring process, to provide the citizens with real-time information about air pollution, to allow citizens to participate in the air quality monitoring process, and to help the authorities identify zones of high air pollution and take the most appropriate measures to improve air quality. The sensor nodes have been developed using low-cost off-the-shelf hardware. Both nomadic and mobile sensor nodes have been developed. A novel sensor management middleware has been designed and developed to have the flexibility to remotely control the operational settings of the sensor nodes and to reduce the volume of transmitted data. Finally, two applications have been developed and implemented for data visualization. The first is a mobile friendly air pollution meter. The second offers a spatial visualization of air pollution levels using a Geographic Information System (GIS). The developed platform has been tested in multiple measurement campaigns. The results of one of these campaigns (conducted in Hay Nahda II, Rabat, Morocco) is presented in this paper to showcase the platform.

Automobile manufacturers around the world are increasingly deploying Industrial Internet of Things (IIoT) devices in their factories to accompany the Industrial Revolution 4.0. Security and privacy are the main limitations to the integration of Internet of Things (IoT) into industrial processes. Therefore, it is necessary to protect industrial data contained in IIoT devices and keep them confidential. As a step towards this direction, in this paper, we propose a dynamic trust management model suitable for industrial environments. We propose also to change the traditional centralized architecture of IIoT networks in automotive plants into a hybrid architecture based on a set of new industrial relationship rules. The performance evaluation in this work is done in two parts. In the first part, we compare our proposed architecture with the traditional architecture of the plant's IIoT network. The results of this comparison show that our architecture is more suitable to simplify trust management of IIoT devices. In the second part, we demonstrated the ability, the adaptiveness and the resiliency of our proposed trust model against behavioral changes of IIoT nodes in malicious environments.

Fog computing is emerging as a new paradigm to deal with latency-sensitive applications, by making data processing and analysis close to their source. Due to the heterogeneity of devices in the fog, it is important to devise novel solutions which take into account the diverse physical resources available in each device to efficiently and dynamically distribute the processing. In this paper, we propose a resource representation scheme which allows exposing the resources of each device through Mobile Edge Computing Application Programming Interfaces (MEC APIs) in order to optimize resource allocation by the supervising entity in the fog. Then, we formulate the resource allocation problem as a Lyapunov optimization and we discuss the impact of our proposed approach on latency. Simulation results show that our proposed approach can minimize latency and improve the performance of the system.

Unmanned aerial vehicles (UAVs) are a promising technology to provide an energy-efficient and cost-effective solution for data collection from the ground Internet of Things (IoT) devices. In this paper, the optimal associations that provide reliable connections between UAVs and IoT devices are investigated. We aim at maximizing the IoT devices' benefits, by assigning them to most suitable UAVs. We formulate the problem as a many-to-one matching game where UAVs and IoT devices are the players. In this game, the players rank one another based on individual utility functions that capture their needs. Each IoT device aims to minimize its transmitting energy while meeting its SNR requirements and each UAV aims to maximize the number of served IoT devices while respecting its energy constraints. Simulation results show that the proposed approach provides a low average total transmit power, ensures fast data transmission and optimal utilization of the UAVs' bandwidth.

Distributed architectures of the SDN control plane require a careful design for balancing the load among controllers. Solutions proposed for SDN load balancing usually use switch migration operations. However, an efficient switch migration means triggering the operation at the right moment, and judiciously choosing the migrated switch and the destination controller. Here, we propose a switch migration scheduling algorithm to improve the migration efficiency, and ensure load balancing between controllers. Our algorithm uses a multi-step ARIMA forecasting model to predict the long-term controllers load. When an overload is predicted, a switch migration operation is scheduled in advance. After validating the accuracy of the ARIMA forecasting model, we evaluated the performance of the algorithm by analyzing the response time of controllers. Numerical results confirm the performance of the proposed algorithm.

IoT sensor networks are applied in several areas, including research, security, and monitoring. All these applications are based on data collection. A practical solution is a use of unmanned aerial vehicles (UAVs) as aerial relays to ensure connectivity between IoT-devices and the destination node. Most applications have been based on using a single UAV to transmit information, while our work focuses on forming a network of multiple UAVs. In this paper, a novel deployment of UAVs moving along non-concentric rings is analyzed. First, they collect packets generated by IoT-devices and then transmit them throughout a UAV-to-UAV forwarding schema to the destination. We derive a closed formula for the end-to-end throughput of the proposed gathering network architecture.

The unlicensed band was found to be useful in terms of users applications QoS Regarding the downlink throughput. The unlicensed band, once used exclusively by the military and medical industry, will soon be exploited for the public interests. The large bandwidth guaranteed by the 5GHZ frequency band will help reduce the data transmission latency. It is for these reasons that the LTE-U (Long Term Evolution on Unlicensed band) was developed and will play a key role in the wireless networks of the 2020's. Its deployment, however, will be at the expense of other existent wireless networks technologies, specifically the Wifi. Indeed once the Wifi networks deployed alongside with the LTE-U, their users applications will be constantly interrupted, since the LTE-U is a dominant technology Regarding the channel access. This issue will cause a degradation of a large part of those deployed users in the heterogeneous networks regarding the QoS. Which is why it is important to figure out a solution that allows Wifi-exploiting applications in heterogeneous networks to serve their users with a minimum value of QoS. In this publication, we will detail the important steps towards a coexistence between the Wifi and the LTE-U, using the cooperative game theory, and more specifically the canonical coalition game, for a fairer channels/sub-channels allocation. The paper also solves the time allocation problem between the access points and the small cells, thanks to the bankruptcy game, which is an entitlement problem involving the allocation of a given amount of a perfectly divisible time among the wireless nodes. The publication will also show that the proposed solution provides a higher average user's throughput than the Bargaining Game solution, whether the user is in the Wifi or the LTE-U network.

On-demand services such as traffic management and video streaming are typical Intelligent transport systems (ITS) requiring very low latency and high bandwidth running. Vehicular Ad hoc NETworks (VANETs) represent important opportunities to exploit content that virtual service providers (VSPs) offers. The concept of “Cache as a Service (CaaS)” is a promising technique to minimize the average latency while satisfying the QoE requirements of vehicles. In this paper, we mainly consider the problem of optimizing latency in ITS, exploiting a completely virtualized environment and caching feature. In this perspective, virtual service providers (VSPs) and mobile virtual network operators (MVNOs) are connected in the Cloud through network as a service (NaaS), using distributed infrastructure as a service (IaaS). VSPs provide services according to VANETS movement. The VANET, as service requester will take advantage of emerging caching techniques (CaaS) to accomplish the on-demand low-latency services requiring computing resources and bandwidth. Consequently, we propose a many-to-many matching strategy coupled to CaaS caching capabilities on distributed F-RAN between VANET and virtual service provider (VSP). We exploit the deferred acceptance algorithm to solve this game. To highlight the effectiveness of our approach, we applied it on two typical on-demand services requiring ultra-reliability and low-latency communications (uRLLC): The intelligent transport and the video streaming services. The simulation results demonstrate the effectiveness of our approach in terms of improved latency and bandwidth optimization and especially during periods of traffic congestion.

Summary We consider a mobile delay‐tolerant networks (MDTNs) with energy‐harvesting capabilities. In order to determine energy management policies that will improve network capacity and packet delivery ratio and maximize the system throughput, we consider a source node that seeks to send packets to a destination node. The optimal policy for the source varies according to its system state, which allows it to guarantee a maximum delivery probability rate. Our problem is modeled by decision theory; as a start, we are interested in the MDP to model and solve such sequential decision problems. Our goal is to optimize for each node, a utility depending on a random environment and decisions made by the node. As the MDP formalism reaches its limits when it is necessary to take into account the interactions between different several nodes, we will start using the graph‐based MDP where the state and action spaces are factorizable by variables. The transition functions and rewards are then decomposed into local functions and the dependency relations between the nodes are represented by a graph. To calculate the optimal policy, we propose Mean Field Approximation (MFA) and Approximate linear‐programming (ALP) algorithms for solving GMDP problem.

Fog computing has emerged as a new paradigm in mobile network communications, aiming to equip the edge of the network with the computing and storing capabilities to deal with the huge amount of data and processing needs generated by the users devices and sensors. Optimizing the assignment of users to fogs is, however, still an open issue. In this paper, we formulated the problem of users-fogs association, as a matching game with minimum and maximum quota constraints, and proposed a Multi-Stage Differed Acceptance (MSDA) in order to balance the use of fogs resources and offer a better response time for users. Simulations results show that the performance of the proposed model compared to a baseline matching of users, achieves lowers delays for users.

With the explosive growth of mobile multimedia traffic, the problem of allocating computing and spectral resources for very low latency services become a challenge for next generation mobile networks. We exploit all the benefits of a completely virtualized environment, where mobile virtual network operators (MVNOs) and virtual service providers (VSPs) are connected in the Cloud through network as a service (NaaS) using distributed infrastructure as a service (IaaS). VSPs provide services according to Internet of things (IoT) devices requests including software as a service (SaaS) and emerging caching techniques Cache as a Service (CaaS) to satisfy the quality of service (QoS) requirements. Thus, we propose a many-to-many matching game between the sets of IoT devices and the set of virtual service providers (VSPs). To solve this game, we exploit the deferred acceptance algorithm that enables the players to self-organize into a stable matching and a reasonable number of algorithm iterations. To highlight the effectiveness of our approach for the on-demand services, we applied it on two typical services requiring ultra- reliability and low-latency communications (uRLLC): The intelligent transport and the video on-demand services. Simulation results has demonstrated that our proposed matching strategy coupled to CaaS caching capabilities on distributed F-RAN significantly outperforms the traditional strategies in terms of latency and network traffic load.

5G ultra-dense multi-Radio access technology (Multi-RAT) HetNets are considered recently, by industrials and mobile network operators, as a key solutions for boosting network capacity. Hence, supporting the exponentially increasing demand of data traffic. However, improving multiple conflicting metrics for users, with distinctive quality of service (QoS) and quality of experience (QoE) requirements, remains the main challenge in these network environments. This is due to the inappropriate adaptation of the current user association schemes. To overcome this challenge, we formulated the user association problem, in 5G ultra-dense multi-RAT HetNets, as a multi-objective optimization problem (MOOP), solved by the weighted sum method. Then, we proposed the multi-objective genetic algorithm (MOGA) to reach suitable associations, that respect the individual requirements of each user. Numerical results show the potency of the MOGA with decoupled access (MOGA-DA) compared to the MOGA with coupled access (MOGA-CA) and Max-SINR (Signal-to-Interference-plus-Noise Ratio) association schemes in terms of energy efficiency (EE), cost efficiency (CE) and QoS uplink/downlink throughput.

In order to face the security risks and challenges imposed by the openness nature of the Internet of Things (IoT), this paper proposes a distributed advanced analytical trust model based on a Markov chain. The main objective of this paper is to study, improve and adapt a distributed trust model already proposed for Vehicular Ad hoc Networks (VANETs) to the IoT context. This adaptation uses specific parameters to the IoT networks and considers all limitations of IoT devices as well as the less mobile nature of these devices than vehicles. In this model, a set of neighbor nodes organize themselves into groups to evaluate and monitor another IoT node called monitored node depending on its behavior in the network. Our model uses an estimation algorithm to remove spams provided by malicious neighbors during the process of updating the monitored node trust. The numerical results illustrate the adaptability, the robustness and the strong resistance of our adapted trust model against the dynamic behaviors of the IoT devices and against the various popular attacks related to trust.

SDN technology has arrived to solve several limitations of standard networks such as flexibility, scalability and programmability. In a data center environment, adopting the SDN approach where switches are statically linked to controllers creates load balancing problems. This issue is due to the traffic variation between controllers and switches, which influences the response time of the controllers. In this paper, we propose a dynamic assignment of switches to controllers by formulating the problem as a one-to-many matching game with a minimum quota that each controller has to achieve. This quota represents the utilization of the processing capacity of the controller. In addition, an efficient algorithm is defined to ensure a stable matching between switches and controllers in order to maintain load balancing and reduce the latency of controllers. Numerical results confirm the performance of our proposed model compared to a static assignment of switches in terms of load balancing and minimization of the response time especially, when the network becomes too loaded.

We consider a mobile delay tolerant networks (MDTNs) with energy harvesting and wireless energy transfer capabilities. In order to determine energy management policies that will improve network capacity, packet delivery ratio and maximize the system throughput. We consider that a source node seeks to send packets to a destination node. The optimal policy for the source is varies according to its system state, which will guarantee a maximum delivery probability rate. Each mobile source node transmits wirelessly a portion of its energy as a reward to relay mode. Our problem is modeled by decision theory; as a start, we are interested in the MDP, which are used to model and solve such sequential decision problems. In this paper, for each node, we try to optimize a utility depending on a random environment and decisions made by a node. As the MDP formalism reaches its limits when it is necessary to take into account the interactions between the different several nodes, that's why we chose to use the Multi agents Markov Decision Processes (MMDP) which is a MDP with a large space of states and actions. The set of agents are then considered as a single agents whose goal are to compute an optimal attached policy for MDP. To make a realistic analysis of our model, we assume that the policy of the MMDP is applied in a decentralized way, which makes finding optimal control intractable; thus, we will develop several approximations and evaluate their effectiveness.

Content is not always about medias and files, computing tasks output are also a content that can be described, stored and cached. In this paper, we investigated the problem of edge computing and caching in the fog computing networks. In our scenario, we consider the user requests computing tasks from the fog, that the devices could not handle. Moreover, fogs use their storage and computing capabilities to cache the tasks computation results in order to minimize the latency, and use it storage to offload their resources by caching the significant computing tasks output. We propose a clustering method based on the correlation between the tasks and the cached content to decide what fog will give a better service in term of latency to offer better quality of service and experience. The problem of the users-fogs association was formulated as a coalition game between the users and the fogs. Finally, we use the BoltzmannGibbs learning algorithm to make every entity able to learn the best coalition, in order to enhance the convergence of the system to reach a better and optimal stability.

With the explosive growth of mobile multimedia traffic, content caching is seen as an effective solution to alleviate the heavy traffic burden on back-haul and front-haul and to improve the quality of real-time data services. The concept of ”Cache as a Service (CaaS)” is a framework for caching virtualization for mobile cloud-based networks. Consequently, contents can be distributed and stored based on their popularity, traffic diversity, and diverse user demands. In this paper, we mainly consider the problem of allocating computing resources for very low latency services, as well as high data rate services that require sufficient spectral resources. We plan to exploit all the benefits of a completely virtualized environment, where mobile virtual network operators (MVNOs) and virtual service providers (VSPs) are connected in the Cloud through network as a service (NaaS) using distributed infrastructure as a service (IaaS). VSPs provide service to (Internet of things) IoT devices including software as a service (SaaS). The IoT devices, as a service requester will take advantage of emerging caching techniques (CaaS) to accomplish the on-demand low-latency services that require a large amount of computing resources and a high bandwidth. In order to satisfy the quality of service (QoS) requirements, the radio access network RAN as a Service (RANaaS) is the pivot of this environment that will allocate dynamically networking, computing and storage resources according to the required services in terms of latency and throughput. Thus we propose a many-to-many matching game between the sets of IoT devices and the set of virtual service providers (VSPs). To solve this game, we exploit the deferred acceptance algorithm that enables the players to self-organize into a stable matching and a reasonable number of algorithm iterations. The goal of the proposed manyto-many game theory approach is to optimize the caching spaces that VSP exploit in the edge to store files or software required by IoT devices. Simulation results has demonstrated that our proposed matching strategy coupled to CaaS caching capabilities on distributed F-RAN significantly outperforms the traditional caching strategies in terms of the cache hit ratio, average latency and back-haul traffic load.

The growth of mobile multimedia content consumption opens up a competitive market where the products are digital multimedia provided by the content providers. The content caching capability is deployed as an extension to the small cell network (SCN) in a Cloud radio access network CRAN of the mobile operators. Mobile users and Internet of things IoT devices, as content requesters, will take advantage of emerging caching techniques and fog computing to accomplish the ondemand services that require high throughput and large amount of computing resources. In this perspective, the mobile operators and contents providers find themselves linked in this market profit generating, and consequently in competition for caching contents in shared space storage owned by multiple mobile operators and setting price issues. The economic component of the caching service adds more stringent constraints to the classical learning and prediction algorithms to estimate the content popularity. We formulate this problem as an oligopolistic multimarket Cournot model and use a non-cooperative dynamic game to obtain the optimal amount of caching space. In a dynamic play, contents providers gradually and iteratively adapt their strategies according to their previous strategies and information provided by the consumers. The Nash equilibrium and the stability state of the dynamic play will be proved. Finally, the simulation results are presented to show the performance of the distribution scheme of proposed dynamic caching resources over classical learning and prediction algorithms.

Device-to-Device (D2D) communication is a promising technology to enhance spectrum efficiency and improve system capacity. One of the major problems in spectral reuse is the important interference to the cellular network when they both share the same resources. This paper considers the channel assignment problem for D2D communication underlaying cellular networks. Our goal is to maximize the overall system throughput by applying an approach based on regret-matching learning while the constraints related to the quality of service of the users are respected. We formulate the problem as a non-cooperative game where players (D2D player) choose the channel that maximizes their utility function by learning their best strategy (based on the regret observed by playing an action). The advantage of regret matching is that it is distributed and involves limited information exchanges among players. The algorithm shows relatively fast convergence to the set of correlated equilibrium and near optimal performance after a small number of iterations.

Since it first appears in 2008 as the underlying technology of the Bitcoin payment system, Blockchain has gained further interests, since it offers a distributed peer-to-peer ledger where non-trusting members can interact with each other, in addition, members can now run applications without the need for a central authority with the same level of certainty. however blockchain suffers from certain limitations and challenges, specifically, when combined with Internet of things (iot) infrastructure, namely the high cost needed in the mining process due to restricted computation resources of iot devices, in this paper we will focus on a survey on three of most relevant papers that discuss resources management in mobile blockchain then we we will propose a matching model based on graph theory for optimal resources allocation between services offered by an Edge service provider and demands from miners of a blockchain network.

We consider a mobile delay tolerant networks (MDTNs) with energy harvesting capabilities. In order to determine energy management policies that will improve network capacity, packet delivery ratio and maximize the system throughput. we consider that a source node seeks to send packets to a destination node. The optimal policy for the source varies according to its system state, which allows it to guarantee a maximum delivery probability rate. Our problem is modeled by decision theory; as a start, We are interested in the MDP, which are used to model and solve such sequential decision problems. Our goal is to optimize, for each node, a utility depending on a random environment and decisions made by the node. As the MDP formalism reaches its limits when it is necessary to take into account the interactions between the different several nodes, we will start using the Graph-based MDP where the state and action spaces are factorizable by variables. The transition functions and rewards are then decomposed into local functions and the dependency relations between the nodes are represented by a graph. To calculate the optimal policy, we propose Mean Field Approximation (MFA) and Approximate linear-programming (ALP) algorithms for solving GMDP problem.

In this paper, we propose a new version of the LBRW (Learning based Random Walk), LBRW-Co, for predicting users co-occurrence based on mobility homophily and social links. More precisely, we analyze and mine jointly spatio-temporal and social features with the aim to predict and rank users co-occurrences. Experiments are performed on the Foursquare LBSN with accurate and refined measurements. Experimental results demonstrate that our LBRW-Co model have substantial advantages over baseline approaches in predicting and ranking co-occurrence interactions.

In this work, we focus our study to improving the energy efficiency of mobile cellular users in ultra-dense LTE HetNets. The hyper-dense co-channel deployment of indoor LTE small cell networks (SCNs) within LTE macro cell networks (MCNs) will aggravate the effect of cross-tier interferences caused by the uplink transmissions of macro-indoor users located inside the overlapping zones of small base station (SBS) coverage areas. Hence, degrading the uplink performance at the level of SBSs adopting closed access policy. In order to eliminate the severe cross-tier interferences, each SBS attempts to open the access for macro-indoor users that accept only the SBS with Max-SINR offer. This will lead to network congestion problems in several SCNs. Wherefore, we formulate our problem as a many-to-many matching game. Then, we introduce an algorithm that computes the optimal many-to-many stable matching which consist of assigning each macro-indoor user with multi-homing capabilities to the most suitable set of SBSs and vice versa based on their preference profiles. With regard to the conventional Max-SINR association scheme, our solution can effectively improve the energy efficiency of cellular users. Moreover, it can ensure load balancing in ultra-dense LTE HetNets.

In this paper, we propose a game-theoretic approach to find the community structure in complex networks based on a non-cooperative game. This approach optimizes a node-based modularity for non-overlapping communities. Experiments show that our approach is effective to discover non-overlapping communities and obtain high values of modularity and Normalized Mutual Information (NMI) for real-world and synthetic networks in a reasonable time.

This paper addresses the channel occupation and the channel selection problem for the Long Term Evolution Unlicensed (LTE-U) technology when coexisting with the WiFi system on the unlicensed spectrum (typically the 5 GHz band). For instance, LTE is a communication standard developed by the 3GPP corporation based on the GSM/EDGE and the UMTP/HSPA technologies and conceived for high-spaced wireless communications. If we focus on LTE-U, it is an extension of the LTE-A (LTE-Advanced) scheme in unlicensed bands. In this paper, we will discuss the coexistence matters between the WiFi and LTE-U technologies. Therefore, we review the different solutions discussed in literature that addressed the coexistence issue. These works are either based on fair spectrum allocation between LTE-U and WiFi technologies or on the game theory paradigm where multi channel access and inter-dependance scenarios are discussed. As a future work, we intend to further investigate the coexistence issues between LTE-U and WiFi systems, especially in the case of inter-dependance scenarios and other configurations including hidden/exposed terminal problem in WiFi networks.

Over the past 20 years, data has increased in a large scale in various fields. Internet of Things (IoT), for instance, comprises billions of devices and the data streams coming from these devices challenge the traditional approaches to data management and contribute to the emerging paradigm of big data. To be able to handle such data adequately, it is necessary to reduce their dimensionality to a size more compatible with the resolution methods, even if this reduction can lead to a slight loss of information. The aim of this paper is to study the potential of dimensionality reduction in text categorization of a publicly available dataset CNAE-9.

Many researches and standardization work on the challenges that 5thnetworks generation raised from the radio perspective while employing advanced techniques such as massive MIMO (Multiple-Input-Multiple-Output) and CoMP (Cooperative Multi-points Processes). However the backhaul problems such as bottlenecks has emerged due to the deployment of ultradense and heavy traffic that should be connected to the core networks. In this paper we investigate the caching as a promising solution to deal with the backhaul problems and the offload of the network. By caching the content near the users, at the base stations or at the device side via device-to-device communications or in advanced architecture of the cloud (In the Fog) is a promising solution to bring the interesting content closer to the users. Caching techniques are many, in this paper we grouped the most interesting ones with regard to different architectures, considering the cases and the quality of the solutions.

The main concept behind 5G mobile network is to expand the idea of small cell network (SCN) to create a cooperative network able to cache data in active nodes inside radio access and Core network. Caching technique is a workaround to deal with bottleneck in the Back-haul, as the capacity of the wireless links could not support the increasing demand for rich multimedia. In this perspective multiple contents providers are in competition for caching space of network operator base stations. In fact, the caching space is a limited resource due to the exponential traffic of mobile data and video consumption. It is in this perspective that mobile operators and contents providers find themselves linked in this market profit generating, and consequently linked also in the allocating cache and setting price issues. In this paper we propose a multi-Stackelberg game between multiple MNOs (leaders) and several CPs (followers) computing under the Cournot-Nash assumption. In the first step a multi-leader Stackelberg game between Multiple MNO, considered as the leaders, aims to define the price they charges the CPs to maximize their profit. In the second step a multi-follower Cournot game between the CPs, considered as the followers, compete to increase the space quantity they cache at the MNOs small base stations (SBS) to maximize also their profit and to improve the quality of service (QoS) of their users. Our goal is to find the price the MNOs will set and the quantity of contents that each CP will cache. In the pricing game, each MNO first sets the price. Then the CPs react with proposed quantities of Space to cache. Then after the MNO sets again an optimal price according to the prediction of each CP's optimal strategies. Numerical results describe the structure of the Nash equilibrium and the optimal prices resulting from the MNOs and CPs optimal strategies.

Wireless Sensor Networks are deployed in different fields of application to gather data on the monitored environment. The Wireless Body Sensor Network (WBSN) is a wireless sensor network designed to monitor a human body vital and environment parameters. The design and development of such WBSN systems for health monitoring have been motivated by costly healthcare and propelled by the development of miniature health monitoring devices. This paper presents the architecture design of a preventive health care monitoring system. This architecture is designed for monitoring multiple patients in a hospital. It is based on a set of mobile data collectors and static sensors for analysis of various patient's parameters. The data collectors need to cooperate together in order to gather the data from the sensor nodes. The point of this paper is how to dynamically and effectively appoint and deploy several data collectors in the hospital to gather the measured data in minimal time. We formulate the problem as a coalitional matching game between patients and data collectors, and we propose a patient-data collector association algorithm that ensures fairness and minimum total course in the stable matchings.

In this paper, we present the Enhanced Learning Based Random Walk (ELBRW) recommender system for Places of Interest (POI), which leverages contextual information for providing more relevant POI recommendations. The ELBRW considers a model of contextual factors namely POI crowdedness based on a discrete-time Markov chain and combines user interests and "mobility homophily" for POI recommendation in Location- Based Social Networks (LBSNs). By comparing it to the Learning Based Random Walk (LBRW), a context- free recommender system, the performed experiments using LBSNs data provide promising results in terms of POI recommendation quality.

In this paper, we propose a green solution for cellular users located in hyper dense co-channel deployments of LTE small cell networks (SCNs), randomly distributed within LTE macro cell networks (MCNs). Our solution is based on a distributed sharing time access algorithm executed by small base stations (SBSs), and multi-homing capabilities of macro cellular users to improve the energy efficiency of cellular users and to satisfy their QoS throughput requirements. The theoretical analysis is validated by simulations. Our results demonstrate the improved energy efficiency of cellular users compared to the other access control mechanisms.

Cognitive radio is emerging as a promising technique to improve the utilization of the radio frequency spectrum in wireless networks. In this paper, we propose a hybrid CR (cognitive radio) system where underlay and overlay CR approaches are combined under SINR (Signal to Interference plus Noise Ratio) constraints. This new access type allows to optimize the spectrum sensing time with throughput improvement. The goal of this paper is to provide a new access mode in order to enhance the cognitive system performance while reducing the power spent for detecting the presence of the primary user and the spectrum holes. We consider the problem of spectrum sharing among primary (or “licensed”) users (PUs) and secondary (or “unlicensed”) users (SUs). In this scheme, the spectrum allocation problem is modeled as a non cooperative game, with each CR user acting as a player. Nash equilibrium is considered as the solution of this game.

Game theory is often used to find equilibria where no player can unilaterally increase its own payoff by changing its strategy without changing the strategies of other players. In this paper, we propose to use coalition formation to compute the optimized tours of mobile sinks in charge of collecting data from static wireless sensor nodes. Mobile sinks constitute a very attractive solution for wireless sensor networks, WSNs, where the application requirements in terms of node autonomy are very strong unlike the requirement in terms of latency. Mobile sinks allow wireless sensor nodes to save energy The associated coalition formation problem has a stable solution given by the final partition obtained. However, the order in which the players play has a major impact on the final result. We determine the best order to minimize the number of mobile sinks needed. We evaluate the complexity of this coalitional game as well as the impact of the number of collect points per surface unit on the number of mobile sinks needed and on the maximum tour duration of these mobile sinks. In addition, we show how to extend the coalitional game to support different latencies for different types of data. Finally, we formalize our problem as an optimization problem and we perform a comparative evaluation.

Prolonging the network lifetime is a challenging problem in wireless sensor networks. Using additional mobile agents able to gather data from static sensor nodes reveals to be one of the convenient solutions in order to conserve static sensor nodes energy. In this paper, we focus on a conflictual situation between two Data Collectors (DCs) tending to gather maximum sensed data and deliver them to a sink node before a required deadline. A non-cooperative game is used to model this behavior and evaluate the performance of the gathering scheme when two DCs are competing with each other to gain more sensed data. Under this scheme, aggressive behavior is discouraged since each data gathering operation incurs sensor nodes energy decrease and data delivery delay increase.

Recently Underwater wireless Sensor Networks (UWSNs) have been suggested as a powerful technology for many civilian and military applications, such that tactical surveillance. Geographic routing that uses the position information of nodes to route the packet toward a destination is preferable for UWSNs. In this paper, we propose a New Greedy Forwarding (NGF) strategy using splitting mechanism based on Chinese remainder theorem(CRT) for UWSNs. In the approach, source node reduced the number of bits transmitted using the proposed splitting mechanism based on CRT if there are more than two nodes participate in the forwarding of one packet. This strategy distribute the forwarding task between more nodes instead of selecting one node as next-hop, that reduce the energy consumption per node and maintain node communication for a long time. Thus resulting the increase of network life time and decrease the number of isolated/void nodes. We use topology control through depth adjustment to cope with the problem of isolated and void nodes appeared in geographic routing protocols. Simulation results shows that with the anycast greedy forwarding strategy the network life time is about 500 rounds whereas it is about 1000 rounds using the new greedy forwarding strategy, which means the new strategy increase the network life time and increase the network performance in energy saving.

It is expected that in two-tier LTE heterogeneous networks, an extensive deployment of small cell networks (SCNs) will take place in the near future, especially in dense urban zones; hence a hyper density of SCNs randomly distributed within macro cell networks (MCNs) will emerge with many overlapping zones of neighboring SCNs. Therefore, the problems of interferences in co-channel deployment will be more complicated and then the overall throughput of downlink will substantially decrease. In order to mitigate the effect of interferences in a hyper density of SCNs scenarios, a solution based on a fully distributed algorithm for sharing time access to SCNs and multi-homing capabilities of macro cellular users is proposed to improve the overall data rate of downlink and at the same time to satisfy QoS throughput requirements of macro and home cellular users. Our tentative scheme will also reduce the signaling overhead due to the absence of coordination among small base stations (SBSs) and macro base station (MBS). Results validate our solution and show the improvement attained in a hyper density of SCNs within MCNs compared to open, closed and shared time access mechanisms based on single network selection.

In this paper, we aim to improve the energy efficiency of cellular users located in hyper-dense co-channel deployments of LTE small cell Networks (SCNs), randomly distributed within LTE macro cell networks (MCNs). Avoiding the severe cross-tier interferences at the small base stations (SBSs) levels caused by the uplink transmissions between the macro indoor users (which are inside the SBS coverage area) and the macro base station (MBS), ensuring load balancing, and improving energy efficiency are critical technical challenges in hyper-dense co-channel LTE SCNs deployments. As a solution, we formulate our problem as a matching game, then we propose the deferred acceptance algorithm to compute the optimal stable matching consisting of assigning each macro indoor user to the most suitable SBS and vice versa. Simulation results validate our solution, and show how it can effectively improve the energy efficiency of cellular users in hyper-dense LTE HetNets compared to the default Max-SINR association scheme.

Abstract The core of cognitive radio paradigm is to introduce cognitive devices able to opportunistically access the licensed radio bands. The coexistence of licensed and unlicensed users prescribes an effective spectrum hole‐detection and a non‐interfering sharing of those frequencies. Collaborative resource allocation and spectrum information exchange are required but often costly in terms of energy and delay. In this paper, each secondary user (SU) can achieve spectrum sensing and data transmission through a coalitional game‐based mechanism. SUs are called upon to report their sensing results to the elected coalition head, which properly decides on the channel state and the transmitter in each time slot according to a proposed algorithm. The goal of this paper is to provide a more holistic view on the spectrum and enhance the cognitive system performance through SUs behavior analysis. We formulate the problem as a coalitional game in partition form with non‐transferable utility, and we investigate on the impact of both coalition formation and the combining reports costs. We discuss the Nash Equilibrium solution for our coalitional game and propose a distributed strategic learning algorithm to illustrate a concrete case of coalition formation and the SUs competitive and cooperative behaviors inter‐coalitions and intra‐coalitions. We show through simulations that cognitive network performances, the energy consumption and transmission delay, improve evidently with the proposed scheme. Copyright © 2016 John Wiley & Sons, Ltd.

Abstract Traditional vertical handover schemes postulate that vertical handovers (VHOs) of users come on an individual basis. This enables users to know previously the decision already made by other users, and then the choice will be accordingly made. However, in case of group mobility, almost all VHO decisions of all users, in a given group (e.g., passengers on board a bus or a train equipped with smart phones or laptops), will be made at the same time. This concept is called group vertical handover (GVHO). When all VHO decisions of a large number of users are made at the same time, the system performance may degrade and network congestion may occur. In this paper, we propose two fully decentralized algorithms for network access selection, and that is based on the concept of congestion game to resolve the problem of network congestion in group mobility scenarios. Two learning algorithms, dubbed Sastry Algorithm and Q‐Learning Algorithm, are envisioned. Each one of these algorithms helps mobile users in a group to reach the nash equilibrium in a stochastic environment. The nash equilibrium represents a fair and efficient solution according to which each mobile user is connected to a single network and has no intention to change his decision to improve his throughput. This shall help resolve the problem of network congestion caused by GVHO. Simulation results validate the proposed algorithms and show their efficiency in achieving convergence, even at a slower pace. To achieve fast convergence, we also propose a heuristic method inspired from simulated annealing and incorporated in a hybrid learning algorithm to speed up convergence time and maintain efficient solutions. The simulation results also show the adaptability of our hybrid algorithm with decreasing step size‐simulated annealing (DSS‐SA) for high mobility group scenario. Copyright © 2015 John Wiley & Sons, Ltd.

In heterogeneous network environments, users need to have mechanisms in place to decide which network is the most suitable at each moment in time for every application that the user requires. Always Best Connected is considered as a special concept to allow users to get connected to Internet using the access technology that best suits their needs or profile at any point in time. Clearly, this concept provides multiple access simultaneously for mobile users moving in heterogeneous access network environment. In this paper, we introduce a signaling game approach to achieve an always best connected service in vehicular networks. Under the considered scenario, we consider two smart vehicles named player 1 and player 2 moving in a road network area equipped with heterogeneous access networks. We assume that player 1 (super player) only has complete information on the road network, whilst player 2 has not any information. Player 1 plays first and sends a signal to player 2 which can be accurate or distorted. Based on the received signal and his belief about that signal, player 2 chooses its own action: it computes its suitable path which provides it an always best connected service.

Nowadays, Small-Cells are widely being deployed to assist and improve performance of mobile networks. Indeed, they are a promising solution to improve coverage and to offload data traffic in mobile networks. In this paper, we propose a signaling-less architecture of the heterogeneous network composed of one single Macro Base Station and a Single Small-Cell. First, we construct a game theoretic framework for channel-state independent interaction. We present many conditions for the existence of Pure Nash equilibrium. Next, and in order to capture the continuous change of the channel state, we build a random matrix game where the channel state is considered to be random (potentially ruled by some given distribution). A characterization of Nash equilibrium is provided in terms of pure strategies and mixed strategies. Convergence to Nash equilibrium is furthermore guaranteed using a variant of the well-known Combined fully distributed payoff and strategy learning. Our algorithm converges faster (only 10–20 iterations are required to converge to Nash equilibrium) and only need a limited amount of local information. This is quite promising since it says that our scheme is almost applicable for all environments (fast fading included).

In recent years, Delay Tolerant Networks (DTNs) have successfully presented as a possible extension of the Internet architecture in order to provide communication support to existing networks. However, these networks have a major issue which is the coordination among relays. In this work, we study the cooperative transmission for DTNs using coalitional game theory. We design a new incentive mechanism for heterogeneous system to induce coordination among DTN relays. In particular, we focus on the source packet dissemination to a destination using tow-hop relaying scheme, considering networks resource constraints: the relay buffer, the packet life time, and the energy consumption according to the mobile technologies. Rational mobiles are organized into separate coalition structures to meet a trade off between the source reward and the energy conservation. We discus the Nash equilibria for our game and the stable strategy state in which no mobile can get a higher payoff through changing unilaterally its coalition. Then, we use the distributed imitative Boltzmann-Gibbs learning algorithm enabling relays to learn the Nash equilibrium strategy; grand coalition. The improvement of the global system performance is examined, and a comparison between different inter-node collaboration states is presented.

A Long Term Evolution-Frequency Division Duplexing (LTE-FDD) small cell is one of the promising solutions for improving service quality and data rate in both the uplink and downlink of home users. Small cell (e.g., femtocell, picocell, microcell) is short range, low cost and low power base station installed by the indoor consumers. However, the avoidance of interferences is still an issue that needs to be addressed for successful deployment of small base stations (SBS) within existing macro cell networks mainly in co-channel deployment. Moreover, interferences are strongly dependent on the type of access control of small cells. Closed and open access are in conflict interests for macro users and home users in the uplink and downlink. To mitigate this conflict, we propose a fully distributed algorithm based on the shared time access and executed by LTE-FDD small cells, in order to reduce the effect of interferences, improve QoS of users, and maximize the overall capacity of downlink and uplink in two-tier LTE networks when small cells are deployed randomly. Simulation results validate our algorithm and show the improvement attained in offloading macro cell and satisfying QoS requirements of home users compared to the closed and open access mechanisms in both the uplink and downlink.

As (M2M) Machine-To-Machine, communication continues to grow rapidly, a full study on overload control approach to manage the data and signaling of H2H traffic from massive MTC devices is required. In this paper, a new M2M resource-scheduling algorithm for Long Term Evolution (LTE) is proposed. It provides Quality of Service (QoS) guarantee to Guaranteed Bit Rate (GBR) services, we set priorities for the critical M2M services to guarantee the transportation of GBR services, which have high QoS needs. Additionally, we simulate and compare different methods and offer further observations on the solution design.

In this paper, we propose a hybrid cognitive radio (CR) system where underlay and overlay CR approaches are combined under SIRN constraints. This new access type allows to optimize the spectrum sensing time with throughput improvement. Our proposed access type is based on a merger of these two modes depending on the number of secondary users (SU) in the Underlay Mode and on the access probabilities for these two modes. The number of all users in the Underlay access mode should not exceed a certain number of users. When a new SU arrives and if the number of SUs in Underlay mode is already equal to the predifined max number, the SU decides to switch to the Overlay mode to access the canal using this method. The SU states will be modeled with a Markov chain, the aim of this work is attempting to show that this new method allows to increase the throughput of the SU with energy optimization. The exact outage probability of secondary system is also derived for this model. Simulation results show that the performance, especiallay the throughput can be significantly increased by the proposed hybrid system.

The trend today is to access Internet through mobile devices, and people start using smartphones more than computers. This kind of services requires frequent updates through small messages from editors like in social networks. Although the use of such applications is subject to fees and consumes energy from limited batteries of smartphones. If a user activates his mobile device and has a useful contact opportunity with an access point, an update is received at the expense of monetary and energy costs. Thus, users face a tradeoff between such costs and their utilities. The goal of this paper is to show how a user can cope with such a tradeoff, by deriving a threshold policies. We consider the multi-user case, where each user try to maximize its reward based on a bonus given by the provider. We study the optimal policy in the Nash equilibrium. An optimal policy consists of deciding, based on the age of the last message received and the availability of access points whether to activate the mobile device or not. We model our problem using a POMDP with an average reward criterion. The accuracy of our model is illustrated through simulations.

We investigate energy conservation and system performance of decentralized resource allocation scheme in cognitive radio networks thoroughly based on secondary users competitive behavior. Indeed, the contention on data channel unoccupied by licensed user leads to a single winner, but also involves a loss of energy of all nodes. In this paper, we apply minority game (MG) to the most important phase from the opportunistic spectrum access (OSA) process: the sensing phase. We attempt to carry out a cooperation in a non-cooperative environment with no information exchange. We study the Nash equilibrium solution for pure and fully mixed strategies, and we use distributed learning algorithms enabling cognitive users to learn the Nash equilibrium. Finally, we provide numerical results to validate the proposed approach. The resource allocation based on minority game approach improves secondary users battery life and the performance of the network.

Using Delay Tolerant Networks (DTNs) would facilitate the connection of devices and areas across the world that are under-served by current networks. The DTNs are based on the concept of store-carry-and-forward protocols. A node may store a message in its buffer and carry it for a short or a long period of time, until an appropriate forwarding opportunity arises. A critical challenge is to determine routes through the network without even having an end-to-end connection. In order to increase the probability of message delivery, a known approach is implemented using epidemic message replication. This combination of long-term storage and message replication imposes a high storage and bandwidth overhead. Thus, efficient scheduling and dropping policies are necessary to decide which messages should be discarded when nodes' buffers operate close to their capacity. If a relay buffer is full and needs to store a new packet, it has to decide either to keep the current message or to drop it. This decision depends on the number of transmissions of the message in its buffer and the message which has just arrived. In this paper, a Partially Observed Markov Decision Process (POMDP) framework is proposed to solve the problem of buffer management in DTNs. This modeling technique predicts some properties of the optimal buffer management policy. In addition, numerical examples are presented to illustrate the findings.

To ensure connectivity in highly sparse Wireless Sensor Networks (WSNs), we consider a Ferry-assisted Wireless Sensor Network (FWSN). In our FWSN, message ferries moving along concentric annulus collect the static sensors generated packets and propagate them throughout a ferry-to-ferry forwarding schema to the sink. In this paper we present a queueing model to study and analyze the FWSN behavior. We will adapt a queuing model with finite queues which will allow us to analyze the network behavior in tens of packet loss using an analytic model. The objective of this work is to provide a way to optimize the energy consumption for each individual sensor. In our approach we consider a sensor Ferry, which is a mobile sensor with the capacity to provide the control operation upon the other fixed sensors within to network. Therefor, these fixed sensors will have to perform less control operations which will reduce their individual energy consumption. By doing so, this will automatically impact positively on to the lifetime of the network.

In this paper, we model a Virtual MIMO system using a game-theoretic approach. We are interested in the uplink, considering a non-coopertive game, where each user try to satisfy a quality of service. The uplink of a direct-sequence code division multiple access (DS-CDMA) data network is considered and a non-cooperative game is proposed in which users are allowed to choose their uplink receivers as well as to satisfy their quality of service. The utility function used in this framework is defined so that the throughput used by the user is divided into two components: the throughput received from cellular Network, and throughput received from Virtual MIMO System. In addition, this framework is used to study a constrained Nash equilibrium for the proposed game, and the impact of the interaction among users.

The concept of Participatory Sensing is centered on individuals that collect data using their smart phones (or other dedicated devices) to track the evolution of their work/living places. The main objective is to use the gathered data in order to enhance the offered quality of life. To maximize people involvement in this process we propose a ferry based architecture to leverage the contributors from charges associated with accessing service providers infrastructure to forward collected data. Opportunistic contact with message ferries will be exploited to gather data that will be carried from ferry-to-ferry till it reaches a centralized processing and decision-taking authority. We provide a closed formula for the End-to-End throughput of the proposed gathering network architecture.

Vehicular Ad Hoc Network (VANET) is considered as a special application of Mobile Ad Hoc Networks (MANETs) in road traffic, which can autonomously organize networks without infrastructure. VANETs enable vehicles on the road to communicate with each other and with road infrastructure using wireless capabilities. In the last few years, extensive research has been performed to extend Internet connectivity to VANETs. Indeed, several routing protocols have been proposed to determine routes between vehicles and gateways. In this paper, we propose a routing algorithm which is based on the Congestion Game to resolve the problem of network congestion in VANET and to provide the optimal Internet access paths. The simulation results show that the proposed routing algorithm has better feasibility and effectiveness for communicating VANETs.

The traditional vertical handover schemes postulate that vertical handover of each user comes on an individual basis. This enables the users to know previously the decision already made by other users, and then the choice will be made accordingly. However, in the case of a group vertical handover, almost all the VHO decisions - which will certainly choose the best network, will be made at the same time which will lead to system performance degradation or network congestion. In this paper, we propose a totally decentralized algorithm for network selection which based on the Congestion Game to resolve the problem of network congestion in GVHO. Therefore, the proposed algorithm named Fully Decentralized Nash Learning Algorithm with incomplete information is a prediction done by each mobile in the group that helps them to reach the Nash equilibrium. Simulation results validate the algorithm and show its robustness under two scenarios. In the first one, we examine the algorithm with a fixed number of mobiles in group to evaluate the mixed strategy and the average perceived throughput of mobiles in WIMAX and HSDPA on the basis of iteration. In the second one, we examine the algorithm with different number of mobiles in group for testing the average number of iterations needed to reach the Nash equilibrium. We also compare it with the traditional vertical handover algorithm.

In this paper, we consider a multimedia data transmission system over a wireless channel, where packets are queued at the transmitter. Multimedia data transmission over wireless networks often suffers from delay, jitter and packet loss. The main problem to implement a wireless network is the high and variable bit error rate in the radio link (fading, shadowing etc). Among the most important performance measures for real time applications are the packet loss probability. In order to improve the radio link, one often retransmits packets that have not been well received (using the Automatic Retransmission reQuest). This however may lead to queuing phenomena and increased delay due to retransmissions, and to losses of packets due to buffer overflow. In this paper, we allow that the number of retransmission is finite. We model the system state by a three-dimensional Markov chain which represents the evolution the radio link state, the number of packets in the buffer and the number of transmission of a packet in the service. We use an advanced approach based on the theory of singular perturbation in order to compute some performance of interest.

Recent advances in localization technologies and algorithms for Wireless Sensor Networks (WSN) motivate the exploitation of location information in routing protocols. In this paper we consider the geographic forwarding of sporadically generated alarm messages. Our objective is to optimize sensor's energy consumption while respecting QoS constraints on transmission delay. For instance, we propose an optimal distributed relay selection policy for WSN with duty-cycling sensors based on a Markov Decision Process (MDP) with complete information. Also, we establish sufficient conditions for optimality of threshold policies. Then, end-to-end performances for a heuristic multi-hop relay selection strategy are established. Finally, we extend our model to account for queuing capabilities at sensor level.

In recent years, many approaches and techniques have been explored for the optimization of energy usage in Wireless Sensor Networks (WSN). It is well recognized that a proper energy consumption model is the foundation for developing and evaluating a power management scheme in WSN. In this paper, we propose a new complete information Markov Decision Process (MDP) model to characterize sensors energy levels. We also propose and compare several centralized power control policies to select the more efficient policy that optimizes throughput and energy consumption.

The idea to allow unlicensed users to utilize licensed bands whenever it would not cause any interference is one of the fundaments of cognitive radio. In this paper we propose some allocation channel strategies to provide more QoS for unlicensed users. We assume that one channel is dedicated for unlicensed users that can eventually use licensed users channels if idle. The first allocation channel strategy reconsiders, in case of unavailable licensed users's channels, the request of unlicensed users to be processed by the dedicated channel. The second strategy doesn't reconsider it and simply reject it. QoS performances of the two patterns are evaluated and discussed using analytical models and simulations. For the first communication strategy, we based our model on the work of Habachi and Hayel which we enhance. We proceed to some corrections of mathematical and formulations errors. Analytical and simulation results show that the performances offered by the primary channels to secondary user are better than performances offered by the dedicated channel, however, the performance offered by the primary channels are negatively impacted by the cost of sensing.

In wireless communication systems, the delay remains a crucial factor. In this work, we focus on the design of a new opportunistic cross-layer MAC protocol involving channels allocation and packet scheduling for cognitive networks in order to optimize system performance. Cognitive radio provides the opportunity for secondary users (unlicensed users) to use available portions of the licensed spectrum bands without interfering with primary users (licensed users). We consider that each secondary user is equipped with two transceivers. The role of the first transceiver is to obtain and exchange channels information on the control channel. The second transceiver is devoted to periodically detect and dynamically use the available data channels. The paper deals with channel allocation considering traffic characteristics of secondary users. So, we propose a mechanism of resource reservation to improve Quality of Service (QoS) requirements that favors successful secondary users to transmit data during x time slots without interfering with primary users. We develop a new analytical model, while taking into account the backoff mechanism. We analyze delay parameter for two scenarios with and without resource reservation. We show through simulations that our approach guarantees an optimal response delay.

In this paper, we propose a cognitive radio based Medium Access Control (MAC) protocol for packet scheduling in wireless networks. Cognitive MAC protocols allow a class of users, called secondary users, to identify the unused frequency spectrum and to communicate without interfering with the primary users. In our proposed MAC protocol, each secondary user is equipped with two transceivers. One of the transceivers is used for control messages while the other periodically senses and dynamically utilizes the unused data channel. The secondary users report the status of channels on control channel and negotiate on the selected data channel itself for onward data transmission. Each channel is used by different set of secondary users. Contrary to existing protocols, data transmission takes place in the time slot in which spectrum opportunity is found. We develop a new analytical model, while taking into account the backoff mechanism. Our simulation results show that throughput increases with the increase in the number of channels.

In this paper, we consider a multimedia data transmission system over a wireless channel, where packets are queued at the transmitter. Multimedia data transmission over wireless networks often suffers from delay, jitter and packet loss. The main problem to implement a wireless network is the high and variable bit error rate in the radio link (fading, shadowing etc), making it necessary to use an additional error control mechanism. Among the most important performance measures for real time applications are the packet loss probability and expected delay. In order to improve the radio link, one often retransmits packets that have not been well received (using the Automatic Retransmission reQuest - ARQ). This however may lead to queuing phenomena and increased delay due to retransmissions, and to losses of packets due to buffer overflow. In this paper, we assume that the number of retransmission is finite. We present a queuing analysis which based on matrix-geometric solutions in M/G/1 type Markov chains, in order to compute some performance measures of interest.