In the Metaverse, real-time digital twin (DT) updates enable immersive, interactive environments by reflecting real world states. Metaverse can engage its users to share data in order to ensure the completeness of the DT. However, the limitations and heterogeneity in IoT devices' computation and transmission resources are critical challenges to synchronizing the vast volume of real-world objects with their digital replicas. In this work, we propose a novel adaptive Vision Transformer (ViT)- based semantic communication (SemCom) system to extract semantic information from raw data collected by IoT devices. The system dynamically adjusts the model size and computational complexity according to the resource constraints of individual IoT devices, enables broader participation of heterogeneous devices, enhances feature extraction capabilities, and ensures completeness and efficient DT representation. We formulate our problem as an utility-maximization optimization to manage ViT complexity under resource constraints, allowing the Metaverse Services Provider (MSP) to select high-performing IoT devices. To guide the optimization, we profile ViT models with varying architectural complexities and conduct an empirical analysis to capture the relationship between model scale and performance. To address the scalability and privacy limitations of the optimization, we propose a decentralized solution where each IoT device independently optimizes its own utility under local constraints. The MSP, in turn, selects among these devices to ensure quality and maximize its overall utility. We show that our distributed solution achieves near-optimal performance and significantly outperforms other approaches in terms of MSP utility, semantic data quality, and overall IoT utility. © 2026 IEEE.

The integration of Unmanned Aerial Vehicles (UAVs) into Open Radio Access Networks (O-RAN) enhances communication in disaster management and Search and Rescue (SAR) operations by ensuring connectivity when infrastructure fails. However, SAR scenarios demand stringent security and low-latency communication, as delays or breaches can compromise mission success. While UAVs serve as mobile relays, they introduce challenges in energy consumption and resource management, necessitating intelligent allocation strategies. Existing UAV-assisted O-RAN approaches often overlook the joint optimization of security, latency, and energy efficiency in dynamic environments. This paper proposes a novel Reinforcement Learning (RL)-based framework for dynamic resource allocation in UAV relays, explicitly addressing these trade-offs. Our approach formulates an optimization problem that integrates security-aware resource allocation, latency minimization, and energy efficiency, which is solved using RL. Unlike heuristic or static methods, our framework adapts in real-time to network dynamics, ensuring robust communication. Simulations demonstrate superior performance compared to heuristic baselines, achieving enhanced security and energy efficiency while maintaining ultralow latency in SAR scenarios.

Open Radio Access Networks (O-RAN) are transforming telecommunications by shifting from centralized to distributed architectures, promoting flexibility, interoperability, and innovation through open interfaces and multi-vendor environments. However, O-RAN's reliance on cloud-based architecture and enhanced observability introduces significant security and resource management challenges. Efficient resource management is crucial for secure and reliable communication in O-RAN, within the resource-constrained environment and heterogeneity of requirements, where multiple User Equipment (UE) and O-RAN Radio Units (O-RUs) coexist. This paper develops a framework to manage these aspects, ensuring each O-RU is associated with UEs based on their communication channel qualities and computational resources, and selecting appropriate encryption algorithms to safeguard data confidentiality, integrity, and authentication. A Multi-objective Optimization Problem (MOP) is formulated to minimize latency and maximize security within resource constraints. Different approaches are proposed to relax the complexity of the problem and achieve near-optimal performance, facilitating tradeoffs between latency, security, and solution complexity. Simulation results demonstrate that the proposed approaches are close enough to the optimal solution, proving that our approach is both effective and efficient.

Generative AI (GAI) and Large Language Models (LLMs) have revolutionized natural language processing and content creation. However, their significant computational demands during inference often require cloud servers, which are currently the only viable option for handling complex multi-modal models like GPT-4. The inherent complexity of these models increases latency, posing challenges even within cloud environments. Furthermore, cloud reliance brings other challenges, including high bandwidth consumption to transfer diverse data types. Worse, in personalized GAI applications like virtual assistants, similar prompts frequently occur, causing redundant transmission and computation of replies, which further increases overhead. Accelerating the inference of multi-modal systems is, therefore, critical in artificial intelligence. In this paper, we aim to improve the inference efficiency through prompt caching; if a current prompt is semantically similar to a previous one, the system can reuse the earlier response without invoking the model again. We leverage collaborative edge computing to cache popular replies and store their request embeddings. New prompts are locally processed to extract embeddings, with their qualities determined by the resources available on edge servers. Our problem is formulated as an optimization to manage offloading decisions for GAI tasks, aiming to avoid cloud inferences and minimize latency while maximizing reply quality. Given its non-convex nature, we propose to solve it via Block Successive Upper Bound Minimization (BSUM). Reinforcement learning is employed to actively pre-cache prompts, tackling the complexity of unknown prompt popularity. Our approach demonstrates near-optimal performance, significantly outperforming cloud-only solutions.

In this paper, a robust reconfigurable intelligent surface (RIS)-aided carrier frequency offset (CFO)-based physical layer authentication (PLA) scheme for wireless networks is proposed. The considered network consists of a legitimate transmitter, a spoofer, and a receiver, acting as an authenticator, who aims to identify the sender's legitimacy relying on the estimated CFO from received signals. Thus, the proposed scheme exploits an RIS to increase the received signal-to-noise ratio (SNR) and enhance the authentication performance. A deep reinforcement learning framework is developed to jointly optimize the RIS phase shifts and the preamble length to maximize the authentication performance under a minimal channel capacity constraint. Then, a supervised machine learning classifier is employed for node authentication, exploiting the optimized RIS reflection and preamble length. The results show that the authentication performance is enhanced with the increase in the RIS size and the difference between the transmitters' CFOs. Also, the proposed scheme outperforms the baseline RIS-aided CSI-based one in mobility scenarios.

In the 5G Advanced and 6G era, wireless communication systems face security challenges, notably adversarial interference from unknown jammers in multi-user scenarios. Reconfigurable Intelligent Surfaces (RIS) present a cost-effective solution due to their low power consumption and easy deployment. Existing RIS techniques typically address simple jamming scenarios with a single static jammer, focusing on a single objective. This study introduces a multi-objective optimization approach deploying UAV-mounted RIS to counter jamming threats in wireless communications within a densely populated smart city environment. The proposed solution aims to safeguard essential services from potential disruptions caused by malicious jamming attacks during public events. We employ Proximal Policy Optimization (PPO), a lightweight Deep Reinforcement Learning (DRL) technique, to concurrently optimize the trajectory of UAV and RIS passive beamforming to address computational complexity. The objectives include maximizing the average sum rate and minimizing energy consumption. Our experiments highlight the efficacy of the PPO-based strategy, demonstrating significant improvements in average sum rates and energy efficiency amid numerous mobile devices and moving jammers. Importantly, our proposed system model outperforms a baseline from related works in maximizing the sum rate and minimizing overall energy consumption.

The metaverse, the next digital frontier, demands high-performance models and quick personalization due to the dynamic nature of user tasks despite limited data availability. The frequent user customization is resource-intensive and data-heavy. Meta-learning, especially federated meta-learning (FML) known for its adaptive capabilities, is crucial for addressing the dynamics in metaverse, characterized by user heterogeneity, diverse data structures, and varied tasks. However, the diversity of tasks can compromise global training outcomes due to statistical heterogeneity. Given this, an urgent need arises for smart coalition formation that accounts for these disparities. This paper proposes a game-theoretic framework for managing FML in metaverse services, with meta-learners as workers. A blockchain-based cooperative coalition formation game is introduced, grounded on a reputation metric, the similarity of users, and their incentives. The reputation metric is derived based on our novel reputation system, which takes into account users' historical contributions and potential contributions to current tasks, by exploiting the correlations between past and new tasks. Meanwhile, the incentive mechanism is formulated as an optimization to minimize users energy cost and boost the users contribution for higher federated meta-learning efficacy. Simulations show the framework's resilience against misbehavior and its superiority over other schemes, improving service utility and worker profitability in metaverse meta-learning.

Deep Neural Networks (DNNs) have become the key technique to revolutionize the healthcare sector. However, conducting online remote inference is often impractical due to privacy constraints and latency requirements. To enable local computation, researchers have attempted network pruning with minimal accuracy loss or DNN distribution without affecting the performance. Yet, distributed inference can be inefficient due to the energy overhead and fluctuation of communication channels between participants. On the other hand, given that realistic healthcare systems use pre-trained models, local pruning and retraining relying only on the available scarce data is not possible. Even pre-pruned DNNs are limited in their ability to customize to the local load of data and device dynamics. The online pruning of DNN inferences without retraining is viable; however, it was not considered in the literature as most well-known techniques do not perform well without adjustment. In this paper, we propose a novel pruning strategy using Explainable AI (XAI) to enhance the performance of pruned DNNs without retraining, a necessity due to the scarcity and bias of local healthcare data. We combine distribution and pruning techniques to perform online distributed inference assisted by dynamic pruning when needed for highest accuracy. We use Non-Linear Integer Programming (NLP) to formulate our approach as a trade-off between resources and accuracy, and Reinforcement Learning (RL) to relax the problem and adapt to dynamic requirements. Our pruning criterion shows high performance compared to other reference techniques and ability to assist distribution by reducing resource usage while keeping high accuracy.

The metaverse, envisioned as the next digital frontier for avatar-based virtual interaction, involves high-performance models. In this dynamic environment, users’ tasks frequently shift, requiring fast model personalization despite limited data. This evolution consumes extensive resources and requires vast data volumes. To address this, meta-learning emerges as an invaluable tool for metaverse users, with federated meta-learning (FML), offering even more tailored solutions owing to its adaptive capabilities. However, the metaverse is characterized by users heterogeneity with diverse data structures, varied tasks, and uneven sample sizes, potentially undermining global training outcomes due to statistical difference. Given this, an urgent need arises for smart coalition formation that accounts for these disparities. This paper introduces a dual game-theoretic framework for metaverse services involving meta-learners as workers to manage FML. A blockchain-based cooperative coalition formation game is crafted, grounded on a reputation metric, user similarity, and incentives. We also introduce a novel reputation system based on users’ historical contributions and potential contributions to present tasks, leveraging correlations between past and new tasks. Finally, a Stackelberg game-based incentive mechanism is presented to attract reliable workers to participate in meta-learning, minimizing users’ energy costs, increasing payoffs, boosting FML efficacy, and improving metaverse utility. Results show that our dual game framework outperforms best-effort, random, and non-uniform clustering schemes -improving training performance by up to 10%, cutting completion times by as much as 30%, enhancing metaverse utility by more than 25%, and offering up to 5% boost in training efficiency over non-blockchain systems, effectively countering misbehaving users.

Although Deep Neural Networks (DNN) have become the backbone technology of several Internet of Things (IoT) applications, their execution in resource-constrained devices remains challenging. To cater for these challenges, collaborative deep inference conducted by IoT devices was introduced. However, the prevalence of DNN computation suffers from severe privacy problems, e.g. data-reverse and model leakage. Particularly, malicious participants can accurately recover the received data to access sensitive information. Furthermore, the system is composed of heterogeneous data-sources represented by different DNN models that wish to execute classifications without exposing their data and models. Though, relaying the trained models to a centralized unit managing the collaboration leads to major risks because some features can be revealed through these models, in addition to dependency and scalability problems. In this paper, we present an approach that targets the privacy of collaborative inference via controlling the amount of data assigned to different participants, to prevent them from reversing attempts. Moreover, each independent data-source requesting inference will be responsible to manage the distribution of its DNN locally. In this context, different sources are required to compete over the pervasive resources while cooperating to maintain privacy welfare. We formulate this methodology, as an integer programming problem, where we establish a trade-off between the latency of co-inference and the privacy required by heterogeneous entities. A distributed solution scheme is also developed based on the Lagrangian dual problem. Next, to relax the optimization, we shape our approach as a cooperative and competitive Multi-Agent Reinforcement Learning (MARL) that supports heterogeneous/independent agents. Our comprehensive simulations demonstrated that our method yields results on par with those of a single RL agent in terms of action performance, while maintaining the privacy of individual agents’ information. Additionally, it surpasses the Independent Q-Learning (IQL) approach, where agents operate autonomously, in safeguarding inference privacy.

In this paper, we propose a Deep Reinforcement Learning (DRL) framework to optimize the secrecy performance of a Multi-User (MU)-Multiple-Input Single-Output (MISO) Underwater Optical Wireless Communication (UOWC) system. The network consists of several light-emitting diodes connected with various underwater users through optical beams. The legitimate transmission is threatened by several eavesdroppers attempting to overhear the confidential message sent to each user. Thus, digital precoding is employed to cancel the inter-user interference and maximize the per-user secrecy rate and, consequently, the secrecy sum rate (SSR). Leveraging the developed DRL algorithm, the MU-MISO precoding matrix is optimized for enhancing the system's SSR. Numerical results show the superiority of the proposed DRL framework compared to the baseline zero-forcing and random pre coding schemes, even with corrupted CSI at the transmitter due to seawater dynamics and estimation errors.

The healthcare sector has undergone a significant transformation with the widespread adoption of Deep Neural Networks (DNN). However, due to privacy constraints and stringent latency requirements, online remote inference is not a viable option in healthcare scenarios. Many efforts have been conducted to enable local computation, such as network compression using pruning or DNN distribution among multiple resource-constrained devices. Yet, it is still challenging to conduct distributed inference due to the latency and energy overheads resulting from intermediate shared data. On the other hand, given that realistic healthcare systems use pre-trained models, local pruning and fine-tuning relying only on the scarce and biased data is not possible. Even pre-pruned DNNs are not efficient as they are not customized to the local load of data and the dynamics of devices. The dynamic and online pruning of DNN without fine-tuning is a promising solution; however, it was not considered in the literature as most well-known techniques do not perform well without adjustment. In this paper, driven by the data restrictions in healthcare sector, we propose a novel pruning strategy based on Explainable AI (XAI), with a target to enhance the pruned DNN performance without fine-tuning. Moreover, to maintain the highest possible accuracy, we propose to combine distribution and pruning techniques to perform online distributed inference assisted by dynamic pruning only when needed. Our experiments show the performance of our pruning criterion compared to other reference techniques, in addition to its ability to assist the distribution by reducing the shared data, while keeping high accuracy.

Crowdsourced live streaming services (CLS) present significant challenges due to massive data size and dynamic user behavior. Service providers must accommodate personalized QoE requests, while managing computational burdens on edge servers. Existing CLS approaches use a single edge server for both transcoding and user service, potentially overwhelming the selected node with high computational demands. In response to these challenges, we propose the Reinforcement Learning-based-Collaborative Edge-Assisted Live Streaming (RL-CEALS) framework. This innovative approach fosters collaboration between edge servers, maintaining QoE demands and distributing computational burden cost-effectively. By sharing tasks across multiple edge servers, RL-CEALS makes smart decisions, efficiently scheduling serving and transcoding of CLS. The design aims to minimize the streaming delay, the bitrate mismatch, and the computational and bandwidth costs. Simulation results reveal substantial improvements in the performance of RL-CEALS compared to recent works and baselines, paving the way for a lower cost and higher quality of live streaming experience.

As deep neural networks continue to expand and become more complex, most edge devices are unable to handle their extensive processing requirements. Therefore, the concept of distributed inference is essential to distribute the neural network among a cluster of nodes. However, distribution may lead to additional energy consumption and dependency among devices that suffer from unstable transmission rates. Unstable transmission rates harm real-time performance of IoT devices causing low latency, high energy usage, and potential failures. Hence, for dynamic systems, it is necessary to have a resilient DNN with an adaptive architecture that can downsize as per the available resources. This paper presents an empirical study that identifies the connections in ResNet that can be dropped without significantly impacting the model’s performance to enable distribution in case of resource shortage. Based on the results, a multi-objective optimization problem is formulated to minimize latency and maximize accuracy as per available resources. Our experiments demonstrate that an adaptive ResNet architecture can reduce shared data, energy consumption, and latency throughout the distribution while maintaining high accuracy.

The deployment flexibility and maneuverability of unmanned aerial vehicles (UAVs) increased their adoption in various applications, such as wildfire tracking, border monitoring, etc. In many critical applications, UAVs capture images and other sensory data and then send the captured data to remote servers for inference and data processing tasks. However, this approach is not always practical in real-time applications due to the connection instability, limited bandwidth, and end-to-end latency. One promising solution is to divide the inference requests into multiple parts (layers or segments), with each part being executed in a different UAV based on the available resources. Furthermore, some applications require the UAVs to traverse certain areas and capture incidents; thus, planning their paths becomes critical particularly, to reduce the latency of making the collaborative inference process. Specifically, planning the UAVs trajectory can reduce the data transmission latency by communicating with devices in the same proximity while mitigating the transmission interference. This work aims to design a model for distributed collaborative inference requests and path planning in a UAV swarm while respecting the resource constraints due to the computational load and memory usage of the inference requests. The model is formulated as an optimization problem and aims to minimize latency. The formulated problem is NP-hard so finding the optimal solution is quite complex; thus, this article introduces a real-time and dynamic solution for online applications using deep reinforcement learning. We conduct extensive simulations and compare our results to the state-of-the-art studies demonstrating that our model outperforms the competing models.

Remote monitoring systems analyze the environment dynamics in different smart industrial applications, such as occupational health and safety, and environmental monitoring. Specifically, in Industrial Internet of Things (IIoT) systems, the huge number of devices and the expected performance put pressure on resources, such as computational, network, and device energy. Distributed training of machine and deep learning (ML/DL) models for intelligent industrial IoT applications is very challenging for resource limited devices over heterogeneous wireless networks (HetNets). Hierarchical federated learning (HFL) performs training at multiple layers offloading the tasks to nearby multiaccess edge computing (MEC) units. In this article, we propose a novel energy-efficient HFL framework enabled by wireless energy transfer (WET) and designed for heterogeneous networks with massive multiple-input–multiple-output (MIMO) wireless backhaul. Our energy-efficiency approach is formulated as a mixed-integer nonlinear programming (MINLP) problem, where we optimize the HFL device association and manage the wireless transmitted energy. However due to its high complexity, we design a heuristic resource management algorithm, namely, H2RMA, that respects energy, channel quality, and accuracy constraints, while presenting a low-computational complexity. We also improve the energy consumption of the network using an efficient device scheduling scheme. Finally, we investigate device mobility and its impact on the HFL performance. Our extensive experiments confirm the high performance of the proposed resource management approach in HFL over HetNets, in terms of training loss and grid energy costs.

In this paper, we investigate an energy-efficient Long Range (LoRa) wireless network powered by hybrid energy which consists of an energy harvesting source and the grid. The grid allows to compensate for the randomness and intermittency of the harvested energy. The aim is to propose a dynamic energy-efficient resource management scheme for LoRa wireless networks that enables green Internet of Things (IoT). Hence, we formulate a grid energy cost minimization problem subject to minimum received signal-to-noise ratio (SNR), and channel, spreading factor (SF) and energy availability constraints. The formulated problem is simplified and decoupled into two sub-problems which allows to derive the optimal resource management solution but with high computational complexity. Then, we propose a low complexity heuristic channel and SF assignment, and energy management algorithm for dynamic LoRa wireless networks. Numerical results shows the efficient use of renewable energy in green dynamic LoRa wireless networks.

LoRa wireless networks are considered as a key enabling technology for next-generation Internet of Things (IoT) systems. New IoT deployments (e.g., smart city scenarios) can have thousands of devices per square kilometer leading to huge amount of power consumption to provide connectivity. In this article, we investigate green LoRa wireless networks powered by a hybrid of the grid and renewable energy sources, which can benefit from harvested energy while dealing with the intermittent supply. This article proposes resource management schemes of the limited number of channels and spreading factors (SFs) with the objective of improving the LoRa gateway energy efficiency. First, the problem of grid power consumption minimization while satisfying the system’s quality of service demands is formulated. Specifically, both scenarios the uncorrelated and time-correlated channels are investigated. The optimal resource management problem is solved by decoupling the formulated problem into two subproblems: 1) channel and SF assignment problem and 2) energy management problem. Since the optimal solution is obtained with high complexity, online resource management heuristic algorithms that minimize the grid energy consumption are proposed. Finally, taking into account the channel and energy correlation, adaptable resource management schemes based on reinforcement learning (RL) are developed. Simulation results show that the proposed resource management schemes offer efficient use of renewable energy in LoRa wireless networks.

Unmanned aerial vehicles (UAVs) have attracted great interest in the last few years owing to their ability to cover large areas and access difficult and hazardous target zones, which is not the case of traditional systems relying on direct observations obtained from fixed cameras and sensors. Furthermore, thanks to the advancements in computer vision and machine learning, UAVs are being adopted for a broad range of solutions and applications. However, deep neural networks (DNNs) are progressing toward deeper and complex models that prevent them from being executed onboard. In this article, we propose a DNN distribution methodology within UAVs to enable data classification in resource-constrained devices and avoid extra delays introduced by the server-based solutions due to data communication over air-to-ground links. The proposed method is formulated as an optimization problem that aims to minimize the latency between data collection and decision-making while considering the mobility model and the resource constraints of the UAVs as part of the air-to-air communication. We also introduce the mobility prediction to adapt our system to the dynamics of UAVs and the network variation. The simulation conducted to evaluate the performance and benchmark the proposed methods, namely, optimal UAV-based layer distribution (OULD) and OULD with mobility prediction (OULD-MP), was run in an HPC cluster. The obtained results show that our optimization solution outperforms the existing and heuristic-based approaches.

LoRa supports the exponential growth of connected devices. In this paper, we investigate green LoRa wireless networks powered by both the grid power and a renewable energy source. The grid power compensates for the randomness and intermittency of the harvested energy. We propose an efficient and smart resource management scheme of the limited number of channels and spreading factors (SFs) with the objective of improving the LoRa gateway (LG) energy efficiency. We formulate the problem of grid power consumption minimization while satisfying the quality of service demands. The optimal resource management problem is solved by decoupling the formulated problem into two sub-problems: channel and SF assignment problem and energy management problem. Next, we develop an adaptable resource management schemes based on Reinforcement Learning (RL) taking into account the channel and energy correlation. Simulations results show that the proposed resource management schemes offer efficient use of renewable energy in LoRa wireless networks.

Recognizing the patient's emotions using deep learning techniques has attracted significant attention recently due to technological advancements. Automatically identifying the emotions can help build smart healthcare centers that can detect depression and stress among the patients in order to start the medication early. Using advanced technology to identify emotions is one of the most exciting topics as it defines the relationships between humans and machines. Machines learned how to predict emotions by adopting various methods. In this survey, we present recent research in the field of using neural networks to recognize emotions. We focus on studying emotions' recognition from speech, facial expressions, and audio-visual input and show the different techniques of deploying these algorithms in the real world. These three emotion recognition techniques can be used as a surveillance system in healthcare centers to monitor patients. We conclude the survey with a presentation of the challenges and the related future work to provide an insight into the applications of using emotion recognition.

Unmanned Aerial Vehicles (UAVs) have recently attracted significant attention due to their outstanding ability to be used in different sectors and serve in difficult and dangerous areas. Moreover, the advancements in computer vision and artificial intelligence have increased the use of UAVs in various applications and solutions, such as forest fires detection and borders monitoring. However, using deep neural networks (DNNs) with UAVs introduces several challenges of processing deeper networks and complex models, which restricts their on-board computation. In this work, we present a strategy aiming at distributing inference requests to a swarm of resource-constrained UAVs that classifies captured images on-board and finds the minimum decision-making latency. We formulate the model as an optimization problem that minimizes the latency between acquiring images and making the final decisions. The formulated optimization solution is an NP-hard problem. Hence it is not adequate for online resource allocation. Therefore, we introduce an online heuristic solution, namely DistInference, to find the layers placement strategy that gives the best latency among the available UAVs. The proposed approach is general enough to be used for different low decision-latency applications as well as for all CNN types organized into pipeline of layers (e.g., VGG) or based on residual blocks (e.g., ResNet).

With the emergence of smart cities, Internet of Things (IoT) devices as well as deep learning technologies have witnessed an increasing adoption. To support the requirements of such paradigm in terms of memory and computation, joint and real-time deep co-inference framework with IoT synergy was introduced. However, the distribution of Deep Neural Networks (DNN) has drawn attention to the privacy protection of sensitive data. In this context, various threats have been presented, including black-box attacks, where a malicious participant can accurately recover an arbitrary input fed into his device. In this paper, we introduce a methodology aiming to secure the sensitive data through re-thinking the distribution strategy, without adding any computation overhead. First, we examine the characteristics of the model structure that make it susceptible to privacy threats. We found that the more we divide the model feature maps into a high number of devices, the better we hide proprieties of the original image. We formulate such a methodology, namely DistPrivacy, as an optimization problem, where we establish a trade-off between the latency of co-inference, the privacy level of the data, and the limited-resources of IoT participants. Due to the NP-hardness of the problem, we introduce an online heuristic that supports heterogeneous IoT devices as well as multiple DNNs and datasets, making the pervasive system a general-purpose platform for privacy-aware and low decision-latency applications.

Leveraging video caching to collaborative Mobile Edge Computing (MEC) servers is an emerging paradigm, where cloud computing services are extended to edge networks to allocate multimedia contents close to end-users. However, despite minimizing the traffic over the content delivery networks (CDN), congestions may occur in peak hours characterized by high load demands. Involving users' devices in data offloading through Device-to-Device (D2D) connections has proved its efficiency in relieving the cellular spectrum utilization. In this paper, the Collaborative Edge network (CE) and the devices (D2D) cluster are combined to form a CE-D2D framework aiming at maximizing video caching and efficiently using cellular and backhaul bandwidths. However, since we are dealing with large sized contents, the small storage and bandwidth capacities offered by users limit the number of cached videos and restrict offloading large volume data. This makes the CE-D2D framework, so far, an incomplete solution for multimedia contents. Therefore, we propose a caching strategy to cache only the chunks of videos to be watched and instead of caching or offloading each video content by one edge node (as performed in literature), helpers (MEC and mobiles) will collaborate to store and share different chunks to optimize the storage/transmission resources usage. In this work, we model both CE and D2D frameworks as linear programs and schedule the collaboration between them constrained by resource availability. Due to the NP-hardness of the problem, we introduce an online heuristic that presents a proactive chunks caching (HLPC) and a near-optimal data offloading with polynomial complexity.

With the advancement of rich media generating devices, the proliferation of live Content Providers (CP), and the availability of convenient internet access, crowdsourced live streaming services have witnessed unexpected growth. To ensure a better Quality of Experience (QoE), higher availability, and lower costs, large live streaming CPs are migrating their services to geo-distributed cloud infrastructure. However, because of the dynamics of live broadcasting and the wide geo-distribution of viewers and broadcasters, it is still challenging to satisfy all requests with reasonable resources. To overcome this challenge, we introduce in this paper a prediction driven approach that estimates the potential number of viewers near different cloud sites at the instant of broadcasting. This online and instant prediction of distributed popularity distinguishes our work from previous efforts that provision constant resources or alter their allocation as the popularity of the content changes. Based on the derived predictions, we formulate an Integer-Linear Program (ILP) to proactively and dynamically choose the right data center to allocate exact resources and serve potential viewers, while minimizing the perceived delays. As the optimization is not adequate for online serving, we propose a real-time approach based on Reinforcement Learning (RL), namely RL-OPRA, which adaptively learns to optimize the allocation and serving decisions by interacting with the network environment. Extensive simulation and comparison with the ILP have shown that our RL-based approach is able to present optimal results compared to heuristic-based approaches.

To support multimedia applications, Mobile Edge Computing (MEC) servers offer storage and computing capacities to handle videos close to end-users. However, the high load in peak hours consumes the limited available bandwidth of existing cellular and backhaul links leading to low network performance. Hence, an elastic system model is required to maintain the high Quality of Experience (QoE) as the resource demands increase. Caching popular videos at mobile devices is considered a promising technique for content delivery. Yet, mobile users offer small capacities that are not adequate for large-sized video sharing. In this paper, we extend the collaborative caching and processing framework in edge networks (Collaborative Edge - CE) to include the users' mobile video sharing (Device-to-Device - D2D). We propose a caching strategy to cache only the chunks of videos to be watched and instead of offloading one video content by one edge node, helpers (MEC servers and users) will collaborate to store and share different chunks to optimize the storage/transmission resources usage. To only cache popular contents, we designed a D2D-aware proactive chunks caching on users’ devices based on our chunks popularity model. Next, we formulate this CE-D2D collaborative problem as a linear program. Due to the NP-hardness of the problem, we introduce a sub-optimal relaxation and an online heuristic using the proactive caching and presenting a near optimal data offloading and a profitable payment determination, with polynomial time complexity. The simulation results show that our policies and heuristics outperform other edge caching approaches by more than 10% in terms of hit ratio, average delay, and cost.

During the last decade, empowered by the technological advances of mobile devices and the revolution of wireless mobile network access, crowdsourced live streaming has become more popular. Ensuring a stable high-quality playback experience is necessary to maximize the number of viewers and profits for content providers. Additionally, because of the instability of network conditions and the heterogeneity of the end-users capabilities, transcoding the original video into multiple bitrates is required. Video transcoding is a computationally exhaustive process, where generally a single cloud instance needs to be reserved to produce one single video bitrate representation. On-demand renting of resources or inadequate resources pre-renting may cause delay of the video playback or serving the viewers with a lower quality. On the other hand, if resources provisioning is much higher than required, the extra resources will be wasted. In this paper, we introduce our resources reservation framework for geo-distributed cloud sites, to maximize the Quality of Experience (QoE) of viewers and minimize the cost to the content providers. First, we formulate an offline optimization problem to allocate transcoding resources at the viewers' proximity, while creating a trade off between the network cost and viewers QoE. Second, based on the optimizer resource allocation decisions on historical live videos, we create our time series datasets containing historical records of the optimal resources needed at each geo-distributed cloud site. Finally, we adopt machine learning to build our distributed time series forecasting models to proactively forecast the exact needed transcoding resources ahead of time at each geo-distributed cloud site.

Driven by the tremendous technological advancement of personal devices and the prevalence of wireless mobile network accesses, the world has witnessed an explosion in crowdsourced live streaming. Ensuring a better viewers quality of experience (QoE) is the key to maximize the audiences number and increase streaming providers' profits. This can be achieved by advocating a geo-distributed cloud infrastructure to allocate the multimedia resources as close as possible to viewers, in order to minimize the access delay and video stalls. Moreover, allocating the exact needed resources beforehand avoids over-provisioning, which may lead to significant costs by the service providers. In the contrary, under-provisioning might cause significant delays to the viewers. In this paper, we introduce a prediction driven resource allocation framework, to maximize the QoE of viewers and minimize the resource allocation cost. First, by exploiting the viewers locations available in our unique dataset, we implement a machine learning model to predict the viewers number near each geo-distributed cloud site. Second, based on the predicted results that showed to be close to the actual values, we formulate an optimization problem to proactively allocate resources at the viewers proximity. Additionally, we will present a trade-off between the video access delay and the cost of resource allocation.

The advancement of technology has pushed the cloud computing capabilities to the edge networks, paving the way for network operators and multimedia service providers to leverage video caching and transcoding to the Mobile Edge computing (MEC) servers. However, because of high load demands in peak hours and the limited bandwidth, congestions may occur, leading to lower network performance. This problem is worsened by the redundancies of the same content requests and the expectation for the highest video quality, which exhausts the cellular spectrum and the backhaul links. Collaboration between edge servers, to cache and transcode videos, is proposed and proved its efficiency to store highly requested contents and relieve the load on the backhaul links. Meanwhile, Device-to-device (D2D) offloading is considered to alleviate cellular spectrum utilization. In this paper, we will extend the Collaborative Edge (CE) network to include the mobile users (D2D) caching and offloading and create a CE-D2D dual framework. Still, this framework does not present the perfect solution for exhaustive bandwidth utilization. In fact, users' mobiles have small storage capacities and very scarce bandwidth availability, which limits the number of cached videos and constraints sharing large sized contents. Additionally, existing D2D approaches are considering video requests as simple contents similar to any HTML page request. However, in realistic cases, a lower bitrate version of the content can be offered to the viewers, if the bandwidth (cellular or mobile) is unavailable. Hence, to maximize the caching efficiency, we formulate the CE-D2D framework as a linear program, where MEC servers and users' mobiles collaborate to cache and offload different chunks of the requested content constrained by cache and bandwidth availability. In this way, instead of caching and serving full popular videos, as done in previous works, we will only cache popular chunks within different helpers, which maximizes the efficiency of caching in small storage devices.

Data Center Networks (DCNs) are attracting immense interest from the industry, research and academia to keep pace with the increase of Internet services demands. One of the major concerns that draws the attention of researchers is the exponential growth of the energy consumption and carbon emission of the DCNs. Studies conducted to identify the causes of the increasing energy consumption have proved that the growing size of computing demand, the over-provisioning of the networking resources, the under-utilization of the infrastructure, the fault-tolerance, the high bandwidth exigence and the inefficient hardware and cooling structure are leading to considerable energy waste. Therefore, in recent years, new data center (DC) architectures are proposed where new hardware types and new technologies are implemented for the sake of energy efficiency. Other efforts are focusing on designing algorithms and strategies to enhance the utilization of the network resources. Replacing brown power by renewable energy was also one of the attractive ideas to minimize the energy costs. In this survey paper, we will present energy-related problems in data centers and review the state of the art of the research literature on energy efficient architectures, techniques, technologies, resource management, and thermal control and monitoring. Additionally, we present the challenges facing each approach and the strategies to build a green DC. This paper serves as a specification document that shows step by step how to minimize the energy consumption of different components of the system.

Recently, the growing demand for rich multimedia content such as Video on Demand (VoD) has made the data transmission from content delivery networks (CDN) to end-users quite challenging. Edge networks have been proposed as an extension to CDN networks to alleviate this excessive data transfer through caching and to delegate the computation tasks to edge servers. To maximize the caching efficiency in the edge networks, different Mobile Edge Computing (MEC) servers assist each others to efficiently select which content to store and the appropriate computation tasks to process. In this paper, we adopt a collaborative caching and transcoding model for VoD in MEC networks. However, unlike other models in the literature, different chunks of the same video are not fetched and cached in the same MEC server. Instead, neighboring servers will collaborate to store and transcode different video chunks and consequently optimize the limited resources usage. Since we are dealing with chunks caching and processing, we propose to maximize the edge efficiency by studying the viewers watching pattern and designing a probabilistic model where chunks popularities are evaluated. Based on this model, popularity-aware policies, namely Proactive caching policy (PcP) and Cache replacement Policy (CrP), are introduced to cache only highest probably requested chunks. In addition to PcP and CrP, an online algorithm (PCCP) is proposed to schedule the collaborative caching and processing. The evaluation results prove that our model and policies give better performance than approaches using conventional replacement policies. This improvement reaches up to 50% in some cases.

This paper presents PTNet, a new data center topology that is specifically designed to offer a high and parameterized scalability with just one layer architecture. Furthermore, despite its high scalability, PTNet grants a reduced latency and a high performance in terms of capacity and fault tolerance. Consequently, compared to widely known data center networks, our new topology shows better capacity, robustness, cost-effectiveness and less power consumption. Conducted experiments and theoretical analyses illustrate the performance of the novel system.