Smart grids, powered by Industrial Internet of Things (IIoT) technologies, enable real-time monitoring and intelligent energy management, promoting carbon reduction and sustainability. However, integrating distributed energy resources like solar, wind, and prosumers introduces challenges such as mismatches in production and consumption, grid imbalances, and energy losses. Centralized, static supply-demand balancing approaches struggle to address these complexities in carbon-aware intelligent IIoT. This paper presents a Cooperative Distributed Multi-Agent System (CD-MAS) for Smart Grid IIoT, fostering decentralized, real-time collaboration among producers, distributors, consumers, and prosumers. The system leverages self-optimizing agents to enhance carbon efficiency and grid resilience. Key features include consumer profile analysis to create adaptive Virtual Neighborhoods (VNs) for targeted energy programs and distributed energy prediction for accurate demand forecasting and demand-supply balancing with localized load-shedding strategies. The proposed CD-MAS advances smart grid capabilities, optimizing energy use and achieving net-zero carbon goals.

Artificial intelligence of vehicles (AIVs) is poised to revolutionize transportation by promoting low-carbon alternatives, such as electric vehicles (EVs). However, the deployment of fixed charging stations (FCSs) lags behind the growing demand, particularly in rural areas, causing range anxiety among potential EV owners. This article proposes a smart transportation solution within the Artificial Intelligence of Things (AIoT) framework to establish a sustainable, low-carbon system. AIoT systems enable real-time data acquisition and analysis through extensive embedded IoT EV sensors and communication networks for pattern recognition and decision making on the cloud. The proposed solution integrates sensor information from vehicle-to-vehicle (V2V) charging, smart home charging stations (HCSs), and mobile charging services (MCSs), coordinated by the cloud-fog nodes in geographically distributed zones. This article employs the Hungarian matching algorithm for optimal decision making of matching EVs with charging services. Our approach incorporates AIV and AIoT technologies to enhance decision making by using an ensemble-based machine learning (ML) model for precise EV range estimation. The comprehensive details and specifications of these proposed models are elaborated in this article.

Non-Intrusive Load Monitoring (NILM) is a technique used for identifying individual appliances' energy consumption from a household's total power usage. This study examines a novel energy disaggregation model called Variational Auto-Encoder (VAE) with Federated Learning (FL). Specifically, VAE has a complex structure that resolves the issues in Short Sequence-to-Point (Short S2P) with fewer samples as input windows for each appliance. Short S2P cannot be generalized and might confront some challenges while disaggregating multi-state appliances. To this end, we examine a series of experiments using a real-life dataset of appliance-level power from the UK: UK-DALE. We also investigate additional protection of model parameters using Differential Privacy (DP). The findings show that FL with the VAE model achieves comparable performance to its centralized counterpart and improves all the metrics significantly compared to the Short S2P model.

Electric vehicles (EVs) are the backbone of the future intelligent transportation system (ITS). They are environmentally friendly and can also be integrated as distributed energy resources (DERs) into the smart grid using vehicle-to-grid (V2G) scheme. Specifically, utility companies can push back EV batteries into the electric grid to reduce the peak load. However, integrating EVs into the power grid efficiently requires accurate artificial intelligence (AI) mechanisms to forecast, coordinate, and dispatch the EVs into the grid. This paper proposes a Multi-agent Reinforcement Learning (MARL) mechanism that schedules the day-ahead discharging process of EV batteries to optimize the peak shaving performance of the electric grid. The proposed MARL overcomes the inaccuracy of energy prediction by allowing the agents, i.e. EVs, to make autonomous decisions. These agents are trained in a centralized fashion but make decisions locally to maintain autonomy and privacy. In particular, the model does not require that the EVs communicate with a centralized entity during the execution stage, which assures the model’s integrity and protects the EVs’ private information. To evaluate the model, a comprehensive series of experiments were carried out to prove the effectiveness of the MARL coordination and scheduling mechanism and to show that the model can indeed flatten the peak load.

The future Intelligent Transportation System (ITS) will rely heavily on the advancement of the Internet of Things (IoT). Indeed, the IoT infrastructure paves the way toward connecting vehicles for various benefits including traffic monitoring, crowdsourcing, energy trading, and other ITS services. One of the use cases of the IoV is Vehicle-to-Vehicle (V2V) energy charging/discharging, which is expected to be an integral part of the ITS. In V2V paradigm, Electric Vehicles (EVs) with bidirectional chargers can communicate with a grid edge network or directly with another EV using low-power wide-area networks (LPWAN) or 5G wireless connection to offer or demand energy. V2V energy exchange allows EV owners to make money from selling their battery’s excess energy to other EVs. One of the main challenges for the wide adoption of V2V is the development of mechanisms that maximize the benefits for participants. In the V2V paradigm, energy providers and suppliers require mechanisms that ensure maximal matching for the optimal social welfare of the users. In this paper, we propose a double-sided auction mechanism that matches EVs by pairing bids and asks such that the traded volume and the utilities are maximized. Through theoretical analysis, we show that the proposed model can indeed be truthful, individually rationale, and computationally efficient. Finally, we evaluate the proposed model based on real data and provide performance analysis.

Offline reinforcement learning promises to enable control policy learning directly from previously collected data. It is a challenging problem because data collected from previous policies may not well-represent the distribution of the data when following the learned policy. This work evaluates the performance of recent offline RL algorithms using the library D3RLpy, which includes recent state-of-the-art algorithms as well as relevant datasets to evaluate them. We evaluate the performance of algorithms provided by the library without hyperparameter tuning. We compare performance both in terms of algorithm performance on the environments, as well as considering the effect of the dataset on the performance a few select algorithms.

In this work, we propose a reliable hyperparameter tuning scheme for offline reinforcement learning. We demonstrate our proposed scheme using the simplest antmaze environment from the standard benchmark offline dataset, D4RL. The usual approach for policy evaluation in offline reinforcement learning involves online evaluation, i.e., cherry-picking best performance on the test environment. To mitigate this cherry-picking, we propose an ad-hoc online evaluation metric, which we name "median-median-return". This metric enables more reliable reporting of results because it represents the expected performance of the learned policy by taking the median online evaluation performance across both epochs and training runs. To demonstrate our scheme, we employ the recently state-of-the-art algorithm, IQL, and perform a thorough hyperparameter search based on our proposed metric. The tuned architectures enjoy notably stronger cherry-picked performance, and the best models are able to surpass the reported state-of-the-art performance on average.

Residential households contribute significantly to the overall energy consumption in developed countries. To reduce their energy consumption, they need solutions that help them track the use of their appliances at home. Non-Intrusive Load Monitoring (NILM) with short Sequence-to-Point (Seq2Point) is a deep learning method used to track and recognize what appliances are used in the houses and their respective energy consumption. To apply NILM with short Seq2Point in the real world, a large amount of data from households must be collected and transferred to centralized servers for further analysis. Such a process does not always preserve consumers' security and privacy. To address this challenge, this paper proposes a combination of Federated Learning (FL) and NILM to make the entire system safe and trustworthy in order to avoid the risks of customers' privacy leakage. In FL, local models are developed on each consumer end and trained on local data instead of gathering data from all devices and sending it back to the central server. By using this method, data will be handled locally in a single residence while increasing the system's speed and security. The proposed model is evaluated using a real-life dataset (UK-Dale) of Appliance Level Electricity from the UK. The results show that our system provides better accuracy and preserves the privacy of consumers.

Industrial Internet of Things (IIoT) devices generate and collect massive amounts of industrial data. Monetizing the flood of data generated by the IIoT devices has enabled the creation of the IIoT data ecosystem, where individuals and businesses may trade data. With the rapid expansion of the online data trading industry, the necessity for an edge intelligent reputation system is becoming increasingly important as more individuals and services connect online. In recent years, researchers have proposed blockchain-based reputation systems as a means of offering anonymity, security, transparency, and mutual trust for both providers and customers in Industry 4.0. Unfortunately, they focus on the decentralized reputation system with a single certificate authority, which creates the concern of a single point of failure (SPOF). Moreover, researchers paid little attention to the performance measures of these blockchain-based reputation systems to demonstrate their usability in a real IIoT data ecosystem. This article proposes a robust edge intelligent blockchain-based reputation system capable of avoiding failures by enhancing the Raft consensus mechanism. We provide extensive security analysis and simulation experiments to demonstrate the performance of the blockchain-based reputation system for the IIoT data ecosystem using different metrics, such as transaction throughput, latency, and resource consumption.

The fast-growing electric vehicles (EVs) industry requires a well-designed recommendation system to locate charging stations while ensuring private data protection. This article proposes a secure cloudlet-based recommendation system for EVs. Unlike conventional methods where training a recommender model involves direct data sharing between data holders, our model utilizes a secure vertical federated learning technique, in which EVs data do not leave the platforms. To improve the efficiency of the model and to alleviate the communication-related concerns in our recommender model, cloudlet-based data aggregator(s) are used as a replacement for the existing centralized architectures. To enhance the security of our system, blockchain technology is incorporated to generate a trusted network of cloudlets that are responsible for transmitting the locally computed training parameters. The simulation results achieved from our proposed recommendation system show that the distribution of EVs over a designated area with charging stations is more optimal, and the proposed decentralized recommender with 10 cloudlets is 5.2 s quicker than a conventional centralized model.

Federated Edge Learning (FEL) allows edge nodes to train a global deep learning model collaboratively for edge computing in the Industrial Internet of Things (IIoT), which significantly promotes the development of Industrial 4.0. However, FEL faces two critical challenges: communication overhead and data privacy. FEL suffers from expensive communication overhead when training large-scale multi-node models. Furthermore, due to the vulnerability of FEL to gradient leakage and label-flipping attacks, the training process of the global model is easily compromised by adversaries. To address these challenges, we propose a communication-efficient and privacy-enhanced asynchronous FEL framework for edge computing in IIoT. First, we introduce an asynchronous model update scheme to reduce the computation time that edge nodes wait for global model aggregation. Second, we propose an asynchronous local differential privacy mechanism, which improves communication efficiency and mitigates gradient leakage attacks by adding well-designed noise to the gradients of edge nodes. Third, we design a cloud-side malicious node detection mechanism to detect malicious nodes by testing the local model quality. Such a mechanism can avoid malicious nodes participating in training to mitigate label-flipping attacks. Extensive experimental studies on two real-world datasets demonstrate that the proposed framework can not only improve communication efficiency but also mitigate malicious attacks while its accuracy is comparable to traditional FEL frameworks.

This paper proposes an image-based peak load prediction system for demand response in smart grids. Our aim is to enhance the prediction outcome especially during the season-changing days, also called shoulder-season. These days do not have a specific time throughout the year so we cannot use the season as a variable in the training set. We hypothesize that the approximate curve of the daily power consumption graph has some specific patterns that can be used to separate each day into different groups based on the pattern of the energy consumption curve. To this end, we use a convolution neural network model to classify and extract the features of the curve image. Then, we apply the k-means mechanism for image clustering to select better training sets and optimize the forecasting mechanism. Our results show an overall improvement of prediction during the season-changing period.

On-demand resource management is a challenging prospect, especially in communication networks where the users' requirements of network resources are volatile. This challenge is magnified by the current shift of hardware-dependent network technologies into virtual Network Functions (vNFs) in a cloud-based platform and the proliferation of the Internet of Things (IoT) networks. vNFs are an integral part of edge devices that aims to improve service providers' response time, eliminate redundancy, and end to end latency results in operators' lower operating costs significantly. The efficiency of such a virtualized system largely depends on the optimization of resource allocation between users and Virtual Machines (VM) or hosting devices. One approach to addressed this issue is to minimize the latency of the entire communication network, however, the concern is that this mechanism does not ensure fairness of resource allocation because some of the users may receive enhanced services than others. In the previous study, a Stable Matching Algorithm is proposed to minimize the latency between vNFs and VMs but the algorithm failed to ensure the fairness of the requested services and server elements. In this study, we further extend the previous model to ensure the fairness. Our proposed solution minimizes the maximum latency in the network. This problem is an NP-hard and hence we provide a local search based solution. The experimental result shows improved fairness from other approaches and the fairness index is close to the optimal, which in fact represents the elimination of imbalances to a great extent.

This paper studies incentive-based peer-to-peer (P2P) energy trading between sellers and buyers in a smart grid distributed system. When designing an energy trading model the main goal is to maximize the social welfare of the involved parties. Peer-to-peer energy trading is a novel mechanism of power system operation which allows users to generate and trade renewable energy. Buyers are considered to be consumers and sellers are considered to be prosumers (producers and consumers of energy) with respect to time. Buyers are more motivated to purchase energy from the seller since the energy was generated from a green source and is typically cheaper than the main grid. Furthermore, by producing clean energy and distributing within small networks, the transmission line stress is mitigated since the overall demand required from the main grid is reduced. In this paper, we model the interaction as single-sided auction taking into consideration the grid infrastructure constraints (e.g capacity) and cost while maximizing the profit of the players. We assess through theoretical analysis and simulations the bidding properties including individually rational, truthful, computationally efficient, and fairness.

Multimedia content is massively generated from various applications and devices, and processed in cloud data centers. Multimedia service providers prefer that their data are processed in data centers close to users in order to offer them high performance and reliable multimedia services that meet the requirements specified in the Service Level of Agreement (SLA). This requires transferring huge data sets of video streams, games content, images etc. across geographically distributed cloud data centers using underutilized bandwidth in backbone transport networks. As the amount of multimedia content increases, the demand to transfer big data sets across data centers increases as well. As such, the leftover bandwidth that appears at different times and for different durations in the backbone network becomes insufficient to satisfy the rapidly increasing demand for multimedia big data transfer. This challenge led to the creation of multi-rate Bandwidth on-Demand (BoD) service offerings for communication between geographically distributed cloud data centers. In this paper, we focus on BoD services which are offered by the Dense Wavelength Division Multiplexing (DWDM) layer because of its huge capacity. We propose a BoD broker which employs a scheduling algorithm that considers various deadlines of multimedia big data transfer requests. The broker in our model leverages the concept of standby wavelengths to minimize peak traffic and accommodate time requirements of delay-tolerant and delay-intolerant transfer requests. We also study strategies of routing and wavelength assignment using Mixed Integer Programming (MIP) optimization to rapidly handle volumes of multimedia big data transfer requests.

This article proposes a tree-based deep model for effective load distribution to edge devices without much loss of accuracy. The input image is divided into groups of volumes, and each volume is passed through a tree structure. The tree structure has many branches and levels, each of which is represented by a convolutional layer. The layers are independent of each other. Therefore, various edge devices can update the parameters of the layers in parallel independently. Experiments are performed using a benchmark dataset and a publicly available date fruits database. Experimental results show that the proposed model has a high information density by reducing the number of parameters without much loss of accuracy.

Lately, HTTP Adaptive Streaming (HAS) has become dominant among different video streaming techniques. However, owing to the unpredictable nature of the wireless radio channel and device mobility, using HAS over mobile wireless networks is still very challenging. In this paper, we propose a novel Quality of Experience (QoE) optimization mechanism for HAS in the context of mobile wireless networks. The proposed mechanism leverages recent advances in HAS specification, which includes new features for QoE measurements and reporting. First, we formulate a discrete optimization problem aiming at maximizing the overall average quality, and minimizing the negative impact of temporal video quality changes for all HAS users simultaneously. Second, in order to take advantage of well-known continuous optimization techniques and to decrease the computational complexity, we convert the formulated problem into a continuous form, and propose a gradient based algorithm to solve the continuous optimization problem. The results of our simulations demonstrate that our system attained better perceived video quality by almost 8% on average, while lowering the freezing period by 20% on average across HAS users when compared to other approaches where HAS users only rely on local adaptation logics.

The Hypertext Transfer Protocol (HTTP) Adaptive Streaming (HAS) has now become ubiquitous and accounts for a large amount of video delivery over the Internet. But since the Internet is prone to bandwidth variations, HAS's up and down switching between different video bitrates to keep up with bandwidth variations leads to a reduction in Quality of Experience (QoE). In this article, we propose a video bitrate adaptation and prediction mechanism based on Fuzzy logic for HAS players, which takes into consideration the estimate of available network bandwidth as well as the predicted buffer occupancy level in order to proactively and intelligently respond to current conditions. This leads to two contributions: First, it allows HAS players to take appropriate actions, sooner than existing methods, to prevent playback interruptions caused by buffer underrun, reducing the ON-OFF traffic phenomena associated with current approaches and increasing the QoE. Second, it facilitates fair sharing of bandwidth among competing players at the bottleneck link. We present the implementation of our proposed mechanism and provide both empirical/QoE analysis and performance comparison with existing work. Our results show that, compared to existing systems, our system has (1) better fairness among multiple competing players by almost 50% on average and as much as 80% as indicated by Jain's fairness index and (2) better perceived quality of video by almost 8% on average and as much as 17%, according to the estimate the Mean Opinion Score (eMOS) model.

In this paper, we propose a novel QoE optimization mechanism for Dynamic Adaptive Streaming over HTTP (DASH) in the context of wireless mobile networks. The proposed mechanism leverages recent advances in 3GPP DASH specification, which includes new features for QoE measurements and reporting. The proposed optimization mechanism has two objective functions, the first function maximizes the overall average QoE among DASH clients while the second function minimizes the negative impact of temporal video quality changes, i.e. the up and down switching between different representation during playback. The results of our simulations demonstrate that the proposed method improves the overall QoE and outperform other approaches where DASH clients only rely on local adaptation logic.

As dynamic delivery of video over HTTP becomes prominent, rate adaptation techniques become more challenging due to bandwidth variations. This paper presents a Fuzzy-based controller to dynamically adapt the video bitrate based on both the estimated throughput and the size of the playback buffer. The proposed Fuzzy Logic Controller (FLC) mechanism takes the observed throughput and buffer dynamics as inputs to change the policy of selecting the video bitrate and download scheduling to minimize the negative effect of ON-OFF switching. The experimental results show that our proposed mechanism generates a smoother stream compared to existing methods.

Cloud computing provides reliable, affordable, and flexible resources for many applications and users with constrained computing resources and capabilities. The cloud computing concept is becoming an appealing paradigm for many industries including the gaming industry, leading to the introduction of cloud gaming architectures. Despite its advantages, cloud gaming suffers from unguaranteed end-to-end delay as well as server side's computational complexity. In this paper, a novel algorithm for reducing the computational complexity and hence speeding up the video encoding speed is proposed. Specifically, by performing minimum modifications in the game engine and the video codec, some information from the game engine is fed into the video encoder to bypass the motion estimation (ME) process. Our results show that the proposed method achieves up to 39% speedup in the ME process, leading to a 24% acceleration in the total encoding process.

Throughout various complex processes within hospitals, context-aware services and applications can help to improve the quality of care and reduce costs. For example, sensors and radio frequency identification (RFID) technologies for e-health have been deployed to improve the flow of material, equipment, personal, and patient. Bed tracking, patient monitoring, real-time logistic analysis, and critical equipment tracking are famous applications of real-time location systems (RTLS) in hospitals. In fact, existing case studies show that RTLS can improve service quality and safety, and optimize emergency management and time critical processes. In this paper, we propose a robust system for position and orientation determination of equipment. Our system utilizes passive (RFID) technology mounted on flooring plates and several peripherals for sensor data interpretation. The system is implemented and tested through extensive experiments. The results show that our system's average positioning and orientation measurement outperforms existing systems in terms of accuracy. The details of the system as well as the experimental results are presented in this paper.

Intelligent Environments have the vision of enhancing our everyday environment and interaction with its objects by sensing, computing, and communication capabilities. The major characteristics of such environments are the increasing number of embedded intelligent devices (ubiquity) into the background (transparency). These devices are expected to disappear or blend into the background and will be invisible to the user. However, because of this transparency, users fail to develop an adequate mental model for interaction with such environments. The Ubiquitous Meta User Interfaces (Ubi-MUI) ACM workshop provides a venue for the development of highly intuitive, multimedia supported meta user interfaces that bring transparency, predictability, and control into intelligent environments.

With the rapid development of applications in open distributed environments such as eCommerce, privacy of information is becoming a critical issue. Information about the preferences, activities, and demographic attributes of people using online shopping is very valuable to online businesses. Beyond the general anxieties with sharing personal information, people may more specifically have concerns about becoming increasingly identifiable; as increasing amounts of personal data are acquired. Not only that, but also it is widely known that information about consumers is often sold to online marketing and advertising companies, generally without the knowledge of consumers [4, 16]. In this paper, we introduce a model where people can opt out to share personal information for a payoff value that balances out the costs of privacy. Our model is based on agents working on behalf of consumers to maximize their benefit. The analysis of the model and a proof of concept implementation are presented in this paper.

The market for consumer information is already a lively market, where consumer information and consumer profile data are often among the most valuable assets owned by online retailers. The value of such commodity derives from the ability of firms to identify consumers and charge them personalized prices flj. We argue that if consumers' identity and personal information is such a valuable asset, should not consumers benefit from their asset as well? In this paper, we propose a negotiation process between an online consumer agent and an online seller. The online consumer agent acts on behalf of consumers to maximize their social welfare. In our model, the agent derives a quantified privacy risk for each private data and uses it to determine a cost premium value to make the bargaining process manageable. We also provide a computational example to evaluate the model.