Search

With the rapid development of cloud computing, Internet of things and artificial intelligence, human–computer interaction (HCI) is playing an increasingly important role in the daily life. As an important component of HCI, hand gesture recognition (HGR) system is usually combined with edge computing server, utilizing machine learning, including neural network, decision tree, integrated learning, to achieve low latency and high reliability service. High precision HGR with low computational complexity is prerequisite for the commercialization of gesture recognition. Therefore, this paper proposed a high-precision parameter correction algorithm based on the established scattered-point model and the outlier detection scheme, and a recognition algorithm with multivariable decision tree is then presented for the dynamic hand gestures. The experimental results show that the proposed algorithms can improve the recognition accuracy and effectively reduce the running time, which is conducive to algorithm transplantation and model deployment in edge servers. [ABSTRACT FROM AUTHOR]

The availability of Software Defined Network's (SDNs) flow rule entry in the flow table is considered a key factor in the timely delivery of a certain flow. The controller is approached for instructions on how to deal with the flow when the rule for such flow is missing. The controller then updates the flow table accordingly at the switch so that flow could be dealt with locally. It becomes problematic when no rule is defined yet at the controller by the application plane. In most of these cases, such a situation is handled by programming the controller with wildcard rules. However, handling many flows at once under wildcard rules severely hampers the network performance. Flow rules formation by the application plane is sometimes critical and time-consuming which increases the latency ratio by creating a bottleneck at the switch level. To avoid the bottlenecks due to rule absence, in this paper, rather than waiting for the application plane's response and putting the pending traffic flows in the buffer, which may be dropped, the controller is programmed in a way that has the built-in mechanism of self-flow rule formation. This atomized mechanism is based on the previously available traces of the same flows when they were forwarded on the network using the wildcard rules. To assess the performance of the proposed work, it is emulated and benchmarked with the latest research. The results show considerable performance achievement. [ABSTRACT FROM AUTHOR]

The proliferation of wireless applications at an exponential rate has made spectrum problems worse. Saturation in the unlicensed frequency spectrum is rapidly increasing as a result of the increasing data rates required by new wireless devices. A proposed solution to this problem is cognitive radio, which allows for the opportunistic use of licenced spectrum in less crowded areas. Cognitive network-based security evaluations using mobile edge computing and a Beyond 5G' (B5G) machine learning (ML) model are the focus of this research. In this case, the security study was carried out using cognitive network data transfer and multi-agent reinforcement encoder neural network and mobile edge computing (MRENN-MEC), a multi-agent reinforcement encoder neural network with mobile edge computing. Scalability, quality of service, throughput, and forecast accuracy are some of the network properties that undergo experimental analysis. [ABSTRACT FROM AUTHOR]

The advancement of technology towards the realization of the evolving mobile computing paradigm brings a rapid paradigm shift in its usage, especially in the Internet, computation, and communications, that has a profound impact on businesses, services, and users. With the rise in resource-intensive or edge-based mobile applications such as autonomous driving, Amazon Go, virtual and augmented reality, and healthcare-related applications, countless challenges in computation and communication parameters like latency, bandwidth, and energy consumption are evolving. As a result, the multi-access edge computing (MEC) paradigm receives enormous attention where some portions of the user applications are offloaded to powerful machines for their efficient execution to optimize different evaluation metrics or to achieve performance goals. While a few survey works are available in this direction, none of them focuses explicitly on the emerging machine learning (ML) based computation offloading techniques and various associated sub-problems together. This paper aims to provide a detailed but precise overview of the research on using ML techniques for MEC environments. In this survey, we focus on how authors and researchers utilize the ML models in computation offloading problems on MEC architecture. We extend our study by considering several edge architectures, offloading parameters, ML approaches, and problem formulation strategies concerning computation offloading. Additionally, this paper discusses the potential challenges in the direction of computation offloading on MEC architecture. [ABSTRACT FROM AUTHOR]

Vehicular Edge Computing (VEC) is a promising paradigm for providing low-latency and high-reliability services in the Internet of Vehicles (IoV). The increasing number of mobile devices and the diverse resource requirements of the growing IoV have resulted in a shift from centralized resource management to a decentralized approach. This shift offers improved fault tolerance, scalability, and privacy preservation. However, constructing collaborative awareness and coordination mechanisms between multiple vehicles and edge nodes in a decentralized manner is a challenge. To address this issue, we propose a decentralized many-to-many task offloading method that aims to minimize the average task completion latency of vehicles. In this study, we propose a data-sharing mechanism between vehicles and edge servers using the digital twin service, which provides global environmental perceptions to the vehicles by a low-cost approach. Additionally, we develop a mean-field multi-agent reinforcement learning algorithm to generate coordinated task offloading schemes. Instead of interacting with multiple agents, the vehicle only needs to respond to the mean action of the environment. Based on this transition, the agent generates coordinated task offloading decisions in complex scenarios. We evaluate the performance of our method using real urban traffic data. Experiment results verify the efficiency of our proposed method. [ABSTRACT FROM AUTHOR]

With the rapid advancement of communication technology, unmanned aerial vehicles (UAVs) have gained significant attention. The incorporation of UAVs in mobile edge computing (MEC) systems has proven to be a viable method for improving system throughput and decreasing computing latency. In pursuit of enhancing the quality of service (QoS) for users, we propose a UAV-assisted MEC system. Our approach entails the utilization of a multi-agent graph convolutional deep reinforcement learning algorithm that facilitates the transmission of distributed information and computing tasks from multiple ground users to the UAV. Through individual strategy training, users are able to effectively learn collaborative strategies. Simulation results validate the efficacy of the proposed UAV-assisted MEC scheduling method, which is based on graph neural networks (GNN), in significantly enhancing system throughput. [ABSTRACT FROM AUTHOR]

The Internet of Vehicles (IoV) has witnessed a substantial surge in its expansion during recent years, with the integration of edge servers emerging as an increasingly prevalent practice to cater to content requests that demand a superior quality of service. Nonetheless, conventional cache policies often prove inadequate for IoV applications, primarily due to their incapacity to effectively handle a diverse array of content requests, the high-speed mobility of vehicles, and the inherent instability of network connections. Within the scope of this study, we commence by bifurcating various content requests into two distinct categories: latency-sensitive content request and bandwidth-sensitive content request. Subsequently, we establish a model to evaluate the service quality under the joint consideration of different types of content requests, the mobility of vehicles and storage capacity of edge nodes. Furthermore, we transform the quality-of-service (QoS) penalty function into a system reward function. This adaption enables us to propose an innovative edge cache scheme founded upon the Deep Deterministic Policy Gradient (DDPG) algorithm of Reinforcement Learning (RL) method, which empowers dynamic adjustments in response to the evolving IoV environment. To validate the effectiveness of our proposed approach, we harness the Simulation of Urban Mobility (SUMO) traffic simulation software and construct a traffic road scenario based on a specific part segment of Nanjing Beltway. A wide-ranging set of contrast experiments was performed to ensure the improved performance of the DDPG based deep reinforcement learning method. Simulation experimental results show that the proposed algorithm converges quickly and outperforms existing algorithms in terms of service quality-hit ratio for latency-sensitive content and transmission speed for bandwidth-sensitive content. [ABSTRACT FROM AUTHOR]

Asynchronous Federated Learning (AFL) has been introduced to improve the efficiency of FL by reducing the latency of Machine Learning (ML) model aggregation, particularly in the Internet of Things (IoT) environment. Meanwhile, decentralized FL, such as leveraging blockchain and directed acyclic graph (DAG)-based ledgers, also has drawn much attention to the integration of FL owing to the security benefit against single-point-failure and Byzantine fault tolerant consensus. We observe that the inherent network asynchrony of DAG-based ledgers is beneficial for implementing asynchronous FL particularly in edge computing domains, even though there is no existing survey work that provides a fundamental overview of such integration. This paper surveys on the integration of asynchronous FL with DAG, which we call AFL-DAG, as a promising approach to realize the intersection of decentralized FL and asynchronous FL. Motivated by the lack of a concrete taxonomy of asynchronous FL and a global model, especially with decentralized FL, we introduce universally applicable terminologies and the extensive classification method of FL in terms of asynchrony. Based on the proposed taxonomy, we present a generic system model of AFL-DAG for edge computing applications. To provide a horizontal overview and cover fundamental concepts in AFL-DAG, we identify four critical design factors and the state-of-the-art solutions are discussed accordingly. Future directions to achieve a practical AFL-DAG are also highlighted. Finally, we explore the opportunities of AFL-DAG by investigating a popular edge computing application, the on-device crowdsourcing, and provide a high-level evaluation of AFL-DAG compared to blockchained-FL. [ABSTRACT FROM AUTHOR]

Multi-access Edge Computing (MEC) has recently emerged as a potential technology to serve the needs of mobile devices (MDs) in 5G and 6G cellular networks. By offloading tasks to high-performance servers installed at the edge of the wireless networks, resource-limited MDs can cope with the proliferation of the recent computationally-intensive applications. In this paper, we study the computation offloading problem in a massive multiple-input multiple-output (MIMO)-based MEC system where the base stations are equipped with a large number of antennas. Our objective is to minimize the power consumption and offloading delay at the MDs under the stochastic system environment. To this end, we introduce new formulation of the problem as a Markov Decision Process (MDP) and propose two Deep Reinforcement Learning (DRL) algorithms to learn the optimal offloading policy without any prior knowledge of the environment dynamics. First, a Deep Q-Network (DQN)-based algorithm to solve the curse of the state space explosion is defined. Then, a more general Proximal Policy Optimization (PPO)-based algorithm to solve the problem of discrete action space is introduced. Simulation results show that our DRL-based solutions outperform the state-of-the-art algorithms. Moreover, our PPO algorithm exhibits stable performance and efficient offloading results compared to the benchmarks DQN and Double DQN (DDQN) strategies. [ABSTRACT FROM AUTHOR]

With the continuous development of mobile edge computing, people are more willing to offload tasks to edge servers that are closer to users than cloud services for a better user experience. Due to problems such as user mobility and limited coverage of edge servers, how to ensure service quality and prevent service interruption is a problem we have to consider. This article explores the problem of service migration to decide when, where, and how to migrate an ongoing service from the current edge server to the target edge server. At the same time, we will consider the premise that edge servers can only obtain partial user information, and model the service migration problem as a partially observable Markov decision process. To minimize user delay and system energy consumption, we propose a service migration decision algorithm based on Deep recurrent Q-learning (DRQNSM). Multiple experiments show that our algorithm has a better performance than some classical reinforcement learning algorithms. [ABSTRACT FROM AUTHOR]

Approximately one-third of human food is derived from flowering plants, and 80% of those plants require honeybees for pollination. However, climate change is affecting bees as well as the beekeeping industry. Having continuous information on the condition of honey bee colonies would be helpful in studying new diseases such as colony collapse disorder, as well as in developing novel beekeeping tools to improve the health management of bee colonies. In this study, an audio monitoring system is established based on Internet of Thing (IoT) technology to record the sounds produced by bees. A colony status classification model coupled with a machine learning algorithm is used to analyze the sound data of bee colonies under various treatments to classify the colony conditions, such as queenless, virus-infected, and pesticide-contaminated. With an edge computing framework, the classification is done by the monitoring system, which greatly reduces data transmission delay and cloud computing burden. The field test results show that the accuracy and F 1 -score for the classification model for the beehive under various treatments reach 98.8% and 0.98, respectively. The proposed IoT-based audio monitoring system for beehives can assist beekeepers and users in managing bee colonies more effectively by providing automatic, accurate, and real-time monitoring information regarding the current colony conditions. [ABSTRACT FROM AUTHOR]

This paper explores the synergies between collective intelligence and edge computing, presenting a novel paradigm that harnesses decentralized processing power for collaborative problem-solving. Edge devices, such as sensors and IoT devices, collect spatial data in real time from their local environments. They can incorporate machine learning algorithms to analyze spatial data, enabling quicker and more context-aware decision-making. Spatial clustering, a pivotal strategy in IoT edge computing, is examined to optimize localized data processing, enhance resource efficiency, and enable real-time analytics in decentralized environments. By leveraging the physical proximity of devices, spatial clustering contributes to the effectiveness and sustainability of IoT deployments at the edge. The paper introduces an innovative approach to adaptive spatial clustering by adopting swarm intelligence, drawing inspiration from the collective behavior of a flock of birds. Building upon the classical flock model of Reynolds, our extended model incorporates movement in a multi-dimensional space and introduces different types of birds. In this context, the birds serve as agents for discovering points with specific characteristics in a multidimensional space. The integration of swarm intelligence into spatial clustering presents a promising avenue for addressing the challenges of decentralized processing in edge computing environments, paving the way for more efficient and responsive IoT deployments. [ABSTRACT FROM AUTHOR]

The evolution of edge computing devices has enabled machine intelligence techniques to process data close to its producers (the sensors) and end-users. Although edge devices are usually resource-constrained, the distribution of processing services among several nodes enables a processing capacity similar to cloud environments. However, the edge computing environment is highly dynamic, impacting the availability of nodes in the distributed system. In addition, the processing workload for each node can change constantly. Thus, the scaling of processing services needs to be rapidly adjusted, avoiding bottlenecks or wasted resources while meeting the applications' QoS requirements. This paper presents an auto-scaling subsystem for container-based processing services using online machine learning. The auto-scaling follows the MAPE-K control loop to dynamically adjust the number of containers in response to workload changes. We designed the approach for scenarios where the number of processing requests is unknown beforehand. We developed a hybrid auto-scaling mechanism that behaves reactively while a prediction online machine learning model is continuously trained. When the prediction model reaches a desirable performance, the auto-scaling acts proactively, using predictions to anticipate scaling actions. An experimental evaluation has demonstrated the feasibility of the architecture. Our solution achieved fewer service level agreement (SLA) violations and scaling operations to meet demand than purely reactive and no scaling approaches using an actual application workload. Also, our solution wasted fewer resources compared to the other techniques. • We specify a software subsystem that incorporates an auto-scaling approach. • We designed an online machine learning-based auto-scaling. • The auto-scaling is incorporated into a distributed video analytics system. • The auto-scaling changes its behavior from reactive to proactive and vice versa. • The auto-scaling follows the MAPE-K autonomic control loop. • We present a performance evaluation using a video analytics scenario. [ABSTRACT FROM AUTHOR]

This article investigates 14-MeV neutron radiation soft-error effects on two contrasting hardware architectures for edge computing systems, one based on an 8-bit processor and another on a 64-bit processor, both processing three frequently used types of tiny machine learning (ML) algorithms. Experimental results of 14-MeV neutron radiation tests in the different case-study ML algorithms running on the 64-bit processor show that the artificial neural network version using floating point failed more than the other versions, although most failure situations caused no misclassification in the ML inference. Furthermore, considering tiny ML applications demanding the investigated 8-bit and 64-bit computing systems for the same amount of ML inferences, the overall results suggest a preliminary trend that both case-study systems would have similar tolerance to 14-MeV neutron radiation soft errors. • Case-study Artificial Neural Networks had the most failures yet only one was critical. • Integer versions tend to be more reliable than floating-point versions. • Case-study Random Forests suffered no misclassifications. • The 8-bit processor studied would fail once every 175 million hours approximately. • From the point of view of number of inferences both processors show similar results. [ABSTRACT FROM AUTHOR]

The large increase in the number of Internet of Things (IoT) devices have revolutionised the way data is processed, which added to the current trend from cloud to edge computing has resulted in the need for efficient and reliable data processing near the data sources using energy-efficient devices. Two methods based on low-cost edge-IoT architectures are proposed to implement lightweight Machine Learning (ML) models that estimate indoor environmental quality (IEQ) parameters, such as Artificial Neural Networks of Multilayer Perceptron type. Their implementation is based on centralised and distributed parallel IoT architectures, connected via wireless, which share commercial off-the-self modules for data acquisition and sensing, such as sensors for temperature, humidity, illuminance, CO 2 , and other gases. The centralised method uses a Graphics Processing Unit and the Message Queuing Telemetry Transport protocol, but the distributed method utilises low-performance ARM-based devices and the Message Passing Interface protocol. Although multiple IEQ parameters are measured, the training and testing of ML models is accomplished with experiments focused on small temperature and illuminance datasets to reduce data processing load, obtained from sudden spikes, square profiles and sawteeth test cases. The results show a high estimation performance with F-score and Accuracy values close to 0.95, and an almost theorical Speedup with a reduction in power consumption close to 37% in the distributed parallel approach. In addition, similar or slightly better performance is achieved compared to equivalent IoT architectures from related research, but error reduction of 35–76% is accomplished with an adequate balance between performance and energy efficiency. • Computing methods based on Artificial Neural Networks of Multilayer Perceptron type (ANN-MLP). • Lightweight Machine Learning (ML) models that estimate IEQ parameters, such as temperature, humidity, illuminance, and CO 2. • Implementation of centralised and distributed parallel IoT architectures using COTS devices connected via a wireless LAN. • The centralised method uses a GPU and the MQTT protocol, and the parallel method uses ARM-based devices and the MPI protocol. • Experimental validation focused on temperature tests with high estimation performance (F-score, Accuracy) and energy efficiency. [ABSTRACT FROM AUTHOR]

Smart Grid technology has been considered an attractive research issue due to its efficiency in solving energy demand, storage, and power transmission. The integration of IoT technology in the Smart Grids is a critical way to accelerate the digitization of the power grid and is useful for the efficient performance of the energy grid infrastructure. For efficient real-time data analysis and decision-making, the Internet of Things will incorporate various communication systems seamlessly. To achieve efficient communication between all Internet of Things, devices are expected to use multiple means, including smart sensors, cable and wireless communication. Improved Internet of Things sensor technologies and connectivity could theoretically prevent or minimize the potential to natural disaster transmission lines, improve transmission power capacity and reduce economic losses. A smart grid is a variety of sensors, devices, and data sets that continuously capture high-resolution data equal to individual IoT conditions. A vast amount of data is one of the biggest challenges on the Internet of Things. Edge Computing is trying to process data close to linked sensors to address this problem, where the data is gathered and processed. This paper aims to investigate the edge computing solutions for the smart grid. A comprehensive review of both emerging issues and edge computing in the Smart Grid environment is discussed and explained. There are two primary components to the energy sharing process among Prosumers: information/digital technologies and Artificial Intelligence Scheduling Techniques. Each of them is mentioned in detail to discuss the Prosumer smart Grid. Furthermore, Edge Computing and classifications (cloudlet, Fog computing and Multi-Access) are among the suitable network methods mentioned in this paper. Some techniques and methodologies have been extensively covered to improve reader awareness of the Prosumer smart grid system. [ABSTRACT FROM AUTHOR]

Smart Grid (SG) is a relatively new subject in which intelligence comes from computer systems. Edge Computing (EC) processes data near the SG information source — potentially attenuating latency problems. While Internet of Things (IoT) collects, transmits and stores data, EC allows the employment of data analysis with Machine Learning (ML) at its edge. This work uses a systematic mapping methodology to encompass 15081 works related to four SGs technologies: data analysis, EC, IoT, and context awareness. With 37 works after the filtering process, this review revealed that most papers use one to three approaches, while only two use all four technologies. The results also indicate that EC has been extensively used in SG solutions, with 22 selected studies. Distinctively, only 9 works use context awareness, which may indicate a path for future developments in SG. The study also allowed the learning of 7 lessons that are presented in this paper. [Display omitted] • We review the state of the art related to smart grid computational techniques. • The paper covers internet of things, edge computing, and context-aware computing. • The work considers literature of the last decade through a well-defined methodology. • The study covers 15081 papers, answering one general and eight focused issues. • The paper proposes a taxonomy for smart grid Technologies. [ABSTRACT FROM AUTHOR]

Unmanned Aerial Vehicles (UAVs), which can operate autonomously in dynamic and complex environments, are becoming increasingly common. Deep learning techniques for motion control have recently taken a major qualitative step since vision-based inference tasks can be executed directly on edge. The goal is to fully integrate the machine learning (ML) element into small UAVs. However, given the limited payload capacity and energy available on small UAVs, integrating computing resources sufficient to host ML and vehicle control functions is still challenging. This paper presents a modular and generic system that can control the UAV by evaluating vision-based ML tasks directly inside the resource-constrained UAV. Two different vision-based navigation configurations were tested and demonstrated. The first configuration implements an autonomous landing site detection system, tested with two models based on LeNet-5 and MobileNetV2, respectively. This allows the UAV to change its planned path accordingly and approach the target to land. Moreover, a model for people detection based on a custom MobileNetV2 network was evaluated in the second configuration. Finally, the execution time and power consumption were measured and compared with a cloud computing approach. The results show the ability of the developed system to dynamically react to the environment to provide the necessary maneuver after detecting the target exploiting only the constrained computational resources of the UAV controller. Furthermore, we demonstrated that moving to the edge, instead of using cloud computing inference, decreases the energy requirement of the system without reducing the quality of service. • Autonomous navigation in lightweight drones can be possible with Raspberry Pi SBC equipped with a neural accelerator. • The best configuration in terms of execution performance and energy consumption lead to Raspberry Pi4 and LeNet-5 architecture. • The tested applications are executed in near real time with a low energy profile. • Comparing the proposed platform with the cloud computing approach reveals its energy sustainability: it consumes 10x energy less. [ABSTRACT FROM AUTHOR]

The optimisation of 3D printing parameters for manufacturing biomedical devices is an emerging interdisciplinary field that incorporates artificial intelligence techniques such as machine learning and deep learning. In this particular study, the focus is on the fabrication of biocompatible finger splints using digital light processing 3D printing technology, followed by UV curing, to evaluate their quality. By leveraging vibration data from printers, which cannot be captured through visual inspection of layer defects, this study aims to develop a predictive model for assessing the failures of printed parts. Here, a closed-loop detection system is proposed to identify failure phenomena in 3D resin printing, combining both cloud and edge computing technologies to effectively detect and address potential failures in the printing process. [ABSTRACT FROM AUTHOR]

Forecasting precise water usage corresponding to various beneficial usages is important for optimal and sustainable planning and management of water resources. Due to rapid population growth, there is an urgent need for devising water saving solutions. In this paper, we propose a blockchain based incentivized edge computing framework for water saving using soft computing methodologies. The framework facilitates decision makers in creating awareness among people about water savings in a easily understandable scientific way. Our incentivized blockchain based model uses edge computing at the house nodes of the network to predict the actual usage of a particular household in the locality based on several factors such as number of people, average income of family, profession of the members and previous water demands. By using Feed Forward Networks and Mixture Density Networks, we predict the water usage in terms of input factors and historical usage respectively, thus incorporating machine computing into the framework. With the two values from these methods, a comparison is made with the actual amount of water used by the householders. This research proposes deployment of the smart contract on the blockchain network for efficient and accurate reward distribution. Incentives and rewards are given in the blockchain network to houses with lesser consumption and penalties are imposed when usage crosses predicted and historic usage. The model ensures that accurate incentives are provided to the people in order to motivate them to avoid wastage of water. Results show that the methods used in our work perform better than other relevant networks on a self-synthesized dataset. The proposed methods converge well and show higher spatio-temporal accuracy. [ABSTRACT FROM AUTHOR]

Emerging distributed applications recently boosted the development of decentralized machine learning, especially in IoT and edge computing fields. In real-world scenarios, the common problems of non-convexity and data heterogeneity result in inefficiency, performance degradation, and development stagnation. The bulk of studies concentrate on one of the issues mentioned above without having a more general framework that has been proven optimal. To this end, we propose a unified paradigm called UMP, which comprises two algorithms D-SUM and GT-DSUM based on the momentum technique with decentralized stochastic gradient descent (SGD). The former provides a convergence guarantee for general non-convex objectives, while the latter is extended by introducing gradient tracking, which estimates the global optimization direction to mitigate data heterogeneity (i.e. , distribution drift). We can cover most momentum-based variants based on the classical heavy ball or Nesterov's acceleration with different parameters in UMP. In theory, we rigorously provide the convergence analysis of these two approaches for non-convex objectives and conduct extensive experiments, demonstrating a significant improvement in model accuracy up to 57.6% compared to other methods in practice. [ABSTRACT FROM AUTHOR]

Deep Neural Networks (DNNs) have achieved remarkable success in many computer vision tasks recently, but the huge number of parameters and the high computation overhead hinder their deployments on resource-constrained edge devices. It is worth noting that channel pruning is an effective approach for compressing DNN models. A critical challenge is to determine which channels are to be removed, so that the model accuracy will not be negatively affected. In this paper, we first propose Spatial and Channel Attention (SCA), a new attention module combining both spatial and channel attention that respectively focuses on "where" and "what" are the most informative parts. Guided by the scale values generated by SCA for measuring channel importance, we further propose a new channel pruning approach called Channel Pruning guided by Spatial and Channel Attention (CPSCA). Experimental results indicate that SCA achieves the best inference accuracy, while incurring negligibly extra resource consumption, compared to other state-of-the-art attention modules. Our evaluation on two benchmark datasets shows that, with the guidance of SCA, our CPSCA approach achieves higher inference accuracy than other state-of-the-art pruning methods under the same pruning ratios. [ABSTRACT FROM AUTHOR]

More recently, there has been a growing interest in the blockchain technique for emerging multi-access edge computing (MEC) applications in relation to security and privacy, considering the benefits of using it in edge computing. In some blockchain-enabled MEC using the best-known consensus algorithm proof-of-work (PoW), the computational efforts increase dramatically with the number of transactions, since a large amount of computational tasks should be conducted by miners with the limited computing resources in edge devices. Hence, improving the computational performance as an important but challenging issue in the design of blockchain system for MEC applications, has attracted intensive attention within last years. To further improve the performance of PoW algorithm, we present a novel implementation mechanism in this paper. Here, motivated by proof-of-deep-learning (PoDL) method in which the deep learning algorithm is used to maintain blockchain, through the design of a two-stage model to achieving computational tasks in PoDL-based blockchain systems, a novel computational-efficient consensus mechanism, named separate-proof-of-deep-learning (S-PoDL), is accordingly proposed. Thus, an energy-efficient blockchain-enabled MEC could be developed with our proposed S-PoDL, which arranges miners to carry out a two-stage-based computation on the basis of accounting_queue technique, while presenting the achieved models as proofs in MEC network. The comparative experiments are conducted between S-PoDL and PoDL, and the experimental results verify the feasibility and efficiency of our consensus mechanism S-PoDL for some blockchain-enabled MEC applications in relation to PoW-based cryptocurrencies, while effectively reducing computing burden of edge devices. [ABSTRACT FROM AUTHOR]

Towards connectivity and development of reliable systems, smarting sensors are vastly applied in hi-tech industries. The wind energy is a growing market that could benefit from edge processing technology by enhancing monitoring systems, decreasing downtime and guiding predictive maintenance. We proposed an embedded multi-sensor architecture to detect incipient short-circuit in wind turbine electrical generators, that is robust to both false positives and negatives. Five different sensor settings are tested in three feature extraction methods and four classifiers. An analysis of variance (ANOVA) and a Tukey's honestly significant difference (HSD) statistical tests are used to determine which architectures should be embedded in a Raspberry Pi 3, NVIDIA Jetson TX2 and NVIDIA Xavier boards. A three current sensor setting with Fourier-MLP is the most suitable approach, achieving 81.20% of accuracy, 0% of false positive rate (FPR) and 0.08% of false negative rate (FNR), also detecting generator's normal conditions 100% of the time. For a single sensor configuration, current sensor is the most suitable method for detecting fault or non-fault conditions, being 16 times more robust to false negatives than using an axial flux sensor. Comparing the processing time, the system embedded in a NVIDIA Xavier predicts a fault condition 37% faster than in a Raspberry Pi 3, with Fourier-MLP and using a single current sensor, thus being the most suitable configuration in fault detection. • This paper presents a multi-sensor architecture for the identification of incipient shortcircuits in wind turbine generators. • The main advantage of this system embedded is to identify the six types of faults and robust to both false positives and negatives. • The combination Fourier-MLP is the most suitable approach, with 81.20% of accuracy, 0% of false-positive rate and 0.08% of false-negative rate, also detecting generator's normal conditions 100% of the time. [ABSTRACT FROM AUTHOR]

• The combination of mobile devices and edge computing reduces the total computing overhead and makes resource allocation reasonable. • Evaluate system performance by the weighted sum of delay and energy consumption. • Change the parameter update method to accelerate the convergence speed of the deep reinforcement learning algorithm. • A neural network is used to approximate the behavior value function, and a target q network is used to update the target. With the arrival of the 5th generation mobile networks (5 G) era, the data needed by mobile devices (MDs) is explosively growing. High-consumption, low-latency applications are huge challenges for resource-constrained Internet of things (IoT) devices. Mobile edge computing overcomes the limitations of computing resources on MDs by offloading tasks generated by MDs and assigning them to nearby MEC servers. Therefore, mobile edge computing (MEC) becomes important. This paper presents a task offloading strategy for the multi-device multi-server system. To meet the task requirements of different MDs, we formulate an overhead minimization problem to optimize the delay and energy consumption of the system. We propose the Double Deep Q Network (Double-DQN) algorithm to perform location selection strategies for tasks generated on the mobile devices and allocate respective computing resources. Simulation results show that the algorithm can allocate resources reasonably and reduce the overhead of the entire system. [Display omitted] [ABSTRACT FROM AUTHOR]

Today, incorporating advanced machine learning techniques into intrusion detection systems (IDSs) plays a crucial role in securing mobile edge computing systems. However, the mobility demands of our modern society require more advanced IDSs to make a good trade-off between coping with the rapid growth of traffic data and responding to attacks. Thus, in this paper, we propose a lightweight distributed IDS that exploits the advantages of centralized platforms to train and learn from large amounts of data. We investigate the benefits of two promising bio-inspired optimization algorithms, namely Ant Lion Optimization and Ant Colony Optimization, to find the optimal data representation for the classification process. We use Deep Forest as a classifier to detect intrusive actions more robustly and generate as few false positives as possible. The experiment results show that the proposed approach can enhance the reliability of lightweight intrusion detection systems in terms of accuracy and execution time. [Display omitted] • Hybrid learning model has benefits in supporting intrusion detection in mobile edge computing. • Deep Forest is effective in strengthening attack detection predictability of IDSs in mobile edge computing. • Bio-inspired optimization algorithms improve the classification process. • The employed strategy enhances attack recognition of IDSs to deal with different threats in mobile edge computing. [ABSTRACT FROM AUTHOR]

This article proposes a new Artificial Intelligent (AI) and Machine Learning (ML) based framework for 6G-enabled Intelligent edge computing. The framework will be equipped with multiple cognitive controllers to harmoniously control various aspects in distributed intelligence toward edge nodes collaboration. Autonomic cognitive controller for edge computing is a popular computing paradigm where the distributed metadata processing and edge intelligence are performed at edge node in 5G/6G network for management, connectivity and interoperability. Some of studies focused on edge management improvement such as reduce the response time and bandwidth costs. However, the previous approaches are inadequate to support autonomously management for large-scale deployment for connectivity for dynamic and reliable communication. We propose a cognitive controller for edge autonomy and collaboration application development. Finally, we discuss challenges and open issues toward cognitive controller and distributed edge intelligence. [ABSTRACT FROM AUTHOR]

• The property of actual PVCS is investigated by characterizing the EV charging behaviors and PV power distribution to explore the net power data regularity. The external factors that most correlated with PVCS power variation are captured by interpretable algorithms to guide EV behaviors to maximally utilize renewable energy. • At the software algorithm design stage, the forecasting performance trade-off between algorithm complexity and forecasting accuracy is fully considered for edge computing use. The model complexity is maximumly reduced by covering the entire cycle of forecasting process, including input data pruning, lightweight model training, and hyperparameter optimization. • At the hardware configuration stage, forecasting programming is implemented on Raspberry pi-based edge platform for rapid online training. The forecasting model is tested on various scenarios to verify the algorithm feasibility on the edge device. The PV-assisted charging station (PVCS) aggregates the two key resources of electric vehicle (EV) charging load and photovoltaic (PV) system to maximize operation profit. In order to quantify the PVCS impact to grid and the renewable energy utilization by EV charging load, it is crucial to predict the PVCS net power and capture the most correlated factors for power variation. Given that the PVCS is located near the user side, the forecasting model complexity must be restricted to meet the online training demands on edge computing, apart from ensuring the prediction accuracy. A PVCS net power forecasting approach is proposed by simplifying model complexity from the entire cycle of training process, including the input data pruning, lightweight model training and hyperparameter optimization. Finally, a comprehensive analysis of the real data in PVCS shows that the most sensitivity factors affecting the net power of PVCS is time of use price, and the deep auto-encoded extreme learning machine (DA-ELM) can make a satisfactory compromise between prediction accuracy and model complexity for edge computing utilization. The forecasting model has outstanding prediction performance on Raspberry pi-based edge platform, which has enough significance for PVCS promotion. [ABSTRACT FROM AUTHOR]

The Internet of Things (IoTs) has transformed the digital landscape by interconnecting billions of devices worldwide, paving the way for smart cities, homes, and industries. With the exponential growth of IoT devices and the vast amount of data they generate, concerns have arisen regarding efficient task-offloading strategies. Traditional cloud and edge computing methods, paired with basic Machine Learning (ML) algorithms, face several challenges in this regard. In this paper, we propose a novel approach to task offloading in a Device-to-Device (D2D)-Edge-Cloud computing using the Rainbow Deep Q-Network (DQN), an advanced Deep Reinforcement Learning (DRL) algorithm. This algorithm utilizes advanced neural networks to optimize task offloading in the three-tier framework. It balances the trade-offs among D2D, Device-to-Edge (D2E), and Device/Edge-to-Cloud (D2C/E2C) communications, benefiting both end users and servers. These networks leverage Deep Learning (DL) to discern patterns, evaluate potential offloading decisions, and adapt in real time to dynamic environments. We compared our proposed algorithm against other state-of-the-art methods. Through rigorous simulations, we achieved remarkable improvements across key metrics: an increase in energy efficiency by 29.8%, a 27.5% reduction in latency, and a 43.1% surge in utility. • We present a method for task offloading across the D2D, Edge, and Cloud spectrum, crucial for elevating operational efficiency in the IoT systems. • We incorporate the Rainbow DQN, an advanced DRL algorithm. • We establish compelling evidence of noteworthy improvements in key performance indicators. [ABSTRACT FROM AUTHOR]

The clinical diagnosis of cardiovascular diseases extensively relies on ECG, which, in its dynamic form, is vulnerable to noise interference during the acquisition phase, consequently leading to misdiagnosis. Addressing these challenges, signal quality assessment (SQA) algorithms have been employed; nevertheless, the insufficiency of ECG signal quality databases have resulted in suboptimal algorithmic performance. In light of this, the study proposes a Generative Adversarial Network (GAN) framework that combines the Swin Transformer concept to synthesize ECG signal quality data. In the realm of synthesis performance evaluation, the method put forth in this study has demonstrated superior outcomes in contrast to prior investigations. Notably, the method exhibits the utmost similarity in synthesizing heartbeats, attaining a commendable synthetic ECG signal diversity level of 98.71% coupled with a fidelity level reaching 98.46%. Meanwhile, various machine learning and deep learning algorithms were utilized to validate the synthesized data, leading to the attainment of improved results. Ultimately, the introduced algorithm is implemented within an edge computing module, enabling real-time synthesis of ECG data and facilitating the assessment and comparison of SQA algorithms. This endeavor is anticipated to mitigate the challenge of limited high-quality data availability for ECG signals, and exhibits potential for broader utilization in the synthesis of diverse physiological signals. • A Self-Attention GAN based on the Swin Transformer architecture is introduced. • Heartbeat and ECG signals of different categories are generated. • The effectiveness of the synthesized signal is verified on the ECG SQA algorithm. • The proposed algorithm is implemented on the edge computing module. [ABSTRACT FROM AUTHOR]

Machine learning has become an indispensable part of the existing technological domain. Edge computing and Internet of Things (IoT) together presents a new opportunity to imply machine learning techniques at the resource constrained embedded devices at the edge of the network. Conventional machine learning requires enormous amount of power to predict a scenario. Embedded machine learning – TinyML paradigm aims to shift such plethora from traditional high-end systems to low-end clients. Several challenges are paved while doing such transition such as, maintaining the accuracy of learning models, provide train-to-deploy facility in resource frugal tiny edge devices, optimizing processing capacity, and improving reliability. In this paper, we present an intuitive review about such possibilities for TinyML. We firstly, present background of TinyML. Secondly, we list the tool sets for supporting TinyML. Thirdly, we present key enablers for improvement of TinyML systems. Fourthly, we present state-of-the-art about frameworks for TinyML. Finally, we identify key challenges and prescribe a future roadmap for mitigating several research issues of TinyML. [ABSTRACT FROM AUTHOR]

Edge computing and artificial intelligence (AI), especially deep learning algorithms, are gradually intersecting to build the novel system, namely edge intelligence. However, the development of edge intelligence systems encounters several challenges, and one of these challenges is the computational gap between computation-intensive deep learning algorithms and less-capable edge systems. Due to the computational gap, many edge intelligence systems cannot meet the expected performance requirements. To bridge the gap, a plethora of new techniques and optimization methods were proposed in the past years: lightweight deep learning models, network compression, and efficient neural architecture search. Although some reviews or surveys have partially covered this large body of literature, we lack a systematic and comprehensive review to discuss all aspects of these deep learning techniques which are critical for edge intelligence implementation. As various and diverse methods, applicable to edge systems, are proposed, a holistic review would enable edge computing engineers and the community to understand the state-of-the-art deep learning techniques that are instrumental for edge intelligence and to facilitate the development of edge intelligence systems. This paper surveys the representative and latest deep learning techniques that are useful for edge intelligence systems, including hand-crafted models, model compression, hardware-aware neural architecture search, and adaptive deep learning models. Finally, based on observations and simple experiments we conducted, we discuss some future directions. [ABSTRACT FROM AUTHOR]

In the context of the application of artificial intelligence in an intellectual property trading platform, the number of demanders and suppliers that exchange scarce resources is growing continuously. Improvement of computational power promotes matching efficiency significantly. It is necessary to greatly reduce energy consumption in order to realize the machine learning process in terminals and microprocessors in edge computing (smart phones, wearable devices, automobiles, IoT devices, etc.) and reduce the resource burden of data centers. Machine learning algorithms generated in an open community lack standardization in practice, and hence require open innovation participation to reduce computing cost, shorten algorithm running time, and improve human-machine collaborative competitiveness. The purpose of this study was to find an economic range of the granularity in a decision tree, a popular machine learning algorithm. This work addresses the research questions of what the economic tree depth interval is and what the corresponding time cost is with increasing granularity given the number of matches. This study also aimed to balance the efficiency and cost via simulation. Results show that the benefit of decreasing the tree search depth brought by the increased evaluation granularity is not linear, which means that, in a given number of candidate matches, the granularity has a definite and relatively economical range. The selection of specific evaluation granularity in this range can obtain a smaller tree depth and avoid the occurrence of low efficiency, which is the excessive increase in the time cost. Hence, the standardization of an AI algorithm is applicable to edge computing scenarios, such as an intellectual property trading platform. The economic granularity interval can not only save computing resource costs but also save AI decision-making time and avoid human decision-maker time cost. [ABSTRACT FROM AUTHOR]

Edge computing has gained widespread adoption for time-sensitive applications by offloading a portion of IoT system workloads from the cloud to edge nodes. However, the limited resources of IoT edge devices hinder service deployment, making auto-scaling crucial for improving resource utilization in response to dynamic workloads. Recent solutions aim to make auto-scaling proactive by predicting future workloads and overcoming the limitations of reactive approaches. These proactive solutions often rely on time-series data analysis and machine learning techniques, especially Long Short-Term Memory (LSTM), thanks to its accuracy and prediction speed. However, existing auto-scaling solutions often suffer from oscillation issues, even when using a cooling-down strategy. Consequently, the efficiency of proactive auto-scaling depends on the prediction model accuracy and the degree of oscillation in the scaling actions. This paper proposes a novel approach to improve prediction accuracy and deal with oscillation issues. Our approach involves an automatic featurization phase that extracts features from time-series workload data, improving the prediction's accuracy. These extracted features also serve as a grid for controlling oscillation in generated scaling actions. Our experimental results demonstrate the effectiveness of our approach in improving prediction accuracy, mitigating oscillation phenomena, and enhancing the overall auto-scaling performance. [ABSTRACT FROM AUTHOR]

Edge Computing (EC) is a recent architectural paradigm that brings computation close to end-users with the aim of reducing latency and bandwidth bottlenecks, which 5G technologies are committed to further reduce, while also achieving higher reliability. EC enables computation offloading from end devices to edge nodes. Deciding whether a task should be offloaded, or not, is not trivial. Moreover, deciding when and where to offload a task makes things even harder and making inadequate or off-time decisions can undermine the EC approach. Recently, Artificial Intelligence (AI) techniques, such as Machine Learning (ML), have been used to help EC systems cope with this problem. AI promises accurate decisions, higher adaptability and portability, thus diminishing the cost of decision-making and the probability of error. In this work, we perform a literature review on computation offloading in EC systems with and without AI techniques. We analyze several AI techniques, especially ML-based, that display promising results, overcoming the shortcomings of current approaches for computing offloading coordination We sorted the ML algorithms into classes for better analysis and provide an in-depth analysis on the use of AI for offloading, in particular, in the use case of offloading in Vehicular Edge Computing Networks, actually one technology that gained more relevance in the last years, enabling a vast amount of solutions for computation and data offloading. We also discuss the main advantages and limitations of offloading, with and without the use of AI techniques. [ABSTRACT FROM AUTHOR]

Powered by edge computing, the last few years have seen a rapid growth in AIoT applications. Federated learning (FL), as a typical machine learning framework for edge intelligence, has attracted a large number of attention since it can protect user privacy. However, recent studies have shown that FL cannot fully ensure privacy. To address this, differential privacy technique is widely used in FL. Nevertheless, existing works neglect that data on devices are non-independent and identically distributed (Non-IID), which largely degrades model accuracy and convergence speed. In this paper, we propose PADP-FedMeta, a personalized and adaptive differentially private federated meta learning mechanism with a provable privacy and convergence guarantee. PADP-FedMeta mitigates the negative effect of Non-IID upon model accuracy by introducing federated meta learning, and significantly improves the convergence speed with an adaptive privacy parameter. Comprehensive experimental results show the effectiveness of our mechanism and its superior performance over the state-of-the-art differentially private FL schemes. [ABSTRACT FROM AUTHOR]

As the most promising machine learning technology, deep neural networks (DNNs) have garnered significant attention in the field of disease diagnosis, i.e., DNN-empowered disease diagnosis. DNN-empowered disease diagnostic models with complex neural network structures are generally computation-intensive and latency-sensitive. To reduce latency, the partitioning and offloading of DNN-empowered disease diagnostic models in edge computing networks has emerged as a potential solution. However, DNN partitioning still faces two key challenges: the limitation of edge resources and the dynamic of network environments. To address these challenges, we propose a cooperative DNN partitioning system aimed at accelerating the processing of the model in multi-access edge computing networks. Specifically, we model the cooperative DNN partitioning offloading problem as a multi-agent Markov decision process with the goal of minimizing the long-term service latency (i.e., accelerating DNN-empowered disease diagnosis). To tackle this optimization problem, we further propose a swarm reinforcement learning (SRL) algorithm. Each agent of the proposed SRL can learn from local data and generate a judicious offloading action independently. Numerous simulation results show that the proposed SRL outperforms existing offloading algorithms and can significantly accelerate DNN-empowered disease diagnosis. • We present a cooperative DNN partitioning system in end-edge cooperation networks. • We propose a swarm reinforcement learning algorithm to support the proposed system. • Simulation results demonstrate the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Deep neural networks are the state of the art for various machine learning problems dealing with large amounts of rich sensor data. It is often desirable to evaluate these models on edge devices instead of relying on cloud computing. In this paper, we perform traffic counting using surveillance cameras. Edge computing is required as only aggregated counts should leave the device and not the privacy sensitive video frames. Unfortunately, only small object detection models are suited for edge devices which results in sub-optimal performance. We introduce location specific models that are each trained for one specific camera. The model does not need to generalize to other locations. We show that smaller specialized models can outperform large general purpose models. We propose an automated way to train these small models without human intervention. We experimentally show that we can achieve a similar counting accuracy with 5x fewer parameters than state-of-the-art techniques. • We obtain high traffic counting accuracy in the real world using lightweight models. • We introduce location specific models that process the data of a single camera. • Since the models only need to work well for a single camera, they can be very small. • The location specific models can be trained automatically without human intervention. [ABSTRACT FROM AUTHOR]

During the COVID-19 pandemic, the patient care delivery paradigm rapidly shifted to remote technological solutions. Rising rates of life expectancy of older people, and deaths due to chronic diseases (CDs) such as cancer, diabetes and respiratory disease pose many challenges to healthcare. While the feasibility of Remote Patient Monitoring (RPM) with a Smart Healthcare Monitoring (SHM) framework was somewhat questionable before the COVID-19 pandemic, it is now a proven commodity and is on its way to becoming ubiquitous. More health organizations are adopting RPM to enable CD management in the absence of individual monitoring. The current studies on SHM have reviewed the applications of IoT and/or Machine Learning (ML) in the domain, their architecture, security, privacy and other network related issues. However, no study has analyzed the AI and ubiquitous computing advances in SHM frameworks. The objective of this research is to identify and map key technical concepts in the SHM framework. In this context an interesting and meaningful classification of the research articles surveyed for this work is presented. The comprehensive and systematic review is based on the "Preferred Reporting Items for Systematic Review and Meta-Analysis" (PRISMA) approach. A total of 2540 papers were screened from leading research archives from 2016 to March 2021, and finally, 50 articles were selected for review. The major advantages, developments, distinctive architectural structure, components, technical challenges and possibilities in SHM are briefly discussed. A review of various recent cloud and fog computing based architectures, major ML implementation challenges, prospects and future trends is also presented. The survey primarily encourages the data driven predictive analytics aspects of healthcare and the development of ML models for health empowerment. [ABSTRACT FROM AUTHOR]

Federated learning is an attractive distributed learning paradigm, which allows resource-constrained edge computing devices to cooperatively train machine learning models, while keeping data locally. However, the non-IID data distribution across devices is one of the main challenges that affect the performance of federated optimization algorithms. Inspired by knowledge distillation, this paper proposes a federated optimization algorithm, Federated Codistillation (FedCodl), in which a distillation term is added to the local objective function, so that local models can be trained on outputs of the global model. We further extend FedCodl to Federated Two-way Codistillation (Fed2Codl) to personalize local models for each device, meanwhile the global model is retained and iteratively updated in parallel. Then, we theoretically provide convergence guarantees for our approaches when learning strongly convex and smooth models. Finally, extensive experiments on federated benchmark datasets demonstrate that our approaches can achieve superior performance under the challenge of multiple non-IID data distribution settings on traditional and deep learning architectures, in terms of classification accuracy and communication efficiency. The biggest accuracy improvement against the standard Federated Averaging (FedAvg) framework is up to 22.26% over all communication rounds. Our results show the potential of the proposed approaches for further applications in federated learning. • A federated optimization framework based on knowledge codistillation is proposed. • An extension is presented to hold a personalized model for each federated device. • Theoretical convergence guarantees for our algorithms are provided. • Performance of the proposed schemes is evaluated on diverse federated benchmarks. [ABSTRACT FROM AUTHOR]

The increasing data produced by IoT devices and the need to harness intelligence in our environments impose the shift of computing and intelligence at the edge, leading to a novel computing paradigm called Edge Intelligence/Edge AI. This paradigm combines Artificial Intelligence and Edge Computing, enables the deployment of machine learning algorithms to the edge, where data is generated, and is able to overcome the drawbacks of a centralized approach based on the cloud (e.g., performance bottleneck, poor scalability, and single point of failure). Edge AI supports the distributed Federated Learning (FL) model that maintains local training data at the end devices and shares only globally learned model parameters in the cloud. This paper proposes a novel, energy-efficient, and dynamic FL-based approach considering a hierarchical edge FL architecture called HED-FL, which supports a sustainable learning paradigm using model parameters aggregation at different layers and considering adaptive learning rounds at the edge to save energy but still preserving the learning model's accuracy. Performance evaluations of the proposed approach have also been led out considering model accuracy, loss, and energy consumption. [ABSTRACT FROM AUTHOR]

Anomaly detection is concerned with identifying data patterns that deviate remarkably from the expected behavior. This is an important research problem, due to its broad set of application domains, from data analysis to e-health, cybersecurity, predictive maintenance, fault prevention, and industrial automation. Herein, we review state-of-the-art methods that may be employed to detect anomalies in the specific area of sensor systems, which poses hard challenges in terms of information fusion, data volumes, data speed, and network/energy efficiency, to mention but the most pressing ones. In this context, anomaly detection is a particularly hard problem, given the need to find computing-energy-accuracy trade-offs in a constrained environment. We taxonomize methods ranging from conventional techniques (statistical methods, time-series analysis, signal processing, etc.) to data-driven techniques (supervised learning, reinforcement learning, deep learning, etc.). We also look at the impact that different architectural environments (Cloud, Fog, Edge) can have on the sensors ecosystem. The review points to the most promising intelligent-sensing methods, and pinpoints a set of interesting open issues and challenges. • Survey and review of anomaly detection techniques for sensor systems. • Conventional vs data-driven techniques for anomaly detection in sensor systems. • Machine learning in the Internet of Things via Cloud, Fog and Edge computing. • Challenges and open issues for intelligent sensing and anomaly detection. [ABSTRACT FROM AUTHOR]

The Internet of Things (IoT) consists of heterogeneous devices such as wearables and monitoring devices that collect data to provide autonomous decision-making and smart applications. IoT technologies, such as the Internet of Medical Things (IoMT), have become gradually popular for medical purposes, combining IoT and medical devices to achieve good health and well-being. However, IoMT devices are often tight and have resource constraints, which leads to limited local data processing in the device. Edge computing provides access to additional computation and storage resources for IoMT devices, bringing intelligent processing closer to the data sources. This technology opens up great possibilities for IoMT applications, especially when combined with Artificial Intelligence (AI). Edge AI runs AI computations close to the IoT devices and users instead of centralized services such as cloud servers. This paper investigates the potential of Edge AI and IoMT. In this sense, this survey is the first work to further detail Edge AI and Machine Learning Operations in IoMT domains and wearable technology, thus contributing to the literature by comprehensively exploring the potential of ML strategies and operations at the network's edge and intelligence distribution. This study also presents a case study on heart anomaly detection. [ABSTRACT FROM AUTHOR]

In the Unmanned Aerial Vehicle (UAV)-assisted mobile edge computing (MEC) system, most existing studies do not effectively consider the task offloading of user equipment (UEs) and the economic issues related to MEC service providers. Firstly, this paper considers the UE offloading cost and the pricing of MEC server, and establishes the UE, UAV cost and UAV revenue model. For the established non-convex optimization problem, it is further described as a Markov decision process (MDP), and a multi-agent reinforcement deep learning algorithm (MADRL) is proposed to minimize the system energy consumption by jointly optimizing power control, resource allocation and UE association, so as to effectively improve the overall revenue of UAV under the premise of ensuring the system performance. The simulation results show that our scheme can effectively improve the UAV revenue, and is significantly better than the random scheme without energy consumption optimization. In addition, the revenue obtained by our design scheme is 10.7% higher than that of the comparison scheme. [ABSTRACT FROM AUTHOR]

Incremental learning is a paradigm that extends knowledge by learning from new data, often used to add new classes to an existing model or to learn a new domain. It imposes strict limitations on the model's access to data from previous tasks, making it similar to the human learning process. The main challenge of incremental learning is catastrophic forgetting, where previous knowledge is severely forgotten while learning new tasks. In this work, we propose a novel approach for domain-incremental learning. Inspired by the Normal Equation , we accumulate the Gram Matrix from each task's hidden layer output to update a simplified RVFL model. This algorithm achieves performance comparable to joint training while strictly adhering to privacy restrictions. With issues such as forgetting, storage requirements and privacy protection be addressed, this algorithm has the potential to play a crucial role in the field of edge computing and other related fields. • We propose a strategy to train a simplified RVFL network for a more effective domain IL. • Simplified RVFL can achieve performance comparable to joint training in domain-IL. • Training RVFL using our algorithm completely resolves catastrophic forgetting. • Privacy and data security are preserved in our approach. • Our approach offers fast speed while ensuring accuracy, ideal for Edge Computing. [ABSTRACT FROM AUTHOR]

Nowadays, video surveillance systems are commonly found in most public and private spaces. These systems typically consist of a network of cameras that feed into a central node. However, the processing aspect is evolving towards distributed approaches, leveraging edge-computing. These distributed systems are capable of effectively addressing the detection of people or events at each individual node. Most of these systems, rely on the use of deep-learning and segmentation algorithms which enable them to achieve high performance, but usually with a significant computational cost, hindering real-time execution. This paper presents an approach for people detection and action recognition in the wild, optimized for running on the edge, and that is able to work in real-time, in an embedded platform. Human Action Recognition (HAR) is performed by using a Recurrent Neural Network (RNN), specifically a Long Short-Term Memory (LSTM). The input to the LSTM is an ad-hoc, lightweight feature vector obtained from the bounding box of each detected person in the video surveillance image. The resulting system is highly portable and easily scalable, providing a powerful tool for real-world video surveillance applications (in the wild and real-time action recognition). The proposal has been exhaustively evaluated and compared against other state-of-the-art (SOTA) proposals in five datasets, including four widely used (KTH, WEIZMAN, WVU, IXMAX) and a novel one (GBA) recorded in the wild, that includes several people performing different actions simultaneously. The obtained results validate the proposal, since it achieves SOTA accuracy within a much more complicated video surveillance real scenario, and using a lightweight embedded hardware. • Real-time action recognition based on LSTMs with an ultralight feature vector. • Experimental results on for widely used dataset and a novel dataset in-the-wild. • Comparison to other state-of-the-art methods. • Achieves accuracy close to state-of-the-art while reducing computational cost. • Proposal runs in real time in an embedded platform. [ABSTRACT FROM AUTHOR]

The surging pace of Internet of Things (IoT) development and its applications has resulted in significantly large amounts of data (commonly known as big data) being communicated and processed across IoT networks. While cloud computing has led to several possibilities in regard to this computational challenge, there are several security risks and concerns associated with it. Edge computing is a state-of-the-art subject in IoT that attempts to decentralize, distribute and transfer computation to IoT nodes. Furthermore, IoT nodes that perform applications are the primary target vectors which allow cybercriminals to threaten an IoT network. Hence, providing applied and robust methods to detect malicious activities by nodes is a big step to protect all of the network. In this study, we transmute the programs' OpCodes into a vector space and employ fuzzy and fast fuzzy pattern tree methods for malware detection and categorization. We obtained a high degree of accuracy during reasonable run-times especially for the fast fuzzy pattern tree. Both utilized feature extraction and fuzzy classification, which were robust, led to more powerful edge computing malware detection and categorization method. [ABSTRACT FROM AUTHOR]