Search

Edge machine learning (ML) enables localized processing of data on devices and is underpinned by deep neural networks (DNNs). However, DNNs cannot be easily run on devices due to their substantial computing, memory and energy requirements for delivering performance that is comparable to cloud-based ML. Therefore, model compression techniques, such as pruning, have been considered. Existing pruning methods are problematic for edge ML since they: (1) Create compressed models that have limited runtime performance benefits (using unstructured pruning) or compromise the final model accuracy (using structured pruning), and (2) Require substantial compute resources and time for identifying a suitable compressed DNN model (using neural architecture search). In this paper, we explore a new avenue, referred to as Pruning-at-Initialization (PaI), using structured pruning to mitigate the above problems. We develop Reconvene, a system for rapidly generating pruned models suited for edge deployments using structured PaI. Reconvene systematically identifies and prunes DNN convolution layers that are least sensitive to structured pruning. Reconvene rapidly creates pruned DNNs within seconds that are up to 16.21x smaller and 2x faster while maintaining the same accuracy as an unstructured PaI counterpart., Comment: The 24th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing

Modern deep neural networks (DNNs) consist of millions of parameters, necessitating high-performance computing during training and inference. Pruning is one solution that significantly reduces the space and time complexities of DNNs. Traditional pruning methods that are applied post-training focus on streamlining inference, but there are recent efforts to leverage sparsity early on by pruning before training. Pruning methods, such as iterative magnitude-based pruning (IMP) achieve up to a 90% parameter reduction while retaining accuracy comparable to the original model. However, this leads to impractical runtime as it relies on multiple train-prune-reset cycles to identify and eliminate redundant parameters. In contrast, training agnostic early pruning methods, such as SNIP and SynFlow offer fast pruning but fall short of the accuracy achieved by IMP at high sparsities. To bridge this gap, we present Dual Gradient-Based Rapid Iterative Pruning (DRIVE), which leverages dense training for initial epochs to counteract the randomness inherent at the initialization. Subsequently, it employs a unique dual gradient-based metric for parameter ranking. It has been experimentally demonstrated for VGG and ResNet architectures on CIFAR-10/100 and Tiny ImageNet, and ResNet on ImageNet that DRIVE consistently has superior performance over other training-agnostic early pruning methods in accuracy. Notably, DRIVE is 43$\times$ to 869$\times$ faster than IMP for pruning.

Efficient on-device Convolutional Neural Network (CNN) training in resource-constrained mobile and edge environments is an open challenge. Backpropagation is the standard approach adopted, but it is GPU memory intensive due to its strong inter-layer dependencies that demand intermediate activations across the entire CNN model to be retained in GPU memory. This necessitates smaller batch sizes to make training possible within the available GPU memory budget, but in turn, results in substantially high and impractical training time. We introduce NeuroFlux, a novel CNN training system tailored for memory-constrained scenarios. We develop two novel opportunities: firstly, adaptive auxiliary networks that employ a variable number of filters to reduce GPU memory usage, and secondly, block-specific adaptive batch sizes, which not only cater to the GPU memory constraints but also accelerate the training process. NeuroFlux segments a CNN into blocks based on GPU memory usage and further attaches an auxiliary network to each layer in these blocks. This disrupts the typical layer dependencies under a new training paradigm - $\textit{`adaptive local learning'}$. Moreover, NeuroFlux adeptly caches intermediate activations, eliminating redundant forward passes over previously trained blocks, further accelerating the training process. The results are twofold when compared to Backpropagation: on various hardware platforms, NeuroFlux demonstrates training speed-ups of 2.3$\times$ to 6.1$\times$ under stringent GPU memory budgets, and NeuroFlux generates streamlined models that have 10.9$\times$ to 29.4$\times$ fewer parameters., Comment: Accepted to EuroSys 2024

Deep neural networks (DNNs) underpin many machine learning applications. Production quality DNN models achieve high inference accuracy by training millions of DNN parameters which has a significant resource footprint. This presents a challenge for resources operating at the extreme edge of the network, such as mobile and embedded devices that have limited computational and memory resources. To address this, models are pruned to create lightweight, more suitable variants for these devices. Existing pruning methods are unable to provide similar quality models compared to their unpruned counterparts without significant time costs and overheads or are limited to offline use cases. Our work rapidly derives suitable model variants while maintaining the accuracy of the original model. The model variants can be swapped quickly when system and network conditions change to match workload demand. This paper presents DNNShifter, an end-to-end DNN training, spatial pruning, and model switching system that addresses the challenges mentioned above. At the heart of DNNShifter is a novel methodology that prunes sparse models using structured pruning. The pruned model variants generated by DNNShifter are smaller in size and thus faster than dense and sparse model predecessors, making them suitable for inference at the edge while retaining near similar accuracy as of the original dense model. DNNShifter generates a portfolio of model variants that can be swiftly interchanged depending on operational conditions. DNNShifter produces pruned model variants up to 93x faster than conventional training methods. Compared to sparse models, the pruned model variants are up to 5.14x smaller and have a 1.67x inference latency speedup, with no compromise to sparse model accuracy. In addition, DNNShifter has up to 11.9x lower overhead for switching models and up to 3.8x lower memory utilisation than existing approaches., Comment: 14 pages, 7 figures, 5 tables

Efficiently running federated learning (FL) on resource-constrained devices is challenging since they are required to train computationally intensive deep neural networks (DNN) independently. DNN partitioning-based FL (DPFL) has been proposed as one mechanism to accelerate training where the layers of a DNN (or computation) are offloaded from the device to the server. However, this creates significant communication overheads since the intermediate activation and gradient need to be transferred between the device and the server during training. While current research reduces the communication introduced by DNN partitioning using local loss-based methods, we demonstrate that these methods are ineffective in improving the overall efficiency (communication overhead and training speed) of a DPFL system. This is because they suffer from accuracy degradation and ignore the communication costs incurred when transferring the activation from the device to the server. This article proposes EcoFed - a communication efficient framework for DPFL systems. EcoFed eliminates the transmission of the gradient by developing pre-trained initialization of the DNN model on the device for the first time. This reduces the accuracy degradation seen in local loss-based methods. In addition, EcoFed proposes a novel replay buffer mechanism and implements a quantization-based compression technique to reduce the transmission of the activation. It is experimentally demonstrated that EcoFed can reduce the communication cost by up to 133x and accelerate training by up to 21x when compared to classic FL. Compared to vanilla DPFL, EcoFed achieves a 16x communication reduction and 2.86x training time speed-up. EcoFed is available from https://github.com/blessonvar/EcoFed.

Resource management in computing is a very challenging problem that involves making sequential decisions. Resource limitations, resource heterogeneity, dynamic and diverse nature of workload, and the unpredictability of fog/edge computing environments have made resource management even more challenging to be considered in the fog landscape. Recently Artificial Intelligence (AI) and Machine Learning (ML) based solutions are adopted to solve this problem. AI/ML methods with the capability to make sequential decisions like reinforcement learning seem most promising for these type of problems. But these algorithms come with their own challenges such as high variance, explainability, and online training. The continuously changing fog/edge environment dynamics require solutions that learn online, adopting changing computing environment. In this paper, we used standard review methodology to conduct this Systematic Literature Review (SLR) to analyze the role of AI/ML algorithms and the challenges in the applicability of these algorithms for resource management in fog/edge computing environments. Further, various machine learning, deep learning and reinforcement learning techniques for edge AI management have been discussed. Furthermore, we have presented the background and current status of AI/ML-based Fog/Edge Computing. Moreover, a taxonomy of AI/ML-based resource management techniques for fog/edge computing has been proposed and compared the existing techniques based on the proposed taxonomy. Finally, open challenges and promising future research directions have been identified and discussed in the area of AI/ML-based fog/edge computing., Comment: 49 page, 15 figures, 10 tables

Collaborative machine learning (CML) techniques, such as federated learning, have been proposed to train deep learning models across multiple mobile devices and a server. CML techniques are privacy-preserving as a local model that is trained on each device instead of the raw data from the device is shared with the server. However, CML training is inefficient due to low resource utilization. We identify idling resources on the server and devices due to sequential computation and communication as the principal cause of low resource utilization. A novel framework PiPar that leverages pipeline parallelism for CML techniques is developed to substantially improve resource utilization. A new training pipeline is designed to parallelize the computations on different hardware resources and communication on different bandwidth resources, thereby accelerating the training process in CML. A low overhead automated parameter selection method is proposed to optimize the pipeline, maximizing the utilization of available resources. The experimental results confirm the validity of the underlying approach of PiPar and highlight that when compared to federated learning: (i) the idle time of the server can be reduced by up to 64.1x, and (ii) the overall training time can be accelerated by up to 34.6x under varying network conditions for a collection of six small and large popular deep neural networks and four datasets without sacrificing accuracy. It is also experimentally demonstrated that PiPar achieves performance benefits when incorporating differential privacy methods and operating in environments with heterogeneous devices and changing bandwidths.

This paper analyzes the effects of dynamically varying video contents and detection latency on the real-time detection accuracy of a detector and proposes a new run-time accuracy variation model, ROMA, based on the findings from the analysis. ROMA is designed to select an optimal detector out of a set of detectors in real time without label information to maximize real-time object detection accuracy. ROMA utilizing four YOLOv4 detectors on an NVIDIA Jetson Nano shows real-time accuracy improvements by 4 to 37% for a scenario of dynamically varying video contents and detection latency consisting of MOT17Det and MOT20Det datasets, compared to individual YOLOv4 detectors and two state-of-the-art runtime techniques., Comment: Accepted at the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) 2023

Federated learning (FL) trains machine learning (ML) models on devices using locally generated data and exchanges models without transferring raw data to a distant server. This exchange incurs a communication overhead and impacts the performance of FL training. There is limited understanding of how communication protocols specifically contribute to the performance of FL. Such an understanding is essential for selecting the right communication protocol when designing an FL system. This paper presents FedComm, a benchmarking methodology to quantify the impact of optimized application layer protocols, namely Message Queue Telemetry Transport (MQTT), Advanced Message Queuing Protocol (AMQP), and ZeroMQ Message Transport Protocol (ZMTP), and non-optimized application layer protocols, namely as TCP and UDP, on the performance of FL. FedComm measures the overall performance of FL in terms of communication time and accuracy under varying computational and network stress and packet loss rates. Experiments on a lab-based testbed demonstrate that TCP outperforms UDP as a non-optimized application layer protocol with higher accuracy and shorter communication times for 4G and Wi-Fi networks. Optimized application layer protocols such as AMQP, MQTT, and ZMTP outperformed non-optimized application layer protocols in most network conditions, resulting in a 2.5x reduction in communication time compared to TCP while maintaining accuracy. The experimental results enable us to highlight a number of open research issues for further investigation. FedComm is available for download from https://github.com/qub-blesson/FedComm.

Partitioning and deploying Deep Neural Networks (DNNs) across edge nodes may be used to meet performance objectives of applications. However, the failure of a single node may result in cascading failures that will adversely impact the delivery of the service and will result in failure to meet specific objectives. The impact of these failures needs to be minimised at runtime. Three techniques are explored in this paper, namely repartitioning, early-exit and skip-connection. When an edge node fails, the repartitioning technique will repartition and redeploy the DNN thus avoiding the failed nodes. The early-exit technique makes provision for a request to exit (early) before the failed node. The skip connection technique dynamically routes the request by skipping the failed nodes. This paper will leverage trade-offs in accuracy, end-to-end latency and downtime for selecting the best technique given user-defined objectives (accuracy, latency and downtime thresholds) when an edge node fails. To this end, CONTINUER is developed. Two key activities of the framework are estimating the accuracy and latency when using the techniques for distributed DNNs and selecting the best technique. It is demonstrated on a lab-based experimental testbed that CONTINUER estimates accuracy and latency when using the techniques with no more than an average error of 0.28% and 13.06%, respectively and selects the suitable technique with a low overhead of no more than 16.82 milliseconds and an accuracy of up to 99.86%., Comment: 10 pages

Based on the collective input of Dagstuhl Seminar (21342), this paper presents a comprehensive discussion on AI methods and capabilities in the context of edge computing, referred as Edge AI. In a nutshell, we envision Edge AI to provide adaptation for data-driven applications, enhance network and radio access, and allow the creation, optimization, and deployment of distributed AI/ML pipelines with given quality of experience, trust, security and privacy targets. The Edge AI community investigates novel ML methods for the edge computing environment, spanning multiple sub-fields of computer science, engineering and ICT. The goal is to share an envisioned roadmap that can bring together key actors and enablers to further advance the domain of Edge AI., Comment: for ACM SIGCOMM CCR

Distributed machine learning algorithms that employ Deep Neural Networks (DNNs) are widely used in Industry 4.0 applications, such as smart manufacturing. The layers of a DNN can be mapped onto different nodes located in the cloud, edge and shop floor for preserving privacy. The quality of the data that is fed into and processed through the DNN is of utmost importance for critical tasks, such as inspection and quality control. Distributed Data Validation Networks (DDVNs) are used to validate the quality of the data. However, they are prone to single points of failure when an attack occurs. This paper proposes QUDOS, an approach that enhances the security of a distributed DNN that is supported by DDVNs using quorums. The proposed approach allows individual nodes that are corrupted due to an attack to be detected or excluded when the DNN produces an output. Metrics such as corruption factor and success probability of an attack are considered for evaluating the security aspects of DNNs. A simulation study demonstrates that if the number of corrupted nodes is less than a given threshold for decision-making in a quorum, the QUDOS approach always prevents attacks. Furthermore, the study shows that increasing the size of the quorum has a better impact on security than increasing the number of layers. One merit of QUDOS is that it enhances the security of DNNs without requiring any modifications to the algorithm and can therefore be applied to other classes of problems.

Federated learning (FL) is a privacy-preserving distributed machine learning technique that trains models while keeping all the original data generated on devices locally. Since devices may be resource constrained, offloading can be used to improve FL performance by transferring computational workload from devices to edge servers. However, due to mobility, devices participating in FL may leave the network during training and need to connect to a different edge server. This is challenging because the offloaded computations from edge server need to be migrated. In line with this assertion, we present FedFly, which is, to the best of our knowledge, the first work to migrate a deep neural network (DNN) when devices move between edge servers during FL training. Our empirical results on the CIFAR10 dataset, with both balanced and imbalanced data distribution, support our claims that FedFly can reduce training time by up to 33% when a device moves after 50% of the training is completed, and by up to 45% when 90% of the training is completed when compared to state-of-the-art offloading approach in FL. FedFly has negligible overhead of up to two seconds and does not compromise accuracy. Finally, we highlight a number of open research issues for further investigation. FedFly can be downloaded from https://github.com/qub-blesson/FedFly., Comment: 7 pages, 6 figures

Applying Federated Learning (FL) on Internet-of-Things devices is necessitated by the large volumes of data they produce and growing concerns of data privacy. However, there are three challenges that need to be addressed to make FL efficient: (i) execution on devices with limited computational capabilities, (ii) accounting for stragglers due to computational heterogeneity of devices, and (iii) adaptation to the changing network bandwidths. This paper presents FedAdapt, an adaptive offloading FL framework to mitigate the aforementioned challenges. FedAdapt accelerates local training in computationally constrained devices by leveraging layer offloading of deep neural networks (DNNs) to servers. Further, FedAdapt adopts reinforcement learning based optimization and clustering to adaptively identify which layers of the DNN should be offloaded for each individual device on to a server to tackle the challenges of computational heterogeneity and changing network bandwidth. Experimental studies are carried out on a lab-based testbed and it is demonstrated that by offloading a DNN from the device to the server FedAdapt reduces the training time of a typical IoT device by over half compared to classic FL. The training time of extreme stragglers and the overall training time can be reduced by up to 57%. Furthermore, with changing network bandwidth, FedAdapt is demonstrated to reduce the training time by up to 40% when compared to classic FL, without sacrificing accuracy., Comment: 13 pages

Deep Neural Networks (DNNs) may be partitioned across the edge and the cloud to improve the performance efficiency of inference. DNN partitions are determined based on operational conditions such as network speed. When operational conditions change DNNs will need to be repartitioned to maintain the overall performance. However, repartitioning using existing approaches, such as Pause and Resume, will incur a service downtime on the edge. This paper presents the NEUKONFIG framework that identifies the service downtime incurred when repartitioning DNNs and proposes approaches for reducing edge service downtime. The proposed approaches are based on 'Dynamic Switching' in which, when the network speed changes and given an existing edge-cloud pipeline, a new edge-cloud pipeline is initialised with new DNN partitions. Incoming inference requests are switched to the new pipeline for processing data. Two dynamic switching scenarios are considered: when a second edge-cloud pipeline is always running and when a second pipeline is only initialised when the network speed changes. Experimental studies are carried out on a lab-based testbed to demonstrate that Dynamic Switching reduces the downtime by at least an order of magnitude when compared to a baseline using Pause and Resume that has a downtime of 6 seconds. A trade-off in the edge service downtime and memory required is noted. The Dynamic Switching approach that requires the same amount of memory as the baseline reduces the edge service downtime to 0.6 seconds and to less than 1 millisecond in the best case when twice the amount of memory as the baseline is available., Comment: 10 pages

This article argues that low latency, high bandwidth, device proliferation, sustainable digital infrastructure, and data privacy and sovereignty continue to motivate the need for edge computing research even though its initial concepts were formulated more than a decade ago.

Real-time video analytics on the edge is challenging as the computationally constrained resources typically cannot analyse video streams at full fidelity and frame rate, which results in loss of accuracy. This paper proposes a Transprecise Object Detector (TOD) which maximises the real-time object detection accuracy on an edge device by selecting an appropriate Deep Neural Network (DNN) on the fly with negligible computational overhead. TOD makes two key contributions over the state of the art: (1) TOD leverages characteristics of the video stream such as object size and speed of movement to identify networks with high prediction accuracy for the current frames; (2) it selects the best-performing network based on projected accuracy and computational demand using an effective and low-overhead decision mechanism. Experimental evaluation on a Jetson Nano demonstrates that TOD improves the average object detection precision by 34.7 % over the YOLOv4-tiny-288 model on average over the MOT17Det dataset. In the MOT17-05 test dataset, TOD utilises only 45.1 % of GPU resource and 62.7 % of the GPU board power without losing accuracy, compared to YOLOv4-416 model. We expect that TOD will maximise the application of edge devices to real-time object detection, since TOD maximises real-time object detection accuracy given edge devices according to dynamic input features without increasing inference latency in practice.

Industrial Internet of Things (IIoT) applications can benefit from leveraging edge computing. For example, applications underpinned by deep neural networks (DNN) models can be sliced and distributed across the IIoT device and the edge of the network for improving the overall performance of inference and for enhancing privacy of the input data, such as industrial product images. However, low network performance between IIoT devices and the edge is often a bottleneck. In this study, we develop ScissionLite, a holistic framework for accelerating distributed DNN inference using the Transfer Layer (TL). The TL is a traffic-aware layer inserted between the optimal slicing point of a DNN model slice in order to decrease the outbound network traffic without a significant accuracy drop. For the TL, we implement a new lightweight down/upsampling network for performance-limited IIoT devices. In ScissionLite, we develop ScissionTL, the Preprocessor, and the Offloader for end-to-end activities for deploying DNN slices with the TL. They decide the optimal slicing point of the DNN, prepare pre-trained DNN slices including the TL, and execute the DNN slices on an IIoT device and the edge. Employing the TL for the sliced DNN models has a negligible overhead. ScissionLite improves the inference latency by up to 16 and 2.8 times when compared to execution on the local device and an existing state-of-the-art model slicing approach respectively., Comment: 10 pages

Edge computing offers the distinct advantage of harnessing compute capabilities on resources located at the edge of the network to run workloads of relatively weak user devices. This is achieved by offloading computationally intensive workloads, such as deep learning from user devices to the edge. Using the edge reduces the overall communication latency of applications as workloads can be processed closer to where data is generated on user devices rather than sending them to geographically distant clouds. Specialised hardware accelerators, such as Graphics Processing Units (GPUs) available in the cloud-edge network can enhance the performance of computationally intensive workloads that are offloaded from devices on to the edge. The underlying approach required to facilitate this is virtualization of GPUs. This paper therefore sets out to investigate the potential of GPU accelerator virtualization to improve the performance of deep learning workloads in a cloud-edge environment. The AVEC accelerator virtualization framework is proposed that incurs minimum overheads and requires no source-code modification of the workload. AVEC intercepts local calls to a GPU on a device and forwards them to an edge resource seamlessly. The feasibility of AVEC is demonstrated on a real-world application, namely OpenPose using the Caffe deep learning library. It is observed that on a lab-based experimental test-bed AVEC delivers up to 7.48x speedup despite communication overheads incurred due to data transfers., Comment: 8 pages, 13 figures

By mechanically distorting a crystal lattice it is possible to engineer the electronic and optical properties of a material. In graphene, one of the major effects of such a distortion is an energy shift of the Dirac point, often described as a scalar potential. We demonstrate how such a scalar potential can be generated systematically over an entire electronic device and how the resulting changes in the graphene work function can be detected in transport experiments. Combined with Raman spectroscopy, we obtain a characteristic scalar potential consistent with recent theoretical estimates. This direct evidence for a scalar potential on a macroscopic scale due to deterministically generated strain in graphene paves the way for engineering the optical and electronic properties of graphene and similar materials by using external strain.

Partitioning and distributing deep neural networks (DNNs) across end-devices, edge resources and the cloud has a potential twofold advantage: preserving privacy of the input data, and reducing the ingress bandwidth demand beyond the edge. However, for a given DNN, identifying the optimal partition configuration for distributing the DNN that maximizes performance is a significant challenge. This is because the combination of potential target hardware resources that maximizes performance and the sequence of layers of the DNN that should be distributed across the target resources needs to be determined, while accounting for user-defined objectives/constraints for partitioning. This paper presents Scission, a tool for automated benchmarking of DNNs on a given set of target device, edge and cloud resources for determining optimal partitions that maximize DNN performance. The decision-making approach is context-aware by capitalizing on hardware capabilities of the target resources, their locality, the characteristics of DNN layers, and the network condition. Experimental studies are carried out on 18 DNNs. The decisions made by Scission cannot be manually made by a human given the complexity and the number of dimensions affecting the search space. The benchmarking overheads of Scission allow for responding to operational changes periodically rather than in real-time. Scission is available for public download at https://github.com/qub-blesson/Scission., Comment: Accepted to IEEE/ACM UCC 2020

Edge computing offers an additional layer of compute infrastructure closer to the data source before raw data from privacy-sensitive and performance-critical applications is transferred to a cloud data center. Deep Neural Networks (DNNs) are one class of applications that are reported to benefit from collaboratively computing between the edge and the cloud. A DNN is partitioned such that specific layers of the DNN are deployed onto the edge and the cloud to meet performance and privacy objectives. However, there is limited understanding of: (a) whether and how evolving operational conditions (increased CPU and memory utilization at the edge or reduced data transfer rates between the edge and the cloud) affect the performance of already deployed DNNs, and (b) whether a new partition configuration is required to maximize performance. A DNN that adapts to changing operational conditions is referred to as an 'adaptive DNN'. This paper investigates whether there is a case for adaptive DNNs in edge computing by considering three questions: (i) Are DNNs sensitive to operational conditions? (ii) How sensitive are DNNs to operational conditions? (iii) Do individual or a combination of operational conditions equally affect DNNs? (iv) Is DNN partitioning sensitive to hardware architectures on the cloud/edge? The exploration is carried out in the context of 8 pre-trained DNN models and the results presented are from analyzing nearly 8 million data points. The results highlight that network conditions affects DNN performance more than CPU or memory related operational conditions. Repartitioning is noted to provide a performance gain in a number of cases, but a specific trend was not noted in relation to its correlation to the underlying hardware architecture. Nonetheless, the need for adaptive DNNs is confirmed.

Containers are becoming a popular workload deployment mechanism in modern distributed systems. However, there are limited software-based methods (hardware-based methods are expensive requiring hardware level changes) for obtaining the power consumed by containers for facilitating power-aware container scheduling, an essential activity for efficient management of distributed systems. This paper presents WattsApp, a tool underpinned by a six step software-based method for power-aware container scheduling to minimize power cap violations on a server. The proposed method relies on a neural network-based power estimation model and a power capped container scheduling technique. Experimental studies are pursued in a lab-based environment on 10 benchmarks deployed on Intel and ARM processors. The results highlight that the power estimation model has negligible overheads for data collection - nearly 90% of all data samples can be estimated with less than a 10% error, and the Mean Absolute Percentage Error (MAPE) is less than 6%. The power-aware scheduling of WattsApp is more effective than Intel's Running Power Average Limit (RAPL) based power capping for both single and multiple containers as it does not degrade the performance of all containers running on the server. The results confirm the feasibility of WattsApp.

Edge computing is the next Internet frontier that will leverage computing resources located near users, sensors, and data stores to provide more responsive services. Therefore, it is envisioned that a large-scale, geographically dispersed, and resource-rich distributed system will emerge and play a key role in the future Internet. However, given the loosely coupled nature of such complex systems, their operational conditions are expected to change significantly over time. In this context, the performance characteristics of such systems will need to be captured rapidly, which is referred to as performance benchmarking, for application deployment, resource orchestration, and adaptive decision-making. Edge performance benchmarking is a nascent research avenue that has started gaining momentum over the past five years. This article first reviews articles published over the past three decades to trace the history of performance benchmarking from tightly coupled to loosely coupled systems. It then systematically classifies previous research to identify the system under test, techniques analyzed, and benchmark runtime in edge performance benchmarking., Comment: Accepted by ACM Computing Surveys, 16 December 2020

Existing power modelling research focuses on the model rather than the process for developing models. An automated power modelling process that can be deployed on different processors for developing power models with high accuracy is developed. For this, (i) an automated hardware performance counter selection method that selects counters best correlated to power on both ARM and Intel processors, (ii) a noise filter based on clustering that can reduce the mean error in power models, and (iii) a two stage power model that surmounts challenges in using existing power models across multiple architectures are proposed and developed. The key results are: (i) the automated hardware performance counter selection method achieves comparable selection to the manual method reported in the literature, (ii) the noise filter reduces the mean error in power models by up to 55%, and (iii) the two stage power model can predict dynamic power with less than 8% error on both ARM and Intel processors, which is an improvement over classic models., Comment: 10 pages; IEEE/ACM CCGrid 2020. arXiv admin note: text overlap with arXiv:1710.10325

Fog computing has emerged as a computing paradigm aimed at addressing the issues of latency, bandwidth and privacy when mobile devices are communicating with remote cloud services. The concept is to offload compute services closer to the data. However many challenges exist in the realisation of this approach. During offloading, (part of) the application underpinned by the services may be unavailable, which the user will experience as down time. This paper describes work aimed at building models to allow prediction of such down time based on metrics (operational data) of the underlying and surrounding infrastructure. Such prediction would be invaluable in the context of automated Fog offloading and adaptive decision making in Fog orchestration. Models that cater for four container-based stateless and stateful offload techniques, namely Save and Load, Export and Import, Push and Pull and Live Migration, are built using four (linear and non-linear) regression techniques. Experimental results comprising over 42 million data points from multiple lab-based Fog infrastructure are presented. The results highlight that reasonably accurate predictions (measured by the coefficient of determination for regression models, mean absolute percentage error, and mean absolute error) may be obtained when considering 25 metrics relevant to the infrastructure., Comment: arXiv admin note: substantial text overlap with arXiv:1909.04945

Fog computing is an emerging paradigm that aims to meet the increasing computation demands arising from the billions of devices connected to the Internet. Offloading services of an application from the Cloud to the edge of the network can improve the overall Quality-of-Service (QoS) of the application since it can process data closer to user devices. Diverse Fog nodes ranging from Wi-Fi routers to mini-clouds with varying resource capabilities makes it challenging to determine which services of an application need to be offloaded. In this paper, a context-aware mechanism for distributing applications across the Cloud and the Fog is proposed. The mechanism dynamically generates (re)deployment plans for the application to maximise the performance efficiency of the application by taking the QoS and running costs into account. The mechanism relies on deep Q-networks to generate a distribution plan without prior knowledge of the available resources on the Fog node, the network condition and the application. The feasibility of the proposed context-aware distribution mechanism is demonstrated on two use-cases, namely a face detection application and a location-based mobile game. The benefits are increased utility of dynamic distribution in both use cases, when compared to a static distribution approach used in existing research.

Scheduling is important in Edge computing. In contrast to the Cloud, Edge resources are hardware limited and cannot support workload-driven infrastructure scaling. Hence, resource allocation and scheduling for the Edge requires a fresh perspective. Existing Edge scheduling research assumes availability of all needed resources whenever a job request is made. This paper challenges that assumption, since not all job requests from a Cloud server can be scheduled on an Edge node. Thus, guaranteeing fairness among the clients (Cloud servers offloading jobs) while accounting for priorities of the jobs becomes a critical task. This paper presents four scheduling techniques, the first is a naive first come first serve strategy and further proposes three strategies, namely a client fair, priority fair, and hybrid that accounts for the fairness of both clients and job priorities. An evaluation on a target platform under three different scenarios, namely equal, random, and Gaussian job distributions is presented. The experimental studies highlight the low overheads and the distribution of scheduled jobs on the Edge node when compared to the naive strategy. The results confirm the superior performance of the hybrid strategy and showcase the feasibility of fair schedulers for Edge computing., Comment: 10 pages; accepted to IEEE Int. Conf. on Fog and Edge Computing (ICFEC), 2020

Fog computing offloads latency critical application services running on the Cloud in close proximity to end-user devices onto resources located at the edge of the network. The research in this paper is motivated towards characterising and estimating the time taken to offload a service using containers, which is investigated in the context of the `Save and Load' container migration technique. To this end, the research addresses questions such as whether fog offloading can be accurately modelled and which system and network related parameters influence offloading. These are addressed by exploring a catalogue of 21 different metrics both at the system and process levels that is used as input to four estimation techniques using collective model and individual models to predict the time taken for offloading. The study is pursued by collecting over 1.1 million data points and the preliminary results indicate that offloading can be modelled accurately.

Fog computing envisions that deploying services of an application across resources in the cloud and those located at the edge of the network may improve the overall performance of the application when compared to running the application on the cloud. However, there are currently no benchmarks that can directly compare the performance of the application across the cloud-only, edge-only and cloud-edge deployment platform to obtain any insight on performance improvement. This paper proposes DeFog, a first Fog benchmarking suite to: (i) alleviate the burden of Fog benchmarking by using a standard methodology, and (ii) facilitate the understanding of the target platform by collecting a catalogue of relevant metrics for a set of benchmarks. The current portfolio of DeFog benchmarks comprises six relevant applications conducive to using the edge. Experimental studies are carried out on multiple target platforms to demonstrate the use of DeFog for collecting metrics related to application latencies (communication and computation), for understanding the impact of stress and concurrent users on application latencies, and for understanding the performance of deploying different combination of services of an application across the cloud and edge. DeFog is available for public download (https://github.com/qub-blesson/DeFog)., Comment: Accepted to the ACM/IEEE Symposium on Edge Computing, 2019, Washington DC, USA

The Cloud has become integral to most Internet-based applications and user gadgets. This article provides a brief history of the Cloud and presents a researcher's view of the prospects for innovating at the infrastructure, middleware, and application and delivery levels of the already crowded Cloud computing stack., Comment: Accepted to IEEE Computer, 2019

The edge of the network has the potential to host services for supporting a variety of user applications, ranging in complexity from data preprocessing, image and video rendering, and interactive gaming, to embedded systems in autonomous cars and built environments. However, the computational and data resources over which such services are hosted, and the actors that interact with these services, have an intermittent availability and access profile, introducing significant risk for user applications that must rely on them. This article investigates the development of an edge marketplace, which is able to support multiple providers for offering services at the network edge, and to enable demand supply for influencing the operation of such a marketplace. Resilience, cost, and quality of service and experience will subsequently enable such a marketplace to adapt its services over time. This article also describes how distributed-ledger technologies (such as blockchains) provide a promising approach to support the operation of such a marketplace and regulate its behavior (such as the GDPR in Europe) and operation. Two application scenarios provide context for the discussion of how such a marketplace would function and be utilized in practice., Comment: Published in IEEE Cloud Computing, Volume 5, Issue 6, 2018, pp. 9-20

Hardware accelerators are available on the Cloud for enhanced analytics. Next generation Clouds aim to bring enhanced analytics using accelerators closer to user devices at the edge of the network for improving Quality-of-Service by minimizing end-to-end latencies and response times. The collective computing model that utilizes resources at the Cloud-Edge continuum in a multi-tier hierarchy comprising the Cloud, the Edge and user devices is referred to as Fog computing. This article identifies challenges and opportunities in making accelerators accessible at the Edge. A holistic view of the Fog architecture is key to pursuing meaningful research in this area., Comment: IEEE Cloud Computing magazine

Contrary to using distant and centralized cloud data center resources, employing decentralized resources at the edge of a network for processing data closer to user devices, such as smartphones and tablets, is an upcoming computing paradigm, referred to as fog/edge computing. Fog/edge resources are typically resource-constrained, heterogeneous, and dynamic compared to the cloud, thereby making resource management an important challenge that needs to be addressed. This article reviews publications as early as 1991, with 85% of the publications between 2013-2018, to identify and classify the architectures, infrastructure, and underlying algorithms for managing resources in fog/edge computing., Comment: 22 pages

Multi-tenancy in resource-constrained environments is a key challenge in Edge computing. In this paper, we develop 'DYVERSE: DYnamic VERtical Scaling in Edge' environments, which is the first light-weight and dynamic vertical scaling mechanism for managing resources allocated to applications for facilitating multi-tenancy in Edge environments. To enable dynamic vertical scaling, one static and three dynamic priority management approaches that are workload-aware, community-aware and system-aware, respectively are proposed. This research advocates that dynamic vertical scaling and priority management approaches reduce Service Level Objective (SLO) violation rates. An online-game and a face detection workload in a Cloud-Edge test-bed are used to validate the research. The merits of DYVERSE is that there is only a sub-second overhead per Edge server when 32 Edge servers are deployed on a single Edge node. When compared to executing applications on the Edge servers without dynamic vertical scaling, static priorities and dynamic priorities reduce SLO violation rates of requests by up to 4% and 12% for the online game, respectively, and in both cases 6% for the face detection workload. Moreover, for both workloads, the system-aware dynamic vertical scaling method effectively reduces the latency of non-violated requests, when compared to other methods.

This book chapter considers how Edge deployments can be brought to bear in a global context by federating them across multiple geographic regions to create a global Edge-based fabric that decentralizes data center computation. This is currently impractical, not only because of technical challenges, but is also shrouded by social, legal and geopolitical issues. In this chapter, we discuss two key challenges - networking and management in federating Edge deployments. Additionally, we consider resource and modeling challenges that will need to be addressed for a federated Edge., Comment: Book Chapter accepted to the Fog and Edge Computing: Principles and Paradigms; Editors Buyya, Srirama