2023 IEEE International Conference on Edge Computing and Communications (EDGE)最新文献

{"title":"The Potentials of AI Planning on the Edge","authors":"Ilche Georgievski, Marco Aiello","doi":"10.1109/EDGE60047.2023.00055","DOIUrl":"https://doi.org/10.1109/EDGE60047.2023.00055","url":null,"abstract":"Edge computing brings computation closer to sources of data and knowledge by embedding computation in the physical space and close to the end users. Edge computing is becoming the ultimate platform where modern applications based on IoT and AI are deployed in a truly distributed manner. When edge applications require goal-oriented behaviour, AI planning comes into play as a powerful tool for achieving such behaviour. In turn, this necessitates AI planning systems that can be deployed and operate on the edge possibly on a multitude of dispersed nodes. Current approaches to distributed AI planning are mainly designed around the requirements and peculiarities of multi-agent systems, such as communication constraints and the self-interest of agents. In this work, we postulate that edge computing provides new perspectives for distributing AI planning. We propose the concept of edge AI planning where multiple AI planning components are distributed on edge nodes and communicate over a vast network. These components need to have clearly defined requirements of what can be distributed and how in order for the overall AI planning to work effectively, in turn enabling correct and consistent executions across the whole system.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"136 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114663476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Perception Workload Characterization and Prediction on the Edges with Memory Contention for Connected Autonomous Vehicles","authors":"Sihai Tang, Shengze Wang, Song Fu, Qing Yang","doi":"10.1109/edge60047.2023.00026","DOIUrl":"https://doi.org/10.1109/edge60047.2023.00026","url":null,"abstract":"Vehicular Edge computing requires computational power from connected Edge devices in the network to process incoming vehicle work requests. This connection and offloading allows for faster and more efficient data processing and thus improves the safety, performance, and reliability of the connected vehicles. Existing works focus on the processor and its characterization, but they forgo the connecting components. Memory resource and storage resource is limited on Edge devices, and the two combined incur a heavy impact on deep learning. This is prominent as perception-based workloads have yet to be studied deeply. In our characterization, we have found that memory contention can be split into 3 behaviors. Each of these behaviors interacts with the other resources differently. Then, in our deep neural network (DNN) layer analysis, we find several layers that see computation time increases of over 2849% for convolutional layers and 1173.34% for activation layers. Through the characterization, we can model the workload behavior for the Edge based on the device configuration and the workload requirements. Through this, the impacts of memory contention and its impacts are quantified. To the best of our knowledge, this is the first such work that characterizes the memory impacts towards vehicular edge computational workloads with a deep focus on memory and DNN layers.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129510064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Seque: Lean and Energy-aware Data Management for IoT Gateways","authors":"Pierre-Louis Sixdenier, S. Wildermann, Martin Ottens, Jürgen Teich","doi":"10.1109/EDGE60047.2023.00030","DOIUrl":"https://doi.org/10.1109/EDGE60047.2023.00030","url":null,"abstract":"IoT systems with multiple deployed sensor nodes often use gateways to gather, fuse, transform and transmit diverse data acquired from the sensor nodes, e.g., to a cloud server. When being deployed in remote environments, not only the memory and storage, but also energy can be scarce and supply be time-dependent and often unpredictable, e.g. when obtained by energy harvesting. In this realm, this paper proposes a lean and energy-aware methodology called Seque for data management for such gateways. Rather than processing multiple sensor requests at a time and being unconscious of the level of available energy, Seque schedules only one request at a time. Moreover, Seque dynamically decides whether to directly process and transmit data of a request to a cloud server, or alternatively compress and persist data locally on the gateway in expectation of a power failure to postpone the upload to time of recovery from a power shortage. With this scheduling technique, a guarantee can be given that no sensor request admitted will suffer from partial or full loss of data. A reference implementation of Seque is provided with scheduling decisions being calibrated based on energy models of sensor interfaces, CPU system and upload interfaces of a real embedded gateway platform. Presented analysis on whether energy can be sent most by selective compression of data. Finally, the lightweight approach is evaluated in terms of energy consumption, and storage footprint and compared with commercially available database management systems including MongoDB and SQLite. The evaluation shows that Seque provides on average between 51% and 63% lower energy consumption for different data schemas per sensor request and also between 63% and 78% of lower storage requirements, pronouncing its leanness.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122455463","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"iEDGE 2023 Committees","authors":"","doi":"10.1109/edge60047.2023.00013","DOIUrl":"https://doi.org/10.1109/edge60047.2023.00013","url":null,"abstract":"","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123525476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"COGNIT: Challenges and Vision for a Serverless and Multi-Provider Cognitive Cloud-Edge Continuum","authors":"P. Townend, A. P. Martí, Idoia De-la-Iglesia, N. Matskanis, Thomas Ohlson Timoudas, Torsten Hallmann, Antonio Lalaguna, Kaja Swat, Francesco Renzi, Dominik Bocheński, Marco Mancini, M. Bhuyan, Marco González-Hierro, S. Dupont, Johan Kristiansson, R. Montero, E. Elmroth, Ivá Valdés, Philippe Massonet, Daniel Olsson, I. Llorente, Per-Olov Östberg, Michael Abdou","doi":"10.1109/EDGE60047.2023.00015","DOIUrl":"https://doi.org/10.1109/EDGE60047.2023.00015","url":null,"abstract":"Use of the serverless paradigm in cloud application development is growing rapidly, primarily driven by its promise to free developers from the responsibility of provisioning, operating, and scaling the underlying infrastructure. However, modern cloud-edge infrastructures are characterized by large numbers of disparate providers, constrained resource devices, platform heterogeneity, infrastructural dynamicity, and the need to orchestrate geographically distributed nodes and devices over public networks. This presents significant management complexity that must be addressed if serverless technologies are to be used in production systems. This position paper introduces COGNIT, a major new European initiative aiming to integrate AI technology into cloud-edge management systems to create a Cognitive Cloud reference framework and associated tools for serverless computing at the edge. COGNIT aims to: 1) support an innovative new serverless paradigm for edge application management and enhanced digital sovereignty for users and developers; 2) enable on-demand deployment of large-scale, highly distributed and self-adaptive serverless environments using existing cloud resources; 3) optimize data placement according to changes in energy efficiency heuristics and application demands and behavior; 4) enable secure and trusted execution of serverless runtimes. We identify and discuss seven research challenges related to the integration of serverless technologies with multi-provider Edge infrastructures and present our vision for how these challenges can be solved. We introduce a high-level view of our reference architecture for serverless cloud-edge continuum systems, and detail four motivating real-world use cases that will be used for validation, drawing from domains within Smart Cities, Agriculture and Environment, Energy, and Cybersecurity.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121631950","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Weighted Load Balancing Method for Heterogeneous Clusters on Hybrid Clouds","authors":"Keita Hagiwara, Yanzhi Li, Midori Sugaya","doi":"10.1109/EDGE60047.2023.00035","DOIUrl":"https://doi.org/10.1109/EDGE60047.2023.00035","url":null,"abstract":"In recent years, edge device and AI services have been expected to utilize scalable cloud computing to handle large amounts of processing. In the cloud, load-balancing techniques distribute the load evenly to many nodes to achieve high throughput. At the same time, the shift to hybrid cloud computing requires additional nodes with different generations or different types of computational resources to achieve high performance in an environment with heterogeneous computational performance. Heterogeneity raises concerns that the current uniform load-balancing will result in overloaded or underloaded nodes, which degrade responsiveness. Therefore, this study proposes a weighted load-balancing method to improve responsiveness in clusters with nonuniform computational performance. The proposed method is effective in improving the average response time by about 20%, the maximum response time by about 45%, and the response time variance by about 70% compared to load-balancing with a load balancer developed by Google.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121647310","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Service Classification of Network Traffic in 5G Core Networks using Machine Learning","authors":"R. Pell, M. Shojafar, Dimitrios Kosmanos, S. Moschoyiannis","doi":"10.1109/EDGE60047.2023.00053","DOIUrl":"https://doi.org/10.1109/EDGE60047.2023.00053","url":null,"abstract":"Fifth generation mobile networks (5G) leverage the power of edge computing to move vital services closer to end users. With critical 5G core network components located at the edge there is a need for detecting malicious signalling traffic to mitigate potential signalling attacks between the distributed Network Functions (NFs). A prerequisite for detecting anomalous signalling is a network traffic dataset for the identification and classification of normal traffic profiles. To this end, we utilise a 5G Core Network (5GC) simulator to execute test scenarios for different 5G procedures and use the captured network traffic to generate a dataset of normalised service interactions in the form of packet captures. We then apply machine learning techniques (supervised learning) and do a comparative analysis on accuracy, which uses three features from the traffic meta-data. Our results show that the identification of 5G service use by applying ML techniques offer a viable solution to classifying normal services from network traffic metadata alone. This has potential advantages in forecasting service demand for resource allocation in the dynamic 5GC environment and provide a baseline for performing anomaly detection of NF communication for detecting malicious traffic within the 5G Service Based Architecture (SBA).","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"109 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129198409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Comprehensive Performance Evaluation of Procedural Geometry Workloads on Resource-Constrained Devices","authors":"Edon Govori, Ilir Murturi, S. Dustdar","doi":"10.1109/EDGE60047.2023.00049","DOIUrl":"https://doi.org/10.1109/EDGE60047.2023.00049","url":null,"abstract":"In recent years, visualizing high-quality 3D content within modern applications (e.g., Augmented or Virtual Reality) is increasingly being generated procedurally rather than explicitly. This manifests in producing highly detailed geometries entailing resource-intensive computational workloads (i.e., Procedural Geometry Workloads) with particular characteristics. Typically, workloads with resource-intensive demands are executed in environments with powerful resources (i.e., the cloud). However, the enormous amount of data transmission, heterogeneous devices, and networks involved impact overall latency and quality in user-facing applications. To tackle these challenges, computing entities (i.e., edge devices) located near end-users can be utilized to generate 3D content. Our objective within this paper is to evaluate performance and power consumption when executing procedural geometric workloads on resource-constrained edge devices. Through extensive experiments, we aim to comprehend the limitations of different edge devices when generating 3D content under different configurations.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131917058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Using Machine Learning for Detection and Classification of Cyber Attacks in Edge IoT","authors":"Elena Becker, Maanak Gupta, K. Aryal","doi":"10.1109/EDGE60047.2023.00063","DOIUrl":"https://doi.org/10.1109/EDGE60047.2023.00063","url":null,"abstract":"Internet of Things (IoT) devices are omnipresent due to their ease of use and level of connectivity. Because of wide deployment, IoT network traffic security is a large issue, especially as the devices become more common at the edge of the connected ecosystem. In general, low-powered IoT devices themselves are not inherently secure, so tailored security mechanisms are needed to make the ecosystem secure. The incorporation of the cloud also adds new security issues with the cloud service provider (CSP). In addition, several smart applications necessitate deploying edge-based infrastructure due to their real-time computation and communication requirements, while also having the ability to detect and mitigate different cyber attacks and remain light-weight. In this paper, we propose a machine learning-based approach to detect and classify different edge IoT network traffic driven cyber attacks, and evaluate their strengths and weaknesses. Particularly, we will compare eleven machine learning models to determine the best security agent trained for attack detection and classification on an edge IoT cyber security dataset with fourteen different attacks. We also provide experimental evaluation and analysis of our work, followed by our conclusion.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"789 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116416019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"EDGE 2023 General Chairs Message","authors":"Feras M. Awaysheh, Anna Kobusinska","doi":"10.1109/edge60047.2023.00009","DOIUrl":"https://doi.org/10.1109/edge60047.2023.00009","url":null,"abstract":"We welcome you as General Co-Chairs to the 7th IEEE Int. Conference on Edge Computing and Communications (EDGE 2023), organised as part of the IEEE SERVICES Congress. Chicago is a wonderful city in which to host this event, a place which has played a strong role in the social, cultural, economic, and political history of the US. Situated not far from the location of this event are some of the leading Universities in the US (University of Chicago, University of Illinois), the US Dept. of Energy national lab: Argonne National Lab and the Chicago stock exchange. Interestingly, edge computing plays an important part in all these institutions, including in the urban infrastructure that involves the monitoring of air quality and traffic within Chicago. Edge computing enables integration of data from sensing infrastructure, including specialist mobile apps that are used by the citizens of Chicago daily. Understanding how this data can be processed and used to facilitate the next generation of applications is a key theme within this conference and the wider SERVICES congress.","PeriodicalId":369407,"journal":{"name":"2023 IEEE International Conference on Edge Computing and Communications (EDGE)","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114843364","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}