In MEC, because the traffic distribution in MEC is heterogeneous and dynamic, it is difficult for an individual edge server to provide satisfactory computation service anytime and anywhere. These issues motivated the researchers to study the cooperation among edge servers. The previous works have disadvantages since the cooperated region in these algorithms is limited within one-hop. The performance of MEC can be improved further by releasing the restriction of cooperation region, which is challenging. Therefore, we propose a new decentralized and region-free edge server cooperation approach for task offloading in MEC, named CoTree. In CoTree, the cooperation region is not limited. Each server forms its own basic cooperation unit and calculates its announced capability based on the BCU. The server's capability, the processing delay, the task and calculation result forwarding delay are considered during the calculation. The task division strategy bases on the real capability of host-server and the announced capability of cooperation-servers. This cooperation process is recursive and will be terminated once the terminal condition is satisfied. The simulation results demonstrate the advantages of CoTree over previous works.
{"title":"CoTree: Region-free and Decentralized Edge Server Cooperation","authors":"Ning Li, Xin Yuan, Zhaoxin Zhang","doi":"10.1145/3526059.3533616","DOIUrl":"https://doi.org/10.1145/3526059.3533616","url":null,"abstract":"In MEC, because the traffic distribution in MEC is heterogeneous and dynamic, it is difficult for an individual edge server to provide satisfactory computation service anytime and anywhere. These issues motivated the researchers to study the cooperation among edge servers. The previous works have disadvantages since the cooperated region in these algorithms is limited within one-hop. The performance of MEC can be improved further by releasing the restriction of cooperation region, which is challenging. Therefore, we propose a new decentralized and region-free edge server cooperation approach for task offloading in MEC, named CoTree. In CoTree, the cooperation region is not limited. Each server forms its own basic cooperation unit and calculates its announced capability based on the BCU. The server's capability, the processing delay, the task and calculation result forwarding delay are considered during the calculation. The task division strategy bases on the real capability of host-server and the announced capability of cooperation-servers. This cooperation process is recursive and will be terminated once the terminal condition is satisfied. The simulation results demonstrate the advantages of CoTree over previous works.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129539305","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}
point clouds, volumetric data. The process of delivering these assets from an hard-disk to a remote screen is subjected to a set of physical and logistical constraints: storage, bandwidth, latency, memory, computational power. A set of techniques can be used to work around these constraints or to trade one resource for another, in these talk we will try to paint a general picture of this framework and recall some of the main developments in technology and algorithms that let to the present situation.
{"title":"Streaming 3D Content","authors":"F. Ponchio","doi":"10.1145/3526059.3533615","DOIUrl":"https://doi.org/10.1145/3526059.3533615","url":null,"abstract":"point clouds, volumetric data. The process of delivering these assets from an hard-disk to a remote screen is subjected to a set of physical and logistical constraints: storage, bandwidth, latency, memory, computational power. A set of techniques can be used to work around these constraints or to trade one resource for another, in these talk we will try to paint a general picture of this framework and recall some of the main developments in technology and algorithms that let to the present situation.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127722203","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}
S. Bano, E. Carlini, P. Cassará, M. Coppola, Patrizio Dazzi, A. Gotta
Multi-Access Edge Computing (MEC) is attracting a lot of interest because it complements cloud-based approaches. Indeed, MEC is opening up in the direction of reducing both interaction delays and data sharing, called Cyber-Physical Systems (CPSs). In the near fu- ture, edge technologies will be a fundamental tool to better support time-dependent and data-intensive applications. In this context, this work explores existing and emerging platforms for MEC and human-centric applications, and proposes a suitable architecture that can be used in the context of autonomous vehicle systems.The proposed architecture will support scalable communication among sensing devices and edge/cloud computing platforms, as well as orchestrate services for computing, storage, and learning with the use of an Information-centric paradigm such as Apache Kafka
{"title":"A Novel Approach to Distributed Model Aggregation using Apache Kafka","authors":"S. Bano, E. Carlini, P. Cassará, M. Coppola, Patrizio Dazzi, A. Gotta","doi":"10.1145/3526059.3533621","DOIUrl":"https://doi.org/10.1145/3526059.3533621","url":null,"abstract":"Multi-Access Edge Computing (MEC) is attracting a lot of interest because it complements cloud-based approaches. Indeed, MEC is opening up in the direction of reducing both interaction delays and data sharing, called Cyber-Physical Systems (CPSs). In the near fu- ture, edge technologies will be a fundamental tool to better support time-dependent and data-intensive applications. In this context, this work explores existing and emerging platforms for MEC and human-centric applications, and proposes a suitable architecture that can be used in the context of autonomous vehicle systems.The proposed architecture will support scalable communication among sensing devices and edge/cloud computing platforms, as well as orchestrate services for computing, storage, and learning with the use of an Information-centric paradigm such as Apache Kafka","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121831767","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}
E. Carlini, Hanna Kavalionak, Patrizio Dazzi, L. Ferrucci, M. Coppola, M. Mordacchini
Edge computing promises to bring computation and storage close to end-users, opening exciting new areas of improvement for applications with a high level of interactivity and requiring low latency. However, these improvements require careful scheduling of applications in the correct Edge resource. This decision is generally taken by considering multiple parameters, including the network capabilities. In this paper, we discuss an approach that measures latency and bandwidth between multiple clients and Edge servers. The approach is based on recent Serverless computing technologies, and it is meant as a support to take timely and correct scheduling decisions in the Edge. We also provide the description of a proof of concept implementation of the said approach.
{"title":"Network Measurements with Function-as-a-Service for Distributed Low-latency Edge Applications","authors":"E. Carlini, Hanna Kavalionak, Patrizio Dazzi, L. Ferrucci, M. Coppola, M. Mordacchini","doi":"10.1145/3526059.3533622","DOIUrl":"https://doi.org/10.1145/3526059.3533622","url":null,"abstract":"Edge computing promises to bring computation and storage close to end-users, opening exciting new areas of improvement for applications with a high level of interactivity and requiring low latency. However, these improvements require careful scheduling of applications in the correct Edge resource. This decision is generally taken by considering multiple parameters, including the network capabilities. In this paper, we discuss an approach that measures latency and bandwidth between multiple clients and Edge servers. The approach is based on recent Serverless computing technologies, and it is meant as a support to take timely and correct scheduling decisions in the Edge. We also provide the description of a proof of concept implementation of the said approach.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131594244","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}
T. Theodoropoulos, Antonios Makris, John Violos, K. Tserpes
The very fabric of Edge Computing is intertwined with the necessity to be able to orchestrate and manage a huge number of heterogeneous computational resources. On top of that, the rather demanding Quality of Service (QoS) requirements of Internet of Things (IoT) applications that run on these resources, dictate that it is essential to establish robust Fault Tolerance mechanisms. These mechanisms should be able to guarantee that the requirements will be upheld regardless of any potential changes in task production rate. To that end, we suggest an Automated Pipeline for Advanced Fault Tolerance (APAFT) that consists of various components that are designed to operate as functional blocks of an automated closed-control loop. Furthermore, the suggested pipeline is able to carry out the various Horizontal Scaling operations in a proactive manner. These Proactive Scaling capabilities are achieved via the use of a dedicated Deep Learning (DL)-based component that is able to perform multi-step prediction. Our work aims to introduce a number of mechanisms that are able to leverage the benefits that are provided by the multi-step format in a more refined manner. Having access to information regarding multiple future instances allows us to design automated resource orchestration strategies that cater to the specific characteristics of each type of computational node that is part of the Edge Infrastructure.
{"title":"An Automated Pipeline for Advanced Fault Tolerance in Edge Computing Infrastructures","authors":"T. Theodoropoulos, Antonios Makris, John Violos, K. Tserpes","doi":"10.1145/3526059.3533623","DOIUrl":"https://doi.org/10.1145/3526059.3533623","url":null,"abstract":"The very fabric of Edge Computing is intertwined with the necessity to be able to orchestrate and manage a huge number of heterogeneous computational resources. On top of that, the rather demanding Quality of Service (QoS) requirements of Internet of Things (IoT) applications that run on these resources, dictate that it is essential to establish robust Fault Tolerance mechanisms. These mechanisms should be able to guarantee that the requirements will be upheld regardless of any potential changes in task production rate. To that end, we suggest an Automated Pipeline for Advanced Fault Tolerance (APAFT) that consists of various components that are designed to operate as functional blocks of an automated closed-control loop. Furthermore, the suggested pipeline is able to carry out the various Horizontal Scaling operations in a proactive manner. These Proactive Scaling capabilities are achieved via the use of a dedicated Deep Learning (DL)-based component that is able to perform multi-step prediction. Our work aims to introduce a number of mechanisms that are able to leverage the benefits that are provided by the multi-step format in a more refined manner. Having access to information regarding multiple future instances allows us to design automated resource orchestration strategies that cater to the specific characteristics of each type of computational node that is part of the Edge Infrastructure.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115359098","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}
Lorenzo Blasi, Andrea Toro, M. Girolamo, K. Tserpes
We assume a computing infrastructure that implements a cloud-to-edge continuum by federating heterogeneous resources. This infrastructure enables next generation (nextgen) QoE-intensive applications, by facilitating one of their critical requirements: latency. As such each node is identified primarily in terms of latency and secondarily in terms of, more traditional, computational capacity and availability characteristics. Since this infrastructure is dependent on the existing, local, and potentially third-party owned resources, we must address the pertinent problem of resource heterogeneity. This paper delineates the technical specifications of the solution that was designed to tackle the problem, creating a software stack that comprises a complete environment for the orchestration of containerized and VM-based workflows. The solution, dubbed "Minicloud", allows the underlying resource to be federated from a higher-level continuum management layer and be used for the deployment of nextgen apps based on latency requirements. Based on lightweight Kubernetes-oriented technologies and tools, it maintains minimum dependencies allowing a wide range of underlying OSs and system architectures to be supported.
{"title":"A Minicloud Specification Enabling the Federation of Heterogeneous Edge Resources for Latency Sensitive Applications' Requirements","authors":"Lorenzo Blasi, Andrea Toro, M. Girolamo, K. Tserpes","doi":"10.1145/3526059.3533619","DOIUrl":"https://doi.org/10.1145/3526059.3533619","url":null,"abstract":"We assume a computing infrastructure that implements a cloud-to-edge continuum by federating heterogeneous resources. This infrastructure enables next generation (nextgen) QoE-intensive applications, by facilitating one of their critical requirements: latency. As such each node is identified primarily in terms of latency and secondarily in terms of, more traditional, computational capacity and availability characteristics. Since this infrastructure is dependent on the existing, local, and potentially third-party owned resources, we must address the pertinent problem of resource heterogeneity. This paper delineates the technical specifications of the solution that was designed to tackle the problem, creating a software stack that comprises a complete environment for the orchestration of containerized and VM-based workflows. The solution, dubbed \"Minicloud\", allows the underlying resource to be federated from a higher-level continuum management layer and be used for the deployment of nextgen apps based on latency requirements. Based on lightweight Kubernetes-oriented technologies and tools, it maintains minimum dependencies allowing a wide range of underlying OSs and system architectures to be supported.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"165 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133053516","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}
Antonios Makris, E. Psomakelis, T. Theodoropoulos, K. Tserpes
Due to the continuous development of Internet of Things (IoT), the volume of the data these devices generate are expected to grow dramatically in the future. As a result, managing and processing such massive data amounts at the edge becomes a vital issue. Edge computing moves data and computation closer to the client enabling latency- and bandwidth-sensitive applications, that would not be feasible using cloud and remote processing alone. Nevertheless, implementing an efficient edge-enabled storage system is challenging due to the distributed and heterogeneous nature of the edge and its limited resource capabilities. To this end, we propose a lightweight hybrid distributed edge/cloud storage framework which aims to improve the Quality of Experience (QoE) of the end-users by migrating data close to them, thus reducing data transfers delays and network utilization. The proposed edge storage component (ESC) exploits the Dynamic Lifecycle Framework, in order to enable transparent and automated access for containerized applications to remote workloads. The effectiveness of the ESC is evaluated by employing a number of resource utilization and Quality of Service (QoS) metrics.
{"title":"Towards a Distributed Storage Framework for Edge Computing Infrastructures","authors":"Antonios Makris, E. Psomakelis, T. Theodoropoulos, K. Tserpes","doi":"10.1145/3526059.3533617","DOIUrl":"https://doi.org/10.1145/3526059.3533617","url":null,"abstract":"Due to the continuous development of Internet of Things (IoT), the volume of the data these devices generate are expected to grow dramatically in the future. As a result, managing and processing such massive data amounts at the edge becomes a vital issue. Edge computing moves data and computation closer to the client enabling latency- and bandwidth-sensitive applications, that would not be feasible using cloud and remote processing alone. Nevertheless, implementing an efficient edge-enabled storage system is challenging due to the distributed and heterogeneous nature of the edge and its limited resource capabilities. To this end, we propose a lightweight hybrid distributed edge/cloud storage framework which aims to improve the Quality of Experience (QoE) of the end-users by migrating data close to them, thus reducing data transfers delays and network utilization. The proposed edge storage component (ESC) exploits the Dynamic Lifecycle Framework, in order to enable transparent and automated access for containerized applications to remote workloads. The effectiveness of the ESC is evaluated by employing a number of resource utilization and Quality of Service (QoS) metrics.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132481137","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}
The highly dynamic and heterogeneous environment that characterizes the edge of the Cloud/Edge Continuum calls for new intelligent methods for tackling the needs of such a complex scenario. In particular, adaptive and self-organizing decentralized solutions have been advanced for optimizing the placement of applications at the Edge. In this paper, we propose a probabilistic mathematical model that allows to describe one of such solutions. The goal of the model is twofold: i) to make it possible to demonstrate the convergence of the proposed solution; ii) to study the impact of the self-organizing solution without the need of an actual implementation or simulation of the system, allowing to evaluate the suitability of the solution in specific contexts. The paper presents the mathematical formulation of the proposed solution as well as the validation of the proposed model against a simulation of the system.
{"title":"A Mathematical Model for Latency Constrained Self-Organizing Application Placement in the Edge","authors":"M. Mordacchini, E. Carlini, Patrizio Dazzi","doi":"10.1145/3526059.3533620","DOIUrl":"https://doi.org/10.1145/3526059.3533620","url":null,"abstract":"The highly dynamic and heterogeneous environment that characterizes the edge of the Cloud/Edge Continuum calls for new intelligent methods for tackling the needs of such a complex scenario. In particular, adaptive and self-organizing decentralized solutions have been advanced for optimizing the placement of applications at the Edge. In this paper, we propose a probabilistic mathematical model that allows to describe one of such solutions. The goal of the model is twofold: i) to make it possible to demonstrate the convergence of the proposed solution; ii) to study the impact of the self-organizing solution without the need of an actual implementation or simulation of the system, allowing to evaluate the suitability of the solution in specific contexts. The paper presents the mathematical formulation of the proposed solution as well as the validation of the proposed model against a simulation of the system.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"211 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115053643","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}
Jämes Ménétrey, Marcelo Pasin, P. Felber, V. Schiavoni
Over the last decade, the cloud computing landscape has transformed from centralised architecture made of large data centres to a distributed and heterogeneous architecture embracing edge and IoT processing units. This shift has created the so-called cloud-edge continuum, which closes the gap between the large datacentres and the end-user devices. Existing solutions are, however, dominated by proprietary silos and incompatible technologies, built around dedicated devices and run-time stacks. In this position paper, we motivate the need for interoperable solutions that would run seamlessly across hardware devices and software environments, while achieving good performance and a high level of security-a critical requirement for code and data processed off-premises. We argue that the technology provided by WebAssembly running on modern virtual machines and shielded within trusted execution environments, combined with a core set of services and support libraries, allows us to meet both goals. We also present preliminary results from a prototype built with these technologies and deployed on the cloud-edge continuum.
{"title":"WebAssembly as a Common Layer for the Cloud-edge Continuum","authors":"Jämes Ménétrey, Marcelo Pasin, P. Felber, V. Schiavoni","doi":"10.1145/3526059.3533618","DOIUrl":"https://doi.org/10.1145/3526059.3533618","url":null,"abstract":"Over the last decade, the cloud computing landscape has transformed from centralised architecture made of large data centres to a distributed and heterogeneous architecture embracing edge and IoT processing units. This shift has created the so-called cloud-edge continuum, which closes the gap between the large datacentres and the end-user devices. Existing solutions are, however, dominated by proprietary silos and incompatible technologies, built around dedicated devices and run-time stacks. In this position paper, we motivate the need for interoperable solutions that would run seamlessly across hardware devices and software environments, while achieving good performance and a high level of security-a critical requirement for code and data processed off-premises. We argue that the technology provided by WebAssembly running on modern virtual machines and shielded within trusted execution environments, combined with a core set of services and support libraries, allows us to meet both goals. We also present preliminary results from a prototype built with these technologies and deployed on the cloud-edge continuum.","PeriodicalId":351705,"journal":{"name":"Proceedings of the 2nd Workshop on Flexible Resource and Application Management on the Edge","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131723648","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：