The most commonly seen things on streets in any city are vehicles. However, most of them are used to transport people or goods. What if they also carry resources and capabilities for sensing, communications, computing, storage, and intelligence (SCCSI)? We will have a web of sensors to monitor the city, a network of powerful communicators to transport data around, a grid of computing power to conduct data analytics and machine learning (ML), a network of distributed storage to buffer/cache data/job for optimization, and a set of movable AI/ML toolboxes made available for specialized smart applications. This perspective article presents how to leverage SCCSI-empowered vehicles to design such a service network, simply called SCCSI network, to help build a smart city with a cost-effective and sustainable solution. It showcases how multi-dimensional technologies, namely, sensing, communications, computing, storage, and intelligence, converge to a unifying technology to solve grand challenges for resource demands from emerging large-scale applications. Thus, with SCCSI-empowered vehicles on the ground, over the air, and on the sea, SCCSI network can make resources and capabilities on the move, practically pushing SCCSI services to the edge! We hope this article serves as a spark to stimulate more disruptive thinking to address grand challenges of paramount importance.
{"title":"Resources on the Move for Smart City: A Disruptive Perspective on the Grand Convergence of Sensing, Communications, Computing, Storage, and Intelligence","authors":"Yuguang Fang, Yiqin Deng, Xianhao Chen","doi":"arxiv-2409.09417","DOIUrl":"https://doi.org/arxiv-2409.09417","url":null,"abstract":"The most commonly seen things on streets in any city are vehicles. However,\u0000most of them are used to transport people or goods. What if they also carry\u0000resources and capabilities for sensing, communications, computing, storage, and\u0000intelligence (SCCSI)? We will have a web of sensors to monitor the city, a\u0000network of powerful communicators to transport data around, a grid of computing\u0000power to conduct data analytics and machine learning (ML), a network of\u0000distributed storage to buffer/cache data/job for optimization, and a set of\u0000movable AI/ML toolboxes made available for specialized smart applications. This\u0000perspective article presents how to leverage SCCSI-empowered vehicles to design\u0000such a service network, simply called SCCSI network, to help build a smart city\u0000with a cost-effective and sustainable solution. It showcases how\u0000multi-dimensional technologies, namely, sensing, communications, computing,\u0000storage, and intelligence, converge to a unifying technology to solve grand\u0000challenges for resource demands from emerging large-scale applications. Thus,\u0000with SCCSI-empowered vehicles on the ground, over the air, and on the sea,\u0000SCCSI network can make resources and capabilities on the move, practically\u0000pushing SCCSI services to the edge! We hope this article serves as a spark to\u0000stimulate more disruptive thinking to address grand challenges of paramount\u0000importance.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"29 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260562","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}
We present a peer-to-peer (P2P) live-streaming architecture designed to address challenges such as free-riding, malicious peers, churn, and network instability through the integration of a reputation system. The proposed algorithm incentivizes active peer participation while discouraging opportunistic behaviors, with a reputation mechanism that rewards altruistic peers and penalizes free riders and malicious actors. To manage peer dynamics, the algorithm continuously updates the strategies and adjusts to changing neighbors. It also implements a request-to-join mechanism for flash crowd scenarios, allowing the source node to delegate requests to child nodes, forming an interconnected tree structure that efficiently handles high demand and maintains system stability. The decentralized reputation mechanism promotes long-term sustainability in the P2P live streaming system.
{"title":"Reputation-Driven Peer-to-Peer Live Streaming Architecture for Preventing Free-Riding","authors":"Rashmi Kushwaha, Rahul Bhattacharyya, Yatindra Nath Singh","doi":"arxiv-2409.09329","DOIUrl":"https://doi.org/arxiv-2409.09329","url":null,"abstract":"We present a peer-to-peer (P2P) live-streaming architecture designed to\u0000address challenges such as free-riding, malicious peers, churn, and network\u0000instability through the integration of a reputation system. The proposed\u0000algorithm incentivizes active peer participation while discouraging\u0000opportunistic behaviors, with a reputation mechanism that rewards altruistic\u0000peers and penalizes free riders and malicious actors. To manage peer dynamics,\u0000the algorithm continuously updates the strategies and adjusts to changing\u0000neighbors. It also implements a request-to-join mechanism for flash crowd\u0000scenarios, allowing the source node to delegate requests to child nodes,\u0000forming an interconnected tree structure that efficiently handles high demand\u0000and maintains system stability. The decentralized reputation mechanism promotes\u0000long-term sustainability in the P2P live streaming system.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"74 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260232","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}
Distributed applications based on micro-services in edge computing are becoming increasingly popular due to the rapid evolution of mobile networks. While Kubernetes is the default framework when it comes to orchestrating and managing micro-service-based applications in mobile networks, the requirement to run applications between multiple sites at cloud and edge poses new challenges. Since Kubernetes does not natively provide tools to abstract inter-cluster communications at the application level, inter-cluster communication in edge computing is becoming increasingly critical to the application performance. In this paper, we evaluate for the first time the impact of inter-cluster communication on edge computing performance by using three prominent, open source inter-cluster communication projects and tools, i.e., Submariner, ClusterLink and Skupper. We develop a fully open-source testbed that integrates these tools in a modular fashion, and experimentally benchmark sample applications, including the ML class of applications, on their performance running in the multi-cluster edge computing system under varying networking conditions. We experimentally analyze two classes of envisioned mobile applications, i.e., a) industrial automation, b) vehicle decision drive assist. Our results show that Submariner performs best out of the three tools in scenarios with small payloads, regardless of the underlying networking conditions or transmission direction between clusters. When sending larger data to a service, ClusterLink outperforms Submariner once the inter-node networking conditions deteriorate, which may be the case in highly mobile scenarios in edge computing. Finally, Skupper significantly outperforms others in a variety of scenarios with larger payloads.
虽然 Kubernetes 是在移动网络中协调和管理基于微服务的应用的默认框架，但在云和边缘的多个站点之间运行应用的要求带来了新的挑战。由于 Kubernetes 本身不提供在应用层抽象集群间通信的工具，因此边缘计算中的集群间通信对应用性能的影响变得越来越关键。在本文中，我们利用三个著名的开源集群间通信项目和工具（即 Submariner、ClusterLink 和 Skupper），首次评估了集群间通信对边缘计算性能的影响。我们开发了一个完全开源的测试平台，以模块化方式集成了这些工具，并对样本应用（包括 ML 类应用）在不同网络条件下在多集群边缘计算系统中的运行性能进行了实验基准测试。我们对两类设想的移动应用进行了实验分析，即 a) 工业自动化；b) 车辆决策驱动辅助。我们的结果表明，无论底层网络条件或集群间的传输方向如何，Submariner 在有效载荷较小的应用场景中都是三种工具中表现最好的。当向服务发送较大数据时，一旦节点间网络条件恶化，ClusterLink 的性能就会优于 Submariner，在边缘计算的高移动性场景中可能会出现这种情况。最后，在有效载荷较大的各种场景中，Skupper 的性能明显优于其他产品。
{"title":"Evaluating the Impact of Inter-cluster Communications in Edge Computing","authors":"Marc Michalke, Iulisloi Zacarias, Admela Jukan","doi":"arxiv-2409.09278","DOIUrl":"https://doi.org/arxiv-2409.09278","url":null,"abstract":"Distributed applications based on micro-services in edge computing are\u0000becoming increasingly popular due to the rapid evolution of mobile networks.\u0000While Kubernetes is the default framework when it comes to orchestrating and\u0000managing micro-service-based applications in mobile networks, the requirement\u0000to run applications between multiple sites at cloud and edge poses new\u0000challenges. Since Kubernetes does not natively provide tools to abstract\u0000inter-cluster communications at the application level, inter-cluster\u0000communication in edge computing is becoming increasingly critical to the\u0000application performance. In this paper, we evaluate for the first time the\u0000impact of inter-cluster communication on edge computing performance by using\u0000three prominent, open source inter-cluster communication projects and tools,\u0000i.e., Submariner, ClusterLink and Skupper. We develop a fully open-source\u0000testbed that integrates these tools in a modular fashion, and experimentally\u0000benchmark sample applications, including the ML class of applications, on their\u0000performance running in the multi-cluster edge computing system under varying\u0000networking conditions. We experimentally analyze two classes of envisioned\u0000mobile applications, i.e., a) industrial automation, b) vehicle decision drive\u0000assist. Our results show that Submariner performs best out of the three tools\u0000in scenarios with small payloads, regardless of the underlying networking\u0000conditions or transmission direction between clusters. When sending larger data\u0000to a service, ClusterLink outperforms Submariner once the inter-node networking\u0000conditions deteriorate, which may be the case in highly mobile scenarios in\u0000edge computing. Finally, Skupper significantly outperforms others in a variety\u0000of scenarios with larger payloads.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"195 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260233","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}
Yinqiu Liu, Hongyang Du, Dusit Niyato, Jiawen Kang, Zehui Xiong, Yonggang Wen, Dong In Kim
Generative AI (GenAI), exemplified by Large Language Models (LLMs) such as OpenAI's ChatGPT, is revolutionizing various fields. Central to this transformation is Data Center Networking (DCN), which not only provides the computational power necessary for GenAI training and inference but also delivers GenAI-driven services to users. This article examines an interplay between GenAI and DCNs, highlighting their symbiotic relationship and mutual advancements. We begin by reviewing current challenges within DCNs and discuss how GenAI contributes to enhancing DCN capabilities through innovations, such as data augmentation, process automation, and domain transfer. We then focus on analyzing the distinctive characteristics of GenAI workloads on DCNs, gaining insights that catalyze the evolution of DCNs to more effectively support GenAI and LLMs. Moreover, to illustrate the seamless integration of GenAI with DCNs, we present a case study on full-lifecycle DCN digital twins. In this study, we employ LLMs equipped with Retrieval Augmented Generation (RAG) to formulate optimization problems for DCNs and adopt Diffusion-Deep Reinforcement Learning (DRL) for optimizing the RAG knowledge placement strategy. This approach not only demonstrates the application of advanced GenAI methods within DCNs but also positions the digital twin as a pivotal GenAI service operating on DCNs. We anticipate that this article can promote further research into enhancing the virtuous interaction between GenAI and DCNs.
{"title":"Generative AI in Data Center Networking: Fundamentals, Perspectives, and Case Study","authors":"Yinqiu Liu, Hongyang Du, Dusit Niyato, Jiawen Kang, Zehui Xiong, Yonggang Wen, Dong In Kim","doi":"arxiv-2409.09343","DOIUrl":"https://doi.org/arxiv-2409.09343","url":null,"abstract":"Generative AI (GenAI), exemplified by Large Language Models (LLMs) such as\u0000OpenAI's ChatGPT, is revolutionizing various fields. Central to this\u0000transformation is Data Center Networking (DCN), which not only provides the\u0000computational power necessary for GenAI training and inference but also\u0000delivers GenAI-driven services to users. This article examines an interplay\u0000between GenAI and DCNs, highlighting their symbiotic relationship and mutual\u0000advancements. We begin by reviewing current challenges within DCNs and discuss\u0000how GenAI contributes to enhancing DCN capabilities through innovations, such\u0000as data augmentation, process automation, and domain transfer. We then focus on\u0000analyzing the distinctive characteristics of GenAI workloads on DCNs, gaining\u0000insights that catalyze the evolution of DCNs to more effectively support GenAI\u0000and LLMs. Moreover, to illustrate the seamless integration of GenAI with DCNs,\u0000we present a case study on full-lifecycle DCN digital twins. In this study, we\u0000employ LLMs equipped with Retrieval Augmented Generation (RAG) to formulate\u0000optimization problems for DCNs and adopt Diffusion-Deep Reinforcement Learning\u0000(DRL) for optimizing the RAG knowledge placement strategy. This approach not\u0000only demonstrates the application of advanced GenAI methods within DCNs but\u0000also positions the digital twin as a pivotal GenAI service operating on DCNs.\u0000We anticipate that this article can promote further research into enhancing the\u0000virtuous interaction between GenAI and DCNs.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"215 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260230","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}
Sunwoo Kim, Yongjun Ahn, Daeyoung Park, Byonghyo Shim
Recent advances in sensing and computer vision (CV) technologies have opened the door for the application of deep learning (DL)-based CV technologies in the realm of 6G wireless communications. For the successful application of this emerging technology, it is crucial to have a qualified vision dataset tailored for wireless applications (e.g., RGB images containing wireless devices such as laptops and cell phones). An aim of this paper is to propose a large-scale vision dataset referred to as Vision Objects for Millimeter and Terahertz Communications (VOMTC). The VOMTC dataset consists of 20,232 pairs of RGB and depth images obtained from a camera attached to the base station (BS), with each pair labeled with three representative object categories (person, cell phone, and laptop) and bounding boxes of the objects. Through experimental studies of the VOMTC datasets, we show that the beamforming technique exploiting the VOMTC-trained object detector outperforms conventional beamforming techniques.
{"title":"VOMTC: Vision Objects for Millimeter and Terahertz Communications","authors":"Sunwoo Kim, Yongjun Ahn, Daeyoung Park, Byonghyo Shim","doi":"arxiv-2409.09330","DOIUrl":"https://doi.org/arxiv-2409.09330","url":null,"abstract":"Recent advances in sensing and computer vision (CV) technologies have opened\u0000the door for the application of deep learning (DL)-based CV technologies in the\u0000realm of 6G wireless communications. For the successful application of this\u0000emerging technology, it is crucial to have a qualified vision dataset tailored\u0000for wireless applications (e.g., RGB images containing wireless devices such as\u0000laptops and cell phones). An aim of this paper is to propose a large-scale\u0000vision dataset referred to as Vision Objects for Millimeter and Terahertz\u0000Communications (VOMTC). The VOMTC dataset consists of 20,232 pairs of RGB and\u0000depth images obtained from a camera attached to the base station (BS), with\u0000each pair labeled with three representative object categories (person, cell\u0000phone, and laptop) and bounding boxes of the objects. Through experimental\u0000studies of the VOMTC datasets, we show that the beamforming technique\u0000exploiting the VOMTC-trained object detector outperforms conventional\u0000beamforming techniques.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"17 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260229","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}
We study the problem of designing scheduling policies for communication networks. This problem is often addressed with max-weight-type approaches since they are throughput-optimal. However, max-weight policies make scheduling decisions based on the network congestion, which can be sometimes unnecessarily restrictive. In this paper, we present a ``schedule as you learn'' (SYL) approach, where we learn an average rate, and then select schedules that generate such a rate in expectation. This approach is interesting because scheduling decisions do not depend on the size of the queue backlogs, and so it provides increased flexibility to select schedules based on other criteria or rules, such as serving high-priority queues. We illustrate the results with numerical experiments for a cross-bar switch and show that, compared to max-weight, SYL can achieve lower latency to certain flows without compromising throughput optimality.
{"title":"Throughput-Optimal Scheduling via Rate Learning","authors":"Panagiotis Promponas, Víctor Valls, Konstantinos Nikolakakis, Dionysis Kalogerias, Leandros Tassiulas","doi":"arxiv-2409.09198","DOIUrl":"https://doi.org/arxiv-2409.09198","url":null,"abstract":"We study the problem of designing scheduling policies for communication\u0000networks. This problem is often addressed with max-weight-type approaches since\u0000they are throughput-optimal. However, max-weight policies make scheduling\u0000decisions based on the network congestion, which can be sometimes unnecessarily\u0000restrictive. In this paper, we present a ``schedule as you learn'' (SYL)\u0000approach, where we learn an average rate, and then select schedules that\u0000generate such a rate in expectation. This approach is interesting because\u0000scheduling decisions do not depend on the size of the queue backlogs, and so it\u0000provides increased flexibility to select schedules based on other criteria or\u0000rules, such as serving high-priority queues. We illustrate the results with\u0000numerical experiments for a cross-bar switch and show that, compared to\u0000max-weight, SYL can achieve lower latency to certain flows without compromising\u0000throughput optimality.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"23 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260234","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}
Wireless networks are increasingly facing challenges due to their expanding scale and complexity. These challenges underscore the need for advanced AI-driven strategies, particularly in the upcoming 6G networks. In this article, we introduce WirelessAgent, a novel approach leveraging large language models (LLMs) to develop AI agents capable of managing complex tasks in wireless networks. It can effectively improve network performance through advanced reasoning, multimodal data processing, and autonomous decision making. Thereafter, we demonstrate the practical applicability and benefits of WirelessAgent for network slicing management. The experimental results show that WirelessAgent is capable of accurately understanding user intent, effectively allocating slice resources, and consistently maintaining optimal performance.
{"title":"WirelessAgent: Large Language Model Agents for Intelligent Wireless Networks","authors":"Jingwen Tong, Jiawei Shao, Qiong Wu, Wei Guo, Zijian Li, Zehong Lin, Jun Zhang","doi":"arxiv-2409.07964","DOIUrl":"https://doi.org/arxiv-2409.07964","url":null,"abstract":"Wireless networks are increasingly facing challenges due to their expanding\u0000scale and complexity. These challenges underscore the need for advanced\u0000AI-driven strategies, particularly in the upcoming 6G networks. In this\u0000article, we introduce WirelessAgent, a novel approach leveraging large language\u0000models (LLMs) to develop AI agents capable of managing complex tasks in\u0000wireless networks. It can effectively improve network performance through\u0000advanced reasoning, multimodal data processing, and autonomous decision making.\u0000Thereafter, we demonstrate the practical applicability and benefits of\u0000WirelessAgent for network slicing management. The experimental results show\u0000that WirelessAgent is capable of accurately understanding user intent,\u0000effectively allocating slice resources, and consistently maintaining optimal\u0000performance.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"11 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142183981","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}
Network function virtualization leverages programmable data plane switches to deploy in-network implementable functions, to improve QoS. The memories of switches can be extended through remote direct memory access to access external memories. This paper exploits the switches external memories to place VNFs at time intervals with ultra-low latency and high bandwidth demands. The reconfiguration decision is modeled as an optimization to minimize the deployment and reconfiguration cost, while meeting the SFCs deadlines. A DRL based method is proposed to reconfigure service chains adoptable with dynamic network and traffic characteristics. To deal with slow convergence due to the complexity of deployment scenarios, static and dynamic filters are used in policy networks construction to diminish unfeasible placement exploration. Results illustrate improvement in convergence, acceptance ratio and cost.
{"title":"External Memories of PDP Switches for In-Network Implementable Functions Placement: Deep Learning Based Reconfiguration of SFCs","authors":"Somayeh Kianpisheh, Tarik Taleb","doi":"arxiv-2409.08043","DOIUrl":"https://doi.org/arxiv-2409.08043","url":null,"abstract":"Network function virtualization leverages programmable data plane switches to\u0000deploy in-network implementable functions, to improve QoS. The memories of\u0000switches can be extended through remote direct memory access to access external\u0000memories. This paper exploits the switches external memories to place VNFs at\u0000time intervals with ultra-low latency and high bandwidth demands. The\u0000reconfiguration decision is modeled as an optimization to minimize the\u0000deployment and reconfiguration cost, while meeting the SFCs deadlines. A DRL\u0000based method is proposed to reconfigure service chains adoptable with dynamic\u0000network and traffic characteristics. To deal with slow convergence due to the\u0000complexity of deployment scenarios, static and dynamic filters are used in\u0000policy networks construction to diminish unfeasible placement exploration.\u0000Results illustrate improvement in convergence, acceptance ratio and cost.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"2 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142183999","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}
Mobile Charge Scheduling for wirelessly charging nodes in Wireless Rechargeable Sensor Networks (WRSNs) is a promising but still evolving research area. Existing research mostly assumes a symmetric environment, where the routing costs in opposite directions between two locations are considered identical. However, various factors such as terrain restrictions and wind or water flows may invalidate the routing-symmetric assumption in practical environments, thereby significantly limiting the performance of these solutions in routing-asymmetric WRSNs (RA-WRSNs). To address the routing-asymmetric challenges in mobile charge scheduling for WRSNs, this paper systematically investigates the underlying Asymmetric Directional Mobile Charger (DMC) Charge Scheduling (ADMCCS) problem, aiming to minimize energy loss while satisfying the charging demands of the network nodes. The DMC model is assumed because its results can be easily applied to the specialized case of an Omnidirectional Mobile Charger (OMC). To solve the ADMCCS problem, we propose a four-step framework. First, a minimum-size efficient charging position set is selected using our designed K-means-based Charging Position Generation (KCPG) algorithm, addressing the challenge of the unlimited charging position selection space. Next, minimum-size functional-equivalent direction sets at these positions are determined using an optimal algorithm, tackling the challenge of infinite charging directions. Subsequently, the optimal energy transmission time lengths for all directions at the positions are obtained by formulating and solving a Nonlinear Program (NLP) problem. Finally, the Lin-Kernighan Heuristic (LKH) algorithm for the Asymmetric Traveling Salesman Problem is adapted to obtain a highly probable optimal loop tour, addressing the routing-asymmetric challenge.
{"title":"Directional WPT Charging for Routing-Asymmetric WRSNs with a Mobile Charger","authors":"Zhenguo Gao, Qi Zhang, Qingyu Gao, Yunlong Zhao, Hsiao-Chun Wu","doi":"arxiv-2409.07994","DOIUrl":"https://doi.org/arxiv-2409.07994","url":null,"abstract":"Mobile Charge Scheduling for wirelessly charging nodes in Wireless\u0000Rechargeable Sensor Networks (WRSNs) is a promising but still evolving research\u0000area. Existing research mostly assumes a symmetric environment, where the\u0000routing costs in opposite directions between two locations are considered\u0000identical. However, various factors such as terrain restrictions and wind or\u0000water flows may invalidate the routing-symmetric assumption in practical\u0000environments, thereby significantly limiting the performance of these solutions\u0000in routing-asymmetric WRSNs (RA-WRSNs). To address the routing-asymmetric\u0000challenges in mobile charge scheduling for WRSNs, this paper systematically\u0000investigates the underlying Asymmetric Directional Mobile Charger (DMC) Charge\u0000Scheduling (ADMCCS) problem, aiming to minimize energy loss while satisfying\u0000the charging demands of the network nodes. The DMC model is assumed because its\u0000results can be easily applied to the specialized case of an Omnidirectional\u0000Mobile Charger (OMC). To solve the ADMCCS problem, we propose a four-step\u0000framework. First, a minimum-size efficient charging position set is selected\u0000using our designed K-means-based Charging Position Generation (KCPG) algorithm,\u0000addressing the challenge of the unlimited charging position selection space.\u0000Next, minimum-size functional-equivalent direction sets at these positions are\u0000determined using an optimal algorithm, tackling the challenge of infinite\u0000charging directions. Subsequently, the optimal energy transmission time lengths\u0000for all directions at the positions are obtained by formulating and solving a\u0000Nonlinear Program (NLP) problem. Finally, the Lin-Kernighan Heuristic (LKH)\u0000algorithm for the Asymmetric Traveling Salesman Problem is adapted to obtain a\u0000highly probable optimal loop tour, addressing the routing-asymmetric challenge.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"21 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142183980","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}
Louis Van Langendonck, Ismael Castell-Uroz, Pere Barlet-Ros
Foundation models have shown great promise in various fields of study. A potential application of such models is in computer network traffic analysis, where these models can grasp the complexities of network traffic dynamics and adapt to any specific task or network environment with minimal fine-tuning. Previous approaches have used tokenized hex-level packet data and the model architecture of large language transformer models. We propose a new, efficient graph-based alternative at the flow-level. Our approach represents network traffic as a dynamic spatio-temporal graph, employing a self-supervised link prediction pretraining task to capture the spatial and temporal dynamics in this network graph framework. To evaluate the effectiveness of our approach, we conduct a few-shot learning experiment for three distinct downstream network tasks: intrusion detection, traffic classification, and botnet classification. Models finetuned from our pretrained base achieve an average performance increase of 6.87% over training from scratch, demonstrating their ability to effectively learn general network traffic dynamics during pretraining. This success suggests the potential for a large-scale version to serve as an operational foundational model.
{"title":"Towards a graph-based foundation model for network traffic analysis","authors":"Louis Van Langendonck, Ismael Castell-Uroz, Pere Barlet-Ros","doi":"arxiv-2409.08111","DOIUrl":"https://doi.org/arxiv-2409.08111","url":null,"abstract":"Foundation models have shown great promise in various fields of study. A\u0000potential application of such models is in computer network traffic analysis,\u0000where these models can grasp the complexities of network traffic dynamics and\u0000adapt to any specific task or network environment with minimal fine-tuning.\u0000Previous approaches have used tokenized hex-level packet data and the model\u0000architecture of large language transformer models. We propose a new, efficient\u0000graph-based alternative at the flow-level. Our approach represents network\u0000traffic as a dynamic spatio-temporal graph, employing a self-supervised link\u0000prediction pretraining task to capture the spatial and temporal dynamics in\u0000this network graph framework. To evaluate the effectiveness of our approach, we\u0000conduct a few-shot learning experiment for three distinct downstream network\u0000tasks: intrusion detection, traffic classification, and botnet classification.\u0000Models finetuned from our pretrained base achieve an average performance\u0000increase of 6.87% over training from scratch, demonstrating their ability to\u0000effectively learn general network traffic dynamics during pretraining. This\u0000success suggests the potential for a large-scale version to serve as an\u0000operational foundational model.","PeriodicalId":501280,"journal":{"name":"arXiv - CS - Networking and Internet Architecture","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142183882","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：