Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651953
Kristjon Ciko, M. Welzl, P. Teymoori
Deploying a new network architecture in the Internet requires changing some, but not necessarily all elements between communicating applications. One way to achieve gradual deployment is a proxy or gateway which "translates" between the new architecture and TCP/IP. We present such a proxy, called "Performance Enhancing Proxy for Deploying Network Architectures (PEP-DNA)", which allows TCP/IP applications to benefit from advanced features of a new network architecture without having to be redeveloped. Our proxy is a kernel-based Linux implementation which can be installed wherever a translation needs to occur between a new architecture and TCP/IP domains. We discuss the proxy operation in detail and evaluate its efficiency and performance in a local testbed, demonstrating that it achieves high throughput with low additional latency overhead. In our experiments, we use the Recursive InterNetwork Architecture (RINA) and Information-Centric Networking (ICN) as examples, but our proxy is modular and flexible, and hence enables realistic gradual deployment of any new "clean-slate" approaches.
{"title":"PEP-DNA: A Performance Enhancing Proxy for Deploying Network Architectures","authors":"Kristjon Ciko, M. Welzl, P. Teymoori","doi":"10.1109/ICNP52444.2021.9651953","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651953","url":null,"abstract":"Deploying a new network architecture in the Internet requires changing some, but not necessarily all elements between communicating applications. One way to achieve gradual deployment is a proxy or gateway which \"translates\" between the new architecture and TCP/IP. We present such a proxy, called \"Performance Enhancing Proxy for Deploying Network Architectures (PEP-DNA)\", which allows TCP/IP applications to benefit from advanced features of a new network architecture without having to be redeveloped. Our proxy is a kernel-based Linux implementation which can be installed wherever a translation needs to occur between a new architecture and TCP/IP domains. We discuss the proxy operation in detail and evaluate its efficiency and performance in a local testbed, demonstrating that it achieves high throughput with low additional latency overhead. In our experiments, we use the Recursive InterNetwork Architecture (RINA) and Information-Centric Networking (ICN) as examples, but our proxy is modular and flexible, and hence enables realistic gradual deployment of any new \"clean-slate\" approaches.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131512786","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651932
Mingsheng Yin, Tuyen X. Tran, Abhigyan Sharma, M. Mezzavilla, S. Rangan
There is a growing interest in reusing cellular base stations on the ground to provide long range, high-speed wireless connectivity to UAVs. Towards this goal, we present SkyRoute – a novel and powerful simulation platform for rapid and realistic assessment of UAV cellular connectivity. SkyRoute combines real base station locations and antenna data with a lightweight version of the widely-used ns-3 simulation platform for full-stack wireless channel and cellular network simulation. As an exemplary application, we demonstrate realistic coverage and cell selection prediction in a large metropolitan area.
{"title":"Demo: SkyRoute, a Fast and Realistic UAV Cellular Simulation Framework","authors":"Mingsheng Yin, Tuyen X. Tran, Abhigyan Sharma, M. Mezzavilla, S. Rangan","doi":"10.1109/ICNP52444.2021.9651932","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651932","url":null,"abstract":"There is a growing interest in reusing cellular base stations on the ground to provide long range, high-speed wireless connectivity to UAVs. Towards this goal, we present SkyRoute – a novel and powerful simulation platform for rapid and realistic assessment of UAV cellular connectivity. SkyRoute combines real base station locations and antenna data with a lightweight version of the widely-used ns-3 simulation platform for full-stack wireless channel and cellular network simulation. As an exemplary application, we demonstrate realistic coverage and cell selection prediction in a large metropolitan area.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131158390","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651919
Zeqi Lai, Qianxia Wu, Hewu Li, M. Lv, Jianping Wu
Satellite-based Earth Observation (EO) systems are gaining popularity and widely used in many time-sensitive scenarios, including disaster monitoring, emergency response, forecasting and defense. Existing efforts for gathering EO data mainly rely on either ground station networks or geostationary (GEO) satellites. However, our quantitative analysis reveals that existing approaches are either limited as their achievable latency is far away from the desired value due to the insufficient coverage of ground stations, or hard to scale as the number of sensing satellites increases because of the high cost of GEO satellite relays.This paper explores the feasibility and performance of a novel approach that leverages emerging low Earth orbit (LEO) constellations to enable low-latency and scalable EO data delivery from space. We present OrbitCast, a hybrid EO data delivery architecture upon LEO constellations and geo-distributed ground stations to forward EO data from the source remote sensing satellite to a collection of end users. To handle the network dynamicity caused by LEO satellite movements and achieve stable communication over the satellite network, we propose a geo-location driven scheme to forward and deliver data packets. To demonstrate the effectiveness of OrbitCast, we build a testbed driven by public constellation information and implement the OrbitCast prototype on top of the testbed. Extensive realistic-data-driven simulations demonstrate that OrbitCast can significantly reduce the latency as compared to other state-of-the-art approaches, and complete the data delivery within five minutes for representative EO data traffic.
{"title":"OrbitCast: Exploiting Mega-Constellations for Low-Latency Earth Observation","authors":"Zeqi Lai, Qianxia Wu, Hewu Li, M. Lv, Jianping Wu","doi":"10.1109/ICNP52444.2021.9651919","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651919","url":null,"abstract":"Satellite-based Earth Observation (EO) systems are gaining popularity and widely used in many time-sensitive scenarios, including disaster monitoring, emergency response, forecasting and defense. Existing efforts for gathering EO data mainly rely on either ground station networks or geostationary (GEO) satellites. However, our quantitative analysis reveals that existing approaches are either limited as their achievable latency is far away from the desired value due to the insufficient coverage of ground stations, or hard to scale as the number of sensing satellites increases because of the high cost of GEO satellite relays.This paper explores the feasibility and performance of a novel approach that leverages emerging low Earth orbit (LEO) constellations to enable low-latency and scalable EO data delivery from space. We present OrbitCast, a hybrid EO data delivery architecture upon LEO constellations and geo-distributed ground stations to forward EO data from the source remote sensing satellite to a collection of end users. To handle the network dynamicity caused by LEO satellite movements and achieve stable communication over the satellite network, we propose a geo-location driven scheme to forward and deliver data packets. To demonstrate the effectiveness of OrbitCast, we build a testbed driven by public constellation information and implement the OrbitCast prototype on top of the testbed. Extensive realistic-data-driven simulations demonstrate that OrbitCast can significantly reduce the latency as compared to other state-of-the-art approaches, and complete the data delivery within five minutes for representative EO data traffic.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"197 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124403554","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651980
Jinbin Hu, Jiawei Huang, Zhaoyi Li, Yijun Li, Wenchao Jiang, Kai Chen, Jianxin Wang, Tian He
Modern datacenter applications bring fundamental challenges to transport protocols as they simultaneously require low latency and high throughput. Recent receiver-driven trans-port protocols transmit only one data packet once receiving each grant or credit packet from the receiver to achieve ultra-low queueing delay and zero packet loss. However, the round-trip time variation and the highly dynamic background traffic significantly deteriorate the performance of receiver-driven transport protocols, resulting in under-utilized bandwidth. This paper designs a simple yet effective solution called RPO that retains the advantages of receiver-driven transmission while efficiently utilizing the available bandwidth. Specifically, RPO rationally uses low-priority opportunistic packets to ensure high network utilization without increasing the queueing delay of high-priority normal packets. In addition, since RPO only uses Explicit Congestion Notification (ECN) marking function and priority queues, RPO is ready to deploy on switches. We implement RPO in Linux hosts with DPDK. Our small-scale testbed experiments and large-scale simulations show that RPO significantly improves the network utilization by up to 35% under high workload over the state-of-the-art receiver-driven transmission schemes, without introducing additional queueing delay.
{"title":"RPO: Receiver-driven Transport Protocol Using Opportunistic Transmission in Data Center","authors":"Jinbin Hu, Jiawei Huang, Zhaoyi Li, Yijun Li, Wenchao Jiang, Kai Chen, Jianxin Wang, Tian He","doi":"10.1109/ICNP52444.2021.9651980","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651980","url":null,"abstract":"Modern datacenter applications bring fundamental challenges to transport protocols as they simultaneously require low latency and high throughput. Recent receiver-driven trans-port protocols transmit only one data packet once receiving each grant or credit packet from the receiver to achieve ultra-low queueing delay and zero packet loss. However, the round-trip time variation and the highly dynamic background traffic significantly deteriorate the performance of receiver-driven transport protocols, resulting in under-utilized bandwidth. This paper designs a simple yet effective solution called RPO that retains the advantages of receiver-driven transmission while efficiently utilizing the available bandwidth. Specifically, RPO rationally uses low-priority opportunistic packets to ensure high network utilization without increasing the queueing delay of high-priority normal packets. In addition, since RPO only uses Explicit Congestion Notification (ECN) marking function and priority queues, RPO is ready to deploy on switches. We implement RPO in Linux hosts with DPDK. Our small-scale testbed experiments and large-scale simulations show that RPO significantly improves the network utilization by up to 35% under high workload over the state-of-the-art receiver-driven transmission schemes, without introducing additional queueing delay.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125670681","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651922
Xinle Du, Tong Li, Lei Xu, Kai Zheng, Meng Shen, Bo Wu, Ke Xu
TCP incast has become a practical problem for high-bandwidth, low-latency transmissions, resulting in throughput degradation of up to 90% and delays of hundreds of milliseconds, severely impacting application performance. However, in virtualized multi-tenant data centers, host-based advancements in the TCP stack are hard to deploy from the operators perspective. Operators only provide infrastructure in the form of virtual machines, in which only tenants can directly modify the end-host TCP stack. In this paper, we present R-AQM, a switch-powered reverse ACK active queue management (R-AQM) mechanism for enhancing ACK-clocking effects through assisting legacy TCP. Specifically, R-AQM proactively intercepts ACKs and paces the ACK-clocked in-flight data packets, preventing TCP from suffering incast collapse. We implement and evaluate R-AQM in NS-3 simulation and NetFPGA-based hardware switch. Both simulation and testbed results show that R-AQM greatly improves TCP performance under heavy incast workloads by significantly lowering packet loss rate, reducing retransmission timeouts, and supporting 16 times (i.e., 60 → 1000) more senders. Meanwhile, the forward queuing delays are also reduced by 4.6 times.
{"title":"R-AQM: Reverse ACK Active Queue Management in Multi-tenant Data Centers","authors":"Xinle Du, Tong Li, Lei Xu, Kai Zheng, Meng Shen, Bo Wu, Ke Xu","doi":"10.1109/ICNP52444.2021.9651922","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651922","url":null,"abstract":"TCP incast has become a practical problem for high-bandwidth, low-latency transmissions, resulting in throughput degradation of up to 90% and delays of hundreds of milliseconds, severely impacting application performance. However, in virtualized multi-tenant data centers, host-based advancements in the TCP stack are hard to deploy from the operators perspective. Operators only provide infrastructure in the form of virtual machines, in which only tenants can directly modify the end-host TCP stack. In this paper, we present R-AQM, a switch-powered reverse ACK active queue management (R-AQM) mechanism for enhancing ACK-clocking effects through assisting legacy TCP. Specifically, R-AQM proactively intercepts ACKs and paces the ACK-clocked in-flight data packets, preventing TCP from suffering incast collapse. We implement and evaluate R-AQM in NS-3 simulation and NetFPGA-based hardware switch. Both simulation and testbed results show that R-AQM greatly improves TCP performance under heavy incast workloads by significantly lowering packet loss rate, reducing retransmission timeouts, and supporting 16 times (i.e., 60 → 1000) more senders. Meanwhile, the forward queuing delays are also reduced by 4.6 times.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"149 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127267296","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651929
Jacob Davis, C. Deccio
The Domain Name System (DNS) has been frequently abused for distributed denial-of-service (DDoS) attacks and cache poisoning because it relies on the User Datagram Protocol (UDP). Since UDP is connection-less, it is trivial for an attacker to spoof the source of a DNS query or response. While other secure transport mechanisms provide identity management, such as the Transmission Control Protocol (TCP) and DNS Cookies, there is currently no method for a client to state that they only use a given protocol. This paper presents a new method to allow protocol enforcement: DNS Protocol Advertisement Records (DPAR). Advertisement records allow Internet Protocol (IP) address subnets to post a public record in the reverse DNS zone stating which DNS mechanisms are used by their clients. DNS servers may then look up this record and require a client to use the stated mechanism, in turn preventing an attacker from sending spoofed messages over UDP. In this paper, we define the specification for DNS Protocol Advertisement Records, considerations that were made, and comparisons to alternative approaches. We additionally estimate the effectiveness of advertisements in preventing DDoS attacks and the expected burden to DNS servers.
{"title":"Advertising DNS Protocol Use to Mitigate DDoS Attacks","authors":"Jacob Davis, C. Deccio","doi":"10.1109/ICNP52444.2021.9651929","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651929","url":null,"abstract":"The Domain Name System (DNS) has been frequently abused for distributed denial-of-service (DDoS) attacks and cache poisoning because it relies on the User Datagram Protocol (UDP). Since UDP is connection-less, it is trivial for an attacker to spoof the source of a DNS query or response. While other secure transport mechanisms provide identity management, such as the Transmission Control Protocol (TCP) and DNS Cookies, there is currently no method for a client to state that they only use a given protocol. This paper presents a new method to allow protocol enforcement: DNS Protocol Advertisement Records (DPAR). Advertisement records allow Internet Protocol (IP) address subnets to post a public record in the reverse DNS zone stating which DNS mechanisms are used by their clients. DNS servers may then look up this record and require a client to use the stated mechanism, in turn preventing an attacker from sending spoofed messages over UDP. In this paper, we define the specification for DNS Protocol Advertisement Records, considerations that were made, and comparisons to alternative approaches. We additionally estimate the effectiveness of advertisements in preventing DDoS attacks and the expected burden to DNS servers.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130886563","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651963
Siyuan Sheng, Qun Huang, P. Lee
In-band network telemetry (INT) enriches network management at scale through the embedding of complete device-internal states into each packet along its forwarding path, yet such embedding of INT information also incurs significant band-width overhead in the data plane. We propose DeltaINT, a general INT framework that achieves extremely low bandwidth overhead and supports various packet-level and flow-level applications in network management. DeltaINT builds on the insight that state changes are often negligible at most time, so it embeds a state into a packet only when the state change is deemed significant. We theoretically derive the time/space complexities and the bounds of bandwidth mitigation for DeltaINT. We implement DeltaINT in both software and P4. Our evaluation shows that DeltaINT reduces up to 93% of INT bandwidth, and its deployment in a Barefoot Tofino switch incurs limited hardware resource usage.
{"title":"DeltaINT: Toward General In-band Network Telemetry with Extremely Low Bandwidth Overhead","authors":"Siyuan Sheng, Qun Huang, P. Lee","doi":"10.1109/ICNP52444.2021.9651963","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651963","url":null,"abstract":"In-band network telemetry (INT) enriches network management at scale through the embedding of complete device-internal states into each packet along its forwarding path, yet such embedding of INT information also incurs significant band-width overhead in the data plane. We propose DeltaINT, a general INT framework that achieves extremely low bandwidth overhead and supports various packet-level and flow-level applications in network management. DeltaINT builds on the insight that state changes are often negligible at most time, so it embeds a state into a packet only when the state change is deemed significant. We theoretically derive the time/space complexities and the bounds of bandwidth mitigation for DeltaINT. We implement DeltaINT in both software and P4. Our evaluation shows that DeltaINT reduces up to 93% of INT bandwidth, and its deployment in a Barefoot Tofino switch incurs limited hardware resource usage.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133593632","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651989
S. A. Fernandez, M. A. M. Marinho, M. Vakilzadeh, A. Vinel
Due to the ability to support a wide range of applications and to involve infrastructure elements, connected and automated vehicles (CAVs) technology has played an important role in the development of cooperative intelligent transport systems. Thus, with the available sensing system, CAVs can perceive the surrounding environment. Indeed, due to the involvement of CAVs, communication of vehicles to other related devices using vehicle-to-everything (V2X) communication plays critical roles. This paper summarizes the research and development trends when proposing driving models, with a particular attention to highway on-ramp merging scenarios. The challenges and future research directions are also presented.
{"title":"Highway On-Ramp Merging for Mixed Traffic: Recent Advances and Future Trends","authors":"S. A. Fernandez, M. A. M. Marinho, M. Vakilzadeh, A. Vinel","doi":"10.1109/ICNP52444.2021.9651989","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651989","url":null,"abstract":"Due to the ability to support a wide range of applications and to involve infrastructure elements, connected and automated vehicles (CAVs) technology has played an important role in the development of cooperative intelligent transport systems. Thus, with the available sensing system, CAVs can perceive the surrounding environment. Indeed, due to the involvement of CAVs, communication of vehicles to other related devices using vehicle-to-everything (V2X) communication plays critical roles. This paper summarizes the research and development trends when proposing driving models, with a particular attention to highway on-ramp merging scenarios. The challenges and future research directions are also presented.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114211456","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651920
Souvik Das, K. Saraç
The public Internet is a network of autonomously owned and operated networks. Outdated peering policies and lack of end-to-end performance guarantees are causing its ossification which have led large cloud and content providers to build their own global private backbone infrastructures. As much as these private backbones help eliminate public transit for content hosted across their networks, content hosted elsewhere is still carried over the public Internet. In this poster, we propose a model where these private backbone operators collaborate with the access-networks of content providers and consumers to implement end-to-end network services with better performance characteristics than the public Internet. We call the resulting end-to-end service domain as a "Private Internet".
{"title":"Poster: Private Internet: A Global End-to-End Service Model","authors":"Souvik Das, K. Saraç","doi":"10.1109/ICNP52444.2021.9651920","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651920","url":null,"abstract":"The public Internet is a network of autonomously owned and operated networks. Outdated peering policies and lack of end-to-end performance guarantees are causing its ossification which have led large cloud and content providers to build their own global private backbone infrastructures. As much as these private backbones help eliminate public transit for content hosted across their networks, content hosted elsewhere is still carried over the public Internet. In this poster, we propose a model where these private backbone operators collaborate with the access-networks of content providers and consumers to implement end-to-end network services with better performance characteristics than the public Internet. We call the resulting end-to-end service domain as a \"Private Internet\".","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"142 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116353117","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}
Pub Date : 2021-11-01DOI: 10.1109/ICNP52444.2021.9651918
Minghao Ye, Junjie Zhang, Zehua Guo, H. J. Chao
Network operators usually adopt Traffic Engineering (TE) to configure the routing in their networks to achieve good load balancing performance and high resource utilization. While centralized TE can effectively improve network performance with a global view of the network, distributed TE has been considered as an alternative to manage large-scale networks that are usually partitioned into multiple regions. However, it is challenging for distributed TE to reach a global optimal performance since each region can make its local routing decisions only based on partially observed network states. In this paper, we propose a novel distributed TE scheme called FedTe, which leverages supervised learning coupled with a collaborative approach to improve the overall load balancing performance for multi-region networks. FedTe learns from the global optimal routing strategy in a centralized offline manner and predicts the optimal distribution of cross-region traffic among different regions through distributed deployment in real time. The predicted cross-region traffic distribution is integrated with measured local traffic to construct each region’s optimal regional traffic matrix, which is used to perform intra-region TE optimization. FedTe can also handle dynamic traffic variation and link failures with a 2-layer hierarchical graph neural network architecture. To validate the effectiveness of the proposed scheme, we evaluate FedTe with two real-world network topologies and a large-scale synthetic topology. Extensive evaluation results show that FedTe can achieve near-optimal load balancing performance and outperform state-of-the-art distributed TE approaches by up to 28.9% on average.
{"title":"Federated Traffic Engineering with Supervised Learning in Multi-region Networks","authors":"Minghao Ye, Junjie Zhang, Zehua Guo, H. J. Chao","doi":"10.1109/ICNP52444.2021.9651918","DOIUrl":"https://doi.org/10.1109/ICNP52444.2021.9651918","url":null,"abstract":"Network operators usually adopt Traffic Engineering (TE) to configure the routing in their networks to achieve good load balancing performance and high resource utilization. While centralized TE can effectively improve network performance with a global view of the network, distributed TE has been considered as an alternative to manage large-scale networks that are usually partitioned into multiple regions. However, it is challenging for distributed TE to reach a global optimal performance since each region can make its local routing decisions only based on partially observed network states. In this paper, we propose a novel distributed TE scheme called FedTe, which leverages supervised learning coupled with a collaborative approach to improve the overall load balancing performance for multi-region networks. FedTe learns from the global optimal routing strategy in a centralized offline manner and predicts the optimal distribution of cross-region traffic among different regions through distributed deployment in real time. The predicted cross-region traffic distribution is integrated with measured local traffic to construct each region’s optimal regional traffic matrix, which is used to perform intra-region TE optimization. FedTe can also handle dynamic traffic variation and link failures with a 2-layer hierarchical graph neural network architecture. To validate the effectiveness of the proposed scheme, we evaluate FedTe with two real-world network topologies and a large-scale synthetic topology. Extensive evaluation results show that FedTe can achieve near-optimal load balancing performance and outperform state-of-the-art distributed TE approaches by up to 28.9% on average.","PeriodicalId":343813,"journal":{"name":"2021 IEEE 29th International Conference on Network Protocols (ICNP)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116229852","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：