Shreeshrita Patnaik, P. Barford, D. Fratta, Bill Jensen, N. Lord, Matthew Malloy, Herbert Wang
In this paper, we introduce Internet Photonic Sensing (IPS), a new framework for deformation and vibration measurement and monitoring based on signals that are available from standard fiber optic communication hardware deployed in the Internet. IPS is based on the hypothesis that atmospheric, seismic, anthropogenic and other natural activity cause deformations and vibrations in the earth that trigger detectable changes in standard optical signals that transmit data through Internet fiber. We assume a simple system component model for optical communication hardware and identify two candidate signals that may reflect deformation and vibrations and that can be measured through standard interfaces: Optical Signal Strength (OSS) and Bit Error Rate (BER). We investigate the efficacy of IPS through a series of controlled, laboratory experiments that consider how the candidate signals respond when fiber is subjected to a range of mechanical deformations. We believe that advancement of IPS offers the potential to transform the practice of scientific, commercial and public safety-related vibration monitoring applications by providing a highly-sensitive platform that is available at a global scale.
{"title":"Internet Photonic Sensing: Using the Internet Optical Transport Signals for Vibration and Deformation Sensing","authors":"Shreeshrita Patnaik, P. Barford, D. Fratta, Bill Jensen, N. Lord, Matthew Malloy, Herbert Wang","doi":"10.1145/3473938.3474507","DOIUrl":"https://doi.org/10.1145/3473938.3474507","url":null,"abstract":"In this paper, we introduce Internet Photonic Sensing (IPS), a new framework for deformation and vibration measurement and monitoring based on signals that are available from standard fiber optic communication hardware deployed in the Internet. IPS is based on the hypothesis that atmospheric, seismic, anthropogenic and other natural activity cause deformations and vibrations in the earth that trigger detectable changes in standard optical signals that transmit data through Internet fiber. We assume a simple system component model for optical communication hardware and identify two candidate signals that may reflect deformation and vibrations and that can be measured through standard interfaces: Optical Signal Strength (OSS) and Bit Error Rate (BER). We investigate the efficacy of IPS through a series of controlled, laboratory experiments that consider how the candidate signals respond when fiber is subjected to a range of mechanical deformations. We believe that advancement of IPS offers the potential to transform the practice of scientific, commercial and public safety-related vibration monitoring applications by providing a highly-sensitive platform that is available at a global scale.","PeriodicalId":302760,"journal":{"name":"Proceedings of the ACM SIGCOMM 2021 Workshop on Optical Systems","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115012973","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}
All-optical circuit switched network core is the holy grail for the next-generation datacenter architectures, as electrical packet switches are struggling to cope up with increasing challenges posed by the end of Moore's law. However, traffic skewness is the biggest enemy of such all-optical network cores comprising of a simple round-robin circuit-scheduling abstraction. Even though valiant load balancing can theoretically solve the problem, it falls short in most of the practical scenarios. In this paper, we point towards a new research direction to address the skewness problem: why not resolve most of the skewness at the network edge while keeping the optical core simple? This approach is fundamentally different and can potentially enable the all-optical network core to achieve good performance in practice. We discuss relevant strategies and envision that a holistic system design is necessary considering all these strategies together.
{"title":"Towards All-optical Circuit-switched Datacenter Network Cores: The Case for Mitigating Traffic Skewness at the Edge","authors":"Sushovan Das, Weitao Wang, T. Ng","doi":"10.1145/3473938.3474505","DOIUrl":"https://doi.org/10.1145/3473938.3474505","url":null,"abstract":"All-optical circuit switched network core is the holy grail for the next-generation datacenter architectures, as electrical packet switches are struggling to cope up with increasing challenges posed by the end of Moore's law. However, traffic skewness is the biggest enemy of such all-optical network cores comprising of a simple round-robin circuit-scheduling abstraction. Even though valiant load balancing can theoretically solve the problem, it falls short in most of the practical scenarios. In this paper, we point towards a new research direction to address the skewness problem: why not resolve most of the skewness at the network edge while keeping the optical core simple? This approach is fundamentally different and can potentially enable the all-optical network core to achieve good performance in practice. We discuss relevant strategies and envision that a holistic system design is necessary considering all these strategies together.","PeriodicalId":302760,"journal":{"name":"Proceedings of the ACM SIGCOMM 2021 Workshop on Optical Systems","volume":"144 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116603035","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}
Reconfigurable Hybrid (electrical/optical) Network (RHN) [1-4, 6, 8, 10, 11, 13-19] for modern datacenter architectures has gained significant momentum during the last decade. The primary advantage of such RHN architectures is the dynamic topological reconfigurability enabled by optical circuit switches (OCS). On one hand, RHN can benefit throughput-intensive applications by providing on-demand high-bandwidth links between the hosts (CPU/GPU/TPU), such as distributed deep neural network training and recommendation systems, etc. On the other hand, RHN can reduce the hop-count between the host pairs, improving the performance for latency-sensitive applications such as real-time customer interactions with in-memory file system. However, previous works mostly focused on finding a suitable topology to efficiently handle a given traffic demand. Performing such topology update together with SDN policy update in a holistic manner while maintaining per-packet consistency and other network invariants is still an open issue. Existing network maintenance and policy update solutions define the notion of per-packet consistency assuming a pure SDN network where the physical network topology is static. This assumption does not hold for RHN because dynamic topology reconfiguration is inherent to RHN. In this paper, first, we define an extended notion of per-packet consistency and discuss the other critical requirements for RHN updates. Next, we provide an abstraction of RHN update and propose Transtate, a general method to perform such RHN update while satisfying the critical requirements. We believe such innovations remove one of the key obstacles towards reconfigurable-hybrid SDN.
{"title":"Abstractions for Reconfigurable Hybrid Network Update and A Consistent Update Approach","authors":"Weitao Wang, Sushovan Das, T. Ng","doi":"10.1145/3473938.3474506","DOIUrl":"https://doi.org/10.1145/3473938.3474506","url":null,"abstract":"Reconfigurable Hybrid (electrical/optical) Network (RHN) [1-4, 6, 8, 10, 11, 13-19] for modern datacenter architectures has gained significant momentum during the last decade. The primary advantage of such RHN architectures is the dynamic topological reconfigurability enabled by optical circuit switches (OCS). On one hand, RHN can benefit throughput-intensive applications by providing on-demand high-bandwidth links between the hosts (CPU/GPU/TPU), such as distributed deep neural network training and recommendation systems, etc. On the other hand, RHN can reduce the hop-count between the host pairs, improving the performance for latency-sensitive applications such as real-time customer interactions with in-memory file system. However, previous works mostly focused on finding a suitable topology to efficiently handle a given traffic demand. Performing such topology update together with SDN policy update in a holistic manner while maintaining per-packet consistency and other network invariants is still an open issue. Existing network maintenance and policy update solutions define the notion of per-packet consistency assuming a pure SDN network where the physical network topology is static. This assumption does not hold for RHN because dynamic topology reconfiguration is inherent to RHN. In this paper, first, we define an extended notion of per-packet consistency and discuss the other critical requirements for RHN updates. Next, we provide an abstraction of RHN update and propose Transtate, a general method to perform such RHN update while satisfying the critical requirements. We believe such innovations remove one of the key obstacles towards reconfigurable-hybrid SDN.","PeriodicalId":302760,"journal":{"name":"Proceedings of the ACM SIGCOMM 2021 Workshop on Optical Systems","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128682097","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}
M. Hall, P. Barford, Klaus-Tycho Foerster, M. Ghobadi, William Jensen, Ramakrishnan Durairajan
In today's wide-area networks, the optical layer is a relatively static and inflexible commodity. In response, Optical Topology Programming (OTP) has been proposed to enable fast and flexible reconfiguration of wavelengths at the optical layer from higher layers. We answer whether WANs are ready for OTP, concluding they are not. We reach this judgement by measuring reconfiguration delay on a long-haul fiber span. To push the needle on OTP towards feasibility, we show how to reduce the time to provision a circuit by an order of magnitude---from minutes to seconds. Finally, we propose a method to quickly store and load optical network equipment settings, reducing the time to less than 1 second.
{"title":"Are WANs Ready for Optical Topology Programming?","authors":"M. Hall, P. Barford, Klaus-Tycho Foerster, M. Ghobadi, William Jensen, Ramakrishnan Durairajan","doi":"10.1145/3473938.3474510","DOIUrl":"https://doi.org/10.1145/3473938.3474510","url":null,"abstract":"In today's wide-area networks, the optical layer is a relatively static and inflexible commodity. In response, Optical Topology Programming (OTP) has been proposed to enable fast and flexible reconfiguration of wavelengths at the optical layer from higher layers. We answer whether WANs are ready for OTP, concluding they are not. We reach this judgement by measuring reconfiguration delay on a long-haul fiber span. To push the needle on OTP towards feasibility, we show how to reduce the time to provision a circuit by an order of magnitude---from minutes to seconds. Finally, we propose a method to quickly store and load optical network equipment settings, reducing the time to less than 1 second.","PeriodicalId":302760,"journal":{"name":"Proceedings of the ACM SIGCOMM 2021 Workshop on Optical Systems","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121970712","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}
Che-Yu Liu, Xiaoliang Chen, R. Proietti, Zhaohui Li, S. Yoo
This paper presents a self-supervised machine learning approach for cognitive reconfiguration in a Hyper-X-like flexible-bandwidth optical interconnect architecture. The proposed approach makes use of a clustering algorithm to learn the traffic patterns from historical traces. A heuristic algorithm is developed for optimizing the connectivity graph for each identified traffic pattern. Further, to mitigate the scalability issue induced by frequent clustering operations, we parameterize the learned traffic patterns by a deep neural network classifier. The classifier is trained offline by supervised learning to enable classification of traffic matrices during online operations, thereby facilitating cognitive reconfiguration decision making. Simulation results show that compared with a static all-to-all interconnection, the proposed approach can improve throughput by up to 1.76× while reducing end-to-end packet latency and flow completion time by up to 2.8× and 25×, respectively.
{"title":"Reconfigurable Optical Datacom Networks by Self-supervised Learning","authors":"Che-Yu Liu, Xiaoliang Chen, R. Proietti, Zhaohui Li, S. Yoo","doi":"10.1145/3473938.3474509","DOIUrl":"https://doi.org/10.1145/3473938.3474509","url":null,"abstract":"This paper presents a self-supervised machine learning approach for cognitive reconfiguration in a Hyper-X-like flexible-bandwidth optical interconnect architecture. The proposed approach makes use of a clustering algorithm to learn the traffic patterns from historical traces. A heuristic algorithm is developed for optimizing the connectivity graph for each identified traffic pattern. Further, to mitigate the scalability issue induced by frequent clustering operations, we parameterize the learned traffic patterns by a deep neural network classifier. The classifier is trained offline by supervised learning to enable classification of traffic matrices during online operations, thereby facilitating cognitive reconfiguration decision making. Simulation results show that compared with a static all-to-all interconnection, the proposed approach can improve throughput by up to 1.76× while reducing end-to-end packet latency and flow completion time by up to 2.8× and 25×, respectively.","PeriodicalId":302760,"journal":{"name":"Proceedings of the ACM SIGCOMM 2021 Workshop on Optical Systems","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122177581","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}
Zhizhen Zhong, Weiyang Wang, M. Ghobadi, Alexander Sludds, R. Hamerly, Liane Bernstein, D. Englund
We present In-network Optical Inference (IOI), a system providing low-latency machine learning inference by leveraging programmable switches and optical matrix multiplication. IOI consists of a novel transceiver module designed specifically to perform linear operations such as matrix multiplication in the optical domain. IOI's transceivers are plugged into programmable switches to perform non-linear activation and respond to inference queries. We demonstrate how to process inference queries inside the network, without the need to send the queries to cloud or edge inference servers, thus significantly reducing end-to-end inference latency experienced by users. We believe IOI is the next frontier for exploring real-time machine learning systems and opens up exciting new opportunities for low-latency in-network inference.
{"title":"IOI: In-network Optical Inference","authors":"Zhizhen Zhong, Weiyang Wang, M. Ghobadi, Alexander Sludds, R. Hamerly, Liane Bernstein, D. Englund","doi":"10.1145/3473938.3474508","DOIUrl":"https://doi.org/10.1145/3473938.3474508","url":null,"abstract":"We present In-network Optical Inference (IOI), a system providing low-latency machine learning inference by leveraging programmable switches and optical matrix multiplication. IOI consists of a novel transceiver module designed specifically to perform linear operations such as matrix multiplication in the optical domain. IOI's transceivers are plugged into programmable switches to perform non-linear activation and respond to inference queries. We demonstrate how to process inference queries inside the network, without the need to send the queries to cloud or edge inference servers, thus significantly reducing end-to-end inference latency experienced by users. We believe IOI is the next frontier for exploring real-time machine learning systems and opens up exciting new opportunities for low-latency in-network inference.","PeriodicalId":302760,"journal":{"name":"Proceedings of the ACM SIGCOMM 2021 Workshop on Optical Systems","volume":"216 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122175290","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：