With the development of 5G and mobile edge computing, deep neural network (DNN) inference can be distributed at the edge to reduce communication overhead and inference time, namely, DNN distributed inference. DNN distributed inference will pose challenges to the resource allocation problem in metro optical networks (MONs). Efficient cooperative allocation of optical communication and computational resources can facilitate high-bandwidth and low-latency applications. However, it also introduces greater complexity to the resource allocation problem. In this study, we propose a joint resource allocation method using high-performance transfer deep reinforcement learning (T-DRL) to maximize network throughput. When the topologies or characteristics of MONs change, T-DRL requires only a small amount of transfer training to re-converge. Considering that the generalizability of conventional methods is inversely related to optimization performance, we develop two deployment schemes (i.e., single-agent and multi-agent) based on the T-DRL method to explore the performance of T-DRL. Simulation results demonstrate that T-DRL greatly reduces the blocking probability and average inference time of DNN inference requests. Besides, the multi-agent scheme can maintain a lower blocking probability of requests in MONs, while the single-agent has a shorter convergence time after network changes.
{"title":"DNN distributed inference offloading scheme based on transfer reinforcement learning in metro optical networks","authors":"Shan Yin;Lihao Liu;Mengru Cai;Yutong Chai;Yurong Jiao;Zheng Duan;Yian Li;Shanguo Huang","doi":"10.1364/JOCN.533206","DOIUrl":"10.1364/JOCN.533206","url":null,"abstract":"With the development of 5G and mobile edge computing, deep neural network (DNN) inference can be distributed at the edge to reduce communication overhead and inference time, namely, DNN distributed inference. DNN distributed inference will pose challenges to the resource allocation problem in metro optical networks (MONs). Efficient cooperative allocation of optical communication and computational resources can facilitate high-bandwidth and low-latency applications. However, it also introduces greater complexity to the resource allocation problem. In this study, we propose a joint resource allocation method using high-performance transfer deep reinforcement learning (T-DRL) to maximize network throughput. When the topologies or characteristics of MONs change, T-DRL requires only a small amount of transfer training to re-converge. Considering that the generalizability of conventional methods is inversely related to optimization performance, we develop two deployment schemes (i.e., single-agent and multi-agent) based on the T-DRL method to explore the performance of T-DRL. Simulation results demonstrate that T-DRL greatly reduces the blocking probability and average inference time of DNN inference requests. Besides, the multi-agent scheme can maintain a lower blocking probability of requests in MONs, while the single-agent has a shorter convergence time after network changes.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"16 9","pages":"852-867"},"PeriodicalIF":4.0,"publicationDate":"2024-08-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141668118","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jianmin Wang;Bin Li;Haochun Gao;Yang Lin;Zhiqian Su
The acquisition time of tens to hundreds of seconds in the optical link between satellites makes it difficult to meet the needs of constructing spatial optical networks. In addition, as a basic requirement for deep space explorers, autonomous attitude determination and autonomous navigation demand the installation of separate, expensive, and complex inertial devices, and the communication data rate is too low to meet the timely transmission of large amounts of data. In this paper, we proposed and developed a multifunctional fusion space optical communication system for space optical networks and deep space exploration, which has the functions of autonomous attitude determination, autonomous navigation, and high-speed optical communication. The sub-second acquisition time can meet the requirements of space optical network construction, and the ability of autonomous attitude determination and autonomous navigation significantly reduce the amount of R&D expenses of the explorer; decrease the volume, weight, and power consumption of the explorer; and improve the reliability and autonomous survival ability of the explorer. The paper provides the structure, working principle, and main algorithm models and conducts a feasibility analysis and field experiments. The experimental results showed that the average open-loop pointing accuracy of the optical terminal is 95.8 µrad (attitude determination accuracy), which can be improved to 39.1 µrad after filtering, and the acquisition time is less than 1 s. For deep space exploration, the navigation accuracy is less than 67.6 km in the cruise phase and 10 km in the acquisition phase, and field experiments have also proven its feasibility. The significance of our research work lies in proposing what we believe to be a new system operation scheme and design method for optical communication systems, and its results can be widely applied in all fields of space optical communication.
{"title":"Space optical communication system for space optical networks and deep space exploration","authors":"Jianmin Wang;Bin Li;Haochun Gao;Yang Lin;Zhiqian Su","doi":"10.1364/JOCN.520301","DOIUrl":"10.1364/JOCN.520301","url":null,"abstract":"The acquisition time of tens to hundreds of seconds in the optical link between satellites makes it difficult to meet the needs of constructing spatial optical networks. In addition, as a basic requirement for deep space explorers, autonomous attitude determination and autonomous navigation demand the installation of separate, expensive, and complex inertial devices, and the communication data rate is too low to meet the timely transmission of large amounts of data. In this paper, we proposed and developed a multifunctional fusion space optical communication system for space optical networks and deep space exploration, which has the functions of autonomous attitude determination, autonomous navigation, and high-speed optical communication. The sub-second acquisition time can meet the requirements of space optical network construction, and the ability of autonomous attitude determination and autonomous navigation significantly reduce the amount of R&D expenses of the explorer; decrease the volume, weight, and power consumption of the explorer; and improve the reliability and autonomous survival ability of the explorer. The paper provides the structure, working principle, and main algorithm models and conducts a feasibility analysis and field experiments. The experimental results showed that the average open-loop pointing accuracy of the optical terminal is 95.8 µrad (attitude determination accuracy), which can be improved to 39.1 µrad after filtering, and the acquisition time is less than 1 s. For deep space exploration, the navigation accuracy is less than 67.6 km in the cruise phase and 10 km in the acquisition phase, and field experiments have also proven its feasibility. The significance of our research work lies in proposing what we believe to be a new system operation scheme and design method for optical communication systems, and its results can be widely applied in all fields of space optical communication.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"16 9","pages":"843-851"},"PeriodicalIF":4.0,"publicationDate":"2024-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141673905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Beyond 5G, the next-generation wireless communication standard requires an optical communication network with a more reliable point-to-multipoint system. A highly reliable, futuristic point-to-multipoint coherent optical system must update wavelength-division-multiplexing-based passive optical networks to a more disaster-resistant architecture that can switch between bypass and backup links. We propose, to our knowledge, a novel link-pair shared protection that can respond locally to double-link failures that result from a significant disaster. We validate the high availability of several network configurations, assuming a double-link disconnection, can be obtained irrespective of the transmission distance of the feeder fiber. We experimentally demonstrate link-pair shared protection with bidirectional wavelength pre-assignment for two of the four feeder fiber failures and validate a penalty of less than 2 dB for double-link failures. Furthermore, we prove that reconnection can be performed with a penalty of at most 2 dB in an experiment with shared protection with a single-link broadcast-and-select function that can manage partial double-link failures using a simple configuration.
在 5G 之后,下一代无线通信标准要求光通信网络具备更可靠的点对多点系统。高可靠性的未来型点对多点相干光系统必须将基于波分复用技术的无源光网络更新为能在旁路和备份链路之间切换的抗灾能力更强的架构。据我们所知,我们提出了一种新型链路对共享保护方式,可在本地响应重大灾难导致的双链路故障。我们验证了几种网络配置的高可用性,假设双链路断开,无论馈线光纤的传输距离有多远,都能获得高可用性。我们在实验中演示了链路对共享保护与双向波长预分配,可应对四种馈线光纤故障中的两种故障,并验证了双链路故障的惩罚小于 2 dB。此外,我们还证明,在具有单链路广播和选择功能的共享保护实验中,使用简单的配置就能管理部分双链路故障,并能以最多 2 dB 的惩罚进行重新连接。
{"title":"Double-link-failure-tolerant shared protection for fully coupled/half-split switchable point-to-multipoint coherent optical systems","authors":"Takahiro Kodama;Tomoya Nakagawa;Shota Eguchi;Keiji Shimada;Ryosuke Matsumoto","doi":"10.1364/JOCN.511029","DOIUrl":"10.1364/JOCN.511029","url":null,"abstract":"Beyond 5G, the next-generation wireless communication standard requires an optical communication network with a more reliable point-to-multipoint system. A highly reliable, futuristic point-to-multipoint coherent optical system must update wavelength-division-multiplexing-based passive optical networks to a more disaster-resistant architecture that can switch between bypass and backup links. We propose, to our knowledge, a novel link-pair shared protection that can respond locally to double-link failures that result from a significant disaster. We validate the high availability of several network configurations, assuming a double-link disconnection, can be obtained irrespective of the transmission distance of the feeder fiber. We experimentally demonstrate link-pair shared protection with bidirectional wavelength pre-assignment for two of the four feeder fiber failures and validate a penalty of less than 2 dB for double-link failures. Furthermore, we prove that reconnection can be performed with a penalty of at most 2 dB in an experiment with shared protection with a single-link broadcast-and-select function that can manage partial double-link failures using a simple configuration.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"16 9","pages":"832-842"},"PeriodicalIF":4.0,"publicationDate":"2024-08-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141686877","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Carmen Vazquez;Ruben Altuna;Juan Dayron Lopez-Cardona
This paper introduces the concept of power over fiber (PoF) and potential applications envisioned in radio access networks with optical fronthauling using different types of optical fibers. It is an open discussion on exploring PoF technology with current experiments integrating analog radio over fiber using 5G-NR signals in compliance with 3GPP, PoF, and monitoring techniques and general requirements in future deployments along with its potential to be part of the energy efficiency strategy in beyond 5G networks.
{"title":"Power over fiber in radio access networks: 5G and beyond","authors":"Carmen Vazquez;Ruben Altuna;Juan Dayron Lopez-Cardona","doi":"10.1364/JOCN.522900","DOIUrl":"10.1364/JOCN.522900","url":null,"abstract":"This paper introduces the concept of power over fiber (PoF) and potential applications envisioned in radio access networks with optical fronthauling using different types of optical fibers. It is an open discussion on exploring PoF technology with current experiments integrating analog radio over fiber using 5G-NR signals in compliance with 3GPP, PoF, and monitoring techniques and general requirements in future deployments along with its potential to be part of the energy efficiency strategy in beyond 5G networks.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"16 8","pages":"D119-D128"},"PeriodicalIF":4.0,"publicationDate":"2024-08-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141666740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Linsheng Fan;Yanfu Yang;Qun Zhang;Siyu Gong;Yuchen Jia;Chen Cheng;Yong Yao
Optical performance monitoring is vital for enabling dynamically reconfigurable optical networks. This paper introduces a monitoring scheme that is robust to transceiver impairments for coherent digital subcarrier multiplexing (DSCM) systems that simultaneously monitors polarization-dependent loss (PDL), differential group delay (DGD), and chromatic dispersion (CD). In the proposed scheme, a pair of frequency domain pilot tones (FPTs) are inserted into the ${X}$