P. O. Amadi;S. A. Aljunid;N. Ali;S. M. Ammar;N. Rusli;R. Endut;A. M. Alhassan
This study presents the design and analysis of a hybrid multi-user quantum key distribution (QKD) system utilizing spectral amplitude coding optical code division multiple access (SAC-OCDMA) encoding techniques. By assigning unique optical codes to each user, SAC-OCDMA enables spontaneous, asynchronous data transmission without any strict synchronization, which makes it scalable and flexible for quantum networks. In the architecture, each user’s quantum signal, initially prepared as weak coherent pulses and encoded using phase or polarization bases, is further spectrally sliced by a SAC-OCDMA encoder in a zero-based cross-correlation code. The physical impairments, comprising spontaneous Raman scattering, four-wave mixing, and crosstalk, were modeled and analyzed. We report a maximum secret key rate of ${sim}{10^5};{rm bps}$ over a transmission distance of ${sim}58;{rm km}$. Furthermore, our analysis demonstrates that careful selection of launch power, simultaneous users, code weight, and spectral bin width is necessary for optimizing the trade-off between multi-user capacity and security performance. In comparison with wave division multiplexing QKD, which has much better spectral efficiency but strict channel allocation, the flexibility and asynchronous access to secure communication of our OCDMA-QKD design with zero cross-correlation coding are more aligned with the needs of quantum security.
{"title":"Quantum key distribution with spectral amplitude coding OCDMA for a multiple access network","authors":"P. O. Amadi;S. A. Aljunid;N. Ali;S. M. Ammar;N. Rusli;R. Endut;A. M. Alhassan","doi":"10.1364/JOCN.574271","DOIUrl":"https://doi.org/10.1364/JOCN.574271","url":null,"abstract":"This study presents the design and analysis of a hybrid multi-user quantum key distribution (QKD) system utilizing spectral amplitude coding optical code division multiple access (SAC-OCDMA) encoding techniques. By assigning unique optical codes to each user, SAC-OCDMA enables spontaneous, asynchronous data transmission without any strict synchronization, which makes it scalable and flexible for quantum networks. In the architecture, each user’s quantum signal, initially prepared as weak coherent pulses and encoded using phase or polarization bases, is further spectrally sliced by a SAC-OCDMA encoder in a zero-based cross-correlation code. The physical impairments, comprising spontaneous Raman scattering, four-wave mixing, and crosstalk, were modeled and analyzed. We report a maximum secret key rate of <tex>${sim}{10^5};{rm bps}$</tex> over a transmission distance of <tex>${sim}58;{rm km}$</tex>. Furthermore, our analysis demonstrates that careful selection of launch power, simultaneous users, code weight, and spectral bin width is necessary for optimizing the trade-off between multi-user capacity and security performance. In comparison with wave division multiplexing QKD, which has much better spectral efficiency but strict channel allocation, the flexibility and asynchronous access to secure communication of our OCDMA-QKD design with zero cross-correlation coding are more aligned with the needs of quantum security.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 12","pages":"1094-1104"},"PeriodicalIF":4.3,"publicationDate":"2025-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145510223","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}
As the scale and complexity of fiber-optic networks increase, the automatic detection and localization of soft failures have become crucial tasks for maintaining network services and scheduling repair actions. Fortunately, in software-defined optical networks (SDONs), the availability of extensive monitoring data can facilitate the application of machine learning (ML) techniques for effective failure management. However, the detection and localization of erbium-doped fiber amplifier (EDFA)-related failures, such as EDFA gain degradation, are hard to realize without the ubiquitous deployment of additional optical performance monitoring (OPM) devices. To address this issue, we propose an ML-based hierarchical framework for proactive detection and localization of inline EDFA gain degradation in optical networks. The framework is embedded within standard digital coherent receivers and comprises three stages, namely, time series prediction, failure detection, and failure localization. In the first stage, triple exponential smoothing (TES) is used to predict future signal power time series. Next, a normalizing flow (NF)-based neural network model is developed to detect abnormalities in the predicted signal power time series. If a failure is predicted by the two above-mentioned stages, the third stage of failure localization is automatically triggered in which a multi-task adaptive classification network (MTA-CN) is used to pinpoint an EDFA facing gain degradation by analyzing the amplitude histogram of the received symbols obtained after standard digital signal processing in the receiver. The efficacy of the proposed framework is rigorously validated through extensive experiments conducted on optical links with varying numbers of spans, i.e., 3 spans, 4 spans, and 5 spans. The results demonstrate an F1 score of 0.962 for proactive failure detection, along with F1 scores of 0.958, 0.931, and 0.919 for accurately localizing the faulty EDFAs in the three respective link configurations considered.
{"title":"Machine learning-aided hierarchical framework for proactive inline EDFA gain degradation detection and localization in optical networks","authors":"Hongcheng Wu;Qi Hu;Zhuojun Cai;Gai Zhou;Kangping Zhong;Faisal Nadeem Khan","doi":"10.1364/JOCN.572232","DOIUrl":"https://doi.org/10.1364/JOCN.572232","url":null,"abstract":"As the scale and complexity of fiber-optic networks increase, the automatic detection and localization of soft failures have become crucial tasks for maintaining network services and scheduling repair actions. Fortunately, in software-defined optical networks (SDONs), the availability of extensive monitoring data can facilitate the application of machine learning (ML) techniques for effective failure management. However, the detection and localization of erbium-doped fiber amplifier (EDFA)-related failures, such as EDFA gain degradation, are hard to realize without the ubiquitous deployment of additional optical performance monitoring (OPM) devices. To address this issue, we propose an ML-based hierarchical framework for proactive detection and localization of inline EDFA gain degradation in optical networks. The framework is embedded within standard digital coherent receivers and comprises three stages, namely, time series prediction, failure detection, and failure localization. In the first stage, triple exponential smoothing (TES) is used to predict future signal power time series. Next, a normalizing flow (NF)-based neural network model is developed to detect abnormalities in the predicted signal power time series. If a failure is predicted by the two above-mentioned stages, the third stage of failure localization is automatically triggered in which a multi-task adaptive classification network (MTA-CN) is used to pinpoint an EDFA facing gain degradation by analyzing the amplitude histogram of the received symbols obtained after standard digital signal processing in the receiver. The efficacy of the proposed framework is rigorously validated through extensive experiments conducted on optical links with varying numbers of spans, i.e., 3 spans, 4 spans, and 5 spans. The results demonstrate an F1 score of 0.962 for proactive failure detection, along with F1 scores of 0.958, 0.931, and 0.919 for accurately localizing the faulty EDFAs in the three respective link configurations considered.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 12","pages":"1082-1093"},"PeriodicalIF":4.3,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145510239","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}
In his opening OFC plenary talk back in 2021, Alibaba Group’s Yiqun Cai notably added in the follow-up Q&A that today’s complex networks are more than computer science—they grow, they are life. This entails that future networks may be better viewed as techno-social systems that resemble biological superorganisms with brain-like cognitive capabilities. Fast-forwarding, there is now growing awareness that we have to completely change our networks from being static to being a living entity that would act as an AI-powered network “brain,” as recently stated by Bruno Zerbib, Chief Technology and Innovation Officer of France’s Orange, at the Mobile World Congress (MWC) 2025. Even though AI was front and center at both MWC and OFC 2025 and has been widely studied in the context of optical networks, there are currently no publications on active inference in optical (and less so mobile) networks available. Active inference is an ideal methodology for developing more advanced AI systems by biomimicking the way living intelligent systems work while overcoming the limitations of today’s AI related to training, learning, and explainability. Active inference is considered the key to true AI: less artificial, more intelligent. It is a biomimetic mathematical framework that is premised on the first principles of statistical physics found in self-organizing/evolving complex adaptive systems, whether natural, artificial, or hybrid cyborganic ones. The goal of this paper is twofold. First, we aim at enabling optical network researchers to conceptualize new research lines for future optical networks with human-AI interaction capabilities by introducing them to the main mathematical concepts of the active inference framework. Second, we demonstrate how to move AI research beyond the human brain toward the 6G world brain by exploring the role of mycorrhizal networks, the largest living organism on planet Earth, in the AI vision and R&D roadmap for the next decade and beyond laid out by Karl Friston, the father of active inference.
{"title":"From artificial intelligence to active inference: the key to true AI and the 6G world brain [Invited]","authors":"Martin Maier","doi":"10.1364/JOCN.566810","DOIUrl":"https://doi.org/10.1364/JOCN.566810","url":null,"abstract":"In his opening OFC plenary talk back in 2021, Alibaba Group’s Yiqun Cai notably added in the follow-up Q&A that today’s complex networks are more than computer science—they grow, they are life. This entails that future networks may be better viewed as techno-social systems that resemble biological superorganisms with brain-like cognitive capabilities. Fast-forwarding, there is now growing awareness that we have to completely change our networks from being static to being a living entity that would act as an AI-powered network “brain,” as recently stated by Bruno Zerbib, Chief Technology and Innovation Officer of France’s Orange, at the Mobile World Congress (MWC) 2025. Even though AI was front and center at both MWC and OFC 2025 and has been widely studied in the context of optical networks, there are currently no publications on active inference in optical (and less so mobile) networks available. Active inference is an ideal methodology for developing more advanced AI systems by biomimicking the way living intelligent systems work while overcoming the limitations of today’s AI related to training, learning, and explainability. Active inference is considered the key to true AI: less artificial, more intelligent. It is a biomimetic mathematical framework that is premised on the first principles of statistical physics found in self-organizing/evolving complex adaptive systems, whether natural, artificial, or hybrid cyborganic ones. The goal of this paper is twofold. First, we aim at enabling optical network researchers to conceptualize new research lines for future optical networks with human-AI interaction capabilities by introducing them to the main mathematical concepts of the active inference framework. Second, we demonstrate how to move AI research beyond the human brain toward the 6G world brain by exploring the role of mycorrhizal networks, the largest living organism on planet Earth, in the AI vision and R&D roadmap for the next decade and beyond laid out by Karl Friston, the father of active inference.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"18 1","pages":"A28-A43"},"PeriodicalIF":4.3,"publicationDate":"2025-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145510233","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}
Haipeng Zhang;Zhensheng Jia;Luis Alberto Campos;Curtis Knittle
Coherent passive optical networks (PONs) are a promising solution for next-generation optical access networks, offering high capacity, extended reach, and improved spectral efficiency. This paper presents a single-laser bidirectional (BiDi) coherent PON architecture that supports hybrid single-carrier (SC) and digital subcarrier (DSC) transmission, enabling cost-effective coherent PON implementation and adaptive resource allocation within the same network. The system was experimentally evaluated over a 50 km optical distribution network (ODN) using multiple split ratio configurations, reflecting practical PON deployment scenarios, for both 25 GBd SC and 6.25 GBd DSC transmissions, demonstrating negligible back-reflection penalties compared to conventional full-duplex BiDi schemes. An upstream burst digital signal processing (DSP) framework is proposed, featuring a simple modulation format selection method based on burst rising edge detection and synchronization peak index information, supporting flexible-rate burst-mode upstream transmission across multiple optical network units (ONUs). Experimental results validate the system’s performance across various link distances and split ratios, achieving robust transmission with minimal inter-subcarrier interference. The proposed system offers a cost-effective and scalable solution for next-generation high-speed optical access networks.
{"title":"Single-laser bidirectional coherent PON with hybrid SC and DSC transmission for flexible and cost-effective optical access networks","authors":"Haipeng Zhang;Zhensheng Jia;Luis Alberto Campos;Curtis Knittle","doi":"10.1364/JOCN.571757","DOIUrl":"https://doi.org/10.1364/JOCN.571757","url":null,"abstract":"Coherent passive optical networks (PONs) are a promising solution for next-generation optical access networks, offering high capacity, extended reach, and improved spectral efficiency. This paper presents a single-laser bidirectional (BiDi) coherent PON architecture that supports hybrid single-carrier (SC) and digital subcarrier (DSC) transmission, enabling cost-effective coherent PON implementation and adaptive resource allocation within the same network. The system was experimentally evaluated over a 50 km optical distribution network (ODN) using multiple split ratio configurations, reflecting practical PON deployment scenarios, for both 25 GBd SC and 6.25 GBd DSC transmissions, demonstrating negligible back-reflection penalties compared to conventional full-duplex BiDi schemes. An upstream burst digital signal processing (DSP) framework is proposed, featuring a simple modulation format selection method based on burst rising edge detection and synchronization peak index information, supporting flexible-rate burst-mode upstream transmission across multiple optical network units (ONUs). Experimental results validate the system’s performance across various link distances and split ratios, achieving robust transmission with minimal inter-subcarrier interference. The proposed system offers a cost-effective and scalable solution for next-generation high-speed optical access networks.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"18 1","pages":"A19-A27"},"PeriodicalIF":4.3,"publicationDate":"2025-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145456030","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}
Matheus Sena;Mael Flament;Shane Andrewski;Ioannis Caltzidis;Niccolo Bigagli;Thomas Rieser;Gabriel Bello Portmann;Rourke Sekelsky;Ralf-Peter Braun;Alexander N. Craddock;Maximilian Schulz;Klaus D. Jons;Michaela Ritter;Marc Geitz;Oliver Holschke;Mehdi Namazi
The Quantum Internet, a network of quantum-enabled infrastructure, represents the next frontier in telecommunications, promising capabilities that cannot be attained by classical counterparts. A crucial step in realizing such large-scale quantum networks is the integration of entanglement distribution within existing telecommunication infrastructure. Here, we demonstrate a real-world scalable quantum networking testbed deployed within Deutsche Telekom’s metropolitan fibers in Berlin. Using commercially available quantum devices and standard add-drop multiplexing hardware, we distributed polarization-entangled photon pairs over dynamically selectable looped fiber paths ranging from 10 m to 60 km and showed entanglement distribution over up to approximately 100 km. Quantum signals, transmitted at 1324 nm (O-band), coexist with conventional bidirectional C-band traffic without dedicated fibers or infrastructure changes. Active stabilization of the polarization enables robust long-term performance, achieving entanglement Bell-state fidelity bounds between 85% and 99% and Clauser–Horne–Shimony–Holt parameter $S$-values between 2.36 and 2.74 during continuous multiday operation. By achieving a high-fidelity entanglement distribution with less than 1.5% downtime, we confirm the feasibility of hybrid quantum-classical networks under real-world conditions at the metropolitan scale. These results establish deployment benchmarks and provide a practical roadmap for telecom operators to integrate quantum capabilities.
{"title":"High-fidelity quantum entanglement distribution in metropolitan fiber networks with co-propagating classical traffic","authors":"Matheus Sena;Mael Flament;Shane Andrewski;Ioannis Caltzidis;Niccolo Bigagli;Thomas Rieser;Gabriel Bello Portmann;Rourke Sekelsky;Ralf-Peter Braun;Alexander N. Craddock;Maximilian Schulz;Klaus D. Jons;Michaela Ritter;Marc Geitz;Oliver Holschke;Mehdi Namazi","doi":"10.1364/JOCN.575396","DOIUrl":"https://doi.org/10.1364/JOCN.575396","url":null,"abstract":"The Quantum Internet, a network of quantum-enabled infrastructure, represents the next frontier in telecommunications, promising capabilities that cannot be attained by classical counterparts. A crucial step in realizing such large-scale quantum networks is the integration of entanglement distribution within existing telecommunication infrastructure. Here, we demonstrate a real-world scalable quantum networking testbed deployed within Deutsche Telekom’s metropolitan fibers in Berlin. Using commercially available quantum devices and standard add-drop multiplexing hardware, we distributed polarization-entangled photon pairs over dynamically selectable looped fiber paths ranging from 10 m to 60 km and showed entanglement distribution over up to approximately 100 km. Quantum signals, transmitted at 1324 nm (O-band), coexist with conventional bidirectional C-band traffic without dedicated fibers or infrastructure changes. Active stabilization of the polarization enables robust long-term performance, achieving entanglement Bell-state fidelity bounds between 85% and 99% and Clauser–Horne–Shimony–Holt parameter <tex>$S$</tex>-values between 2.36 and 2.74 during continuous multiday operation. By achieving a high-fidelity entanglement distribution with less than 1.5% downtime, we confirm the feasibility of hybrid quantum-classical networks under real-world conditions at the metropolitan scale. These results establish deployment benchmarks and provide a practical roadmap for telecom operators to integrate quantum capabilities.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 12","pages":"1072-1081"},"PeriodicalIF":4.3,"publicationDate":"2025-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145456049","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}
Farhad Arpanaei;Mohammadreza Dibaj;Amirhossein Dibaj;Hamzeh Beyranvand;John S. Vardakas;Christos Verikoukis;Jose Manuel Rivas-Moscoso;Juan Pedro Fernandez-Palacios;Alfonso Sanchez-Macian;David Larrabeiti;Jose Alberto Hernandez
As optical networks evolve toward dynamic, multi-band (C $+$ L) architectures, efficient and QoT-aware resource management becomes essential to ensure scalable and low-service downtime operation. This paper introduces a novel, to our knowledge, unified hybrid grooming framework that addresses the unique challenges of traffic grooming in dynamic multi-band elastic optical networks (MB-EONs). Motivated by the need for cost-effective and adaptive high-capacity infrastructures, we propose a policy-based framework incorporating three heuristic algorithms tailored to distinct optimization goals. The unique challenges of multi-band optical networks, such as the non-uniform QoT performance caused by inter-channel stimulated Raman scattering (ISRS) are explicitly considered in our design, as they directly impact grooming efficiency, spectrum utilization, and achievable modulation formats. The algorithms include (i) Min–Max Channel, which minimizes spectrum fragmentation and reduces the partial bit rate blocking probability by up to 35%; (ii) Max Grooming Capacity, which improves line card interface (LCI) reuse and reduces deployment by 20%; and (iii) Time Aware, which minimizes reconfiguration counts by up to 80%, significantly lowering control overhead and service downtime. Unlike prior works limited to static or single-band scenarios, our framework is the first, to our knowledge, to dynamically integrate routing, band selection, modulation format, grooming, and spectrum assignment (RBMGSA) in a QoT-aware manner. Simulation results over the NSFNET, Japan, and Spain topologies under dynamic traffic conditions demonstrate that our approach supports flexible trade-offs among performance, cost, and reconfiguration complexity. Notably, the reconfigurable variants of our algorithms consistently outperform non-reconfigurable approaches by enhancing resource utilization and reducing blocking. The proposed system also supports partial grooming, enabling improved service accommodation and laying the groundwork for scalable and efficient operation in future multi-band optical networks.
{"title":"Evaluating QoT-aware hybrid grooming schemes in dynamic C + L-band optical networks","authors":"Farhad Arpanaei;Mohammadreza Dibaj;Amirhossein Dibaj;Hamzeh Beyranvand;John S. Vardakas;Christos Verikoukis;Jose Manuel Rivas-Moscoso;Juan Pedro Fernandez-Palacios;Alfonso Sanchez-Macian;David Larrabeiti;Jose Alberto Hernandez","doi":"10.1364/JOCN.571277","DOIUrl":"https://doi.org/10.1364/JOCN.571277","url":null,"abstract":"As optical networks evolve toward dynamic, multi-band (C <tex>$+$</tex> L) architectures, efficient and QoT-aware resource management becomes essential to ensure scalable and low-service downtime operation. This paper introduces a novel, to our knowledge, unified hybrid grooming framework that addresses the unique challenges of traffic grooming in dynamic multi-band elastic optical networks (MB-EONs). Motivated by the need for cost-effective and adaptive high-capacity infrastructures, we propose a policy-based framework incorporating three heuristic algorithms tailored to distinct optimization goals. The unique challenges of multi-band optical networks, such as the non-uniform QoT performance caused by inter-channel stimulated Raman scattering (ISRS) are explicitly considered in our design, as they directly impact grooming efficiency, spectrum utilization, and achievable modulation formats. The algorithms include (i) Min–Max Channel, which minimizes spectrum fragmentation and reduces the partial bit rate blocking probability by up to 35%; (ii) Max Grooming Capacity, which improves line card interface (LCI) reuse and reduces deployment by 20%; and (iii) Time Aware, which minimizes reconfiguration counts by up to 80%, significantly lowering control overhead and service downtime. Unlike prior works limited to static or single-band scenarios, our framework is the first, to our knowledge, to dynamically integrate routing, band selection, modulation format, grooming, and spectrum assignment (RBMGSA) in a QoT-aware manner. Simulation results over the NSFNET, Japan, and Spain topologies under dynamic traffic conditions demonstrate that our approach supports flexible trade-offs among performance, cost, and reconfiguration complexity. Notably, the reconfigurable variants of our algorithms consistently outperform non-reconfigurable approaches by enhancing resource utilization and reducing blocking. The proposed system also supports partial grooming, enabling improved service accommodation and laying the groundwork for scalable and efficient operation in future multi-band optical networks.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 12","pages":"1059-1071"},"PeriodicalIF":4.3,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145456050","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}
With the increasing demand for low-latency deep neural network (DNN) inference, edge-cloud collaborative inference has become a promising paradigm. However, the increasing diversity of models, coupled with the limited and heterogeneous resources of edge nodes, makes it impractical to pre-deploy all models at the edge. These constraints not only intensify the complexity of model partitioning and slice delivery but also impose stricter requirements on scheduling and resource coordination. To address these challenges, this paper proposes a joint optimization approach for model partitioning and slice delivery to improve inference performance. We first formulate a mixed-integer nonlinear programming (MINLP) model for exact optimization. A heuristically seeded genetic algorithm (HSGA) is further developed, which incorporates heuristic initialization and task-driven alternating evolution to improve solution quality and convergence speed. Simulation results demonstrate that the proposed algorithm significantly reduces both the task completion time and blocking rate, validating its effectiveness in complex edge-cloud environments.
{"title":"Joint optimization of DNN model partitioning and slice delivery for distributed edge-cloud inference over optical networks","authors":"Tingting Bao;Xin Li;Yongli Zhao;Meng Lian;Yike Jiang;Jie Zhang","doi":"10.1364/JOCN.575114","DOIUrl":"https://doi.org/10.1364/JOCN.575114","url":null,"abstract":"With the increasing demand for low-latency deep neural network (DNN) inference, edge-cloud collaborative inference has become a promising paradigm. However, the increasing diversity of models, coupled with the limited and heterogeneous resources of edge nodes, makes it impractical to pre-deploy all models at the edge. These constraints not only intensify the complexity of model partitioning and slice delivery but also impose stricter requirements on scheduling and resource coordination. To address these challenges, this paper proposes a joint optimization approach for model partitioning and slice delivery to improve inference performance. We first formulate a mixed-integer nonlinear programming (MINLP) model for exact optimization. A heuristically seeded genetic algorithm (HSGA) is further developed, which incorporates heuristic initialization and task-driven alternating evolution to improve solution quality and convergence speed. Simulation results demonstrate that the proposed algorithm significantly reduces both the task completion time and blocking rate, validating its effectiveness in complex edge-cloud environments.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 11","pages":"1047-1058"},"PeriodicalIF":4.3,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145456029","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}
Roberto Sabella;Luca Valcarenghi;Jun Shan Wey;Yuki Yoshida
This special issue contains 13 papers, of which 5 are invited, relating to hot topics in the area of optical networks, systems, and technologies for future radio access. These topics are gaining increasing importance in mobile network evolutions and related radio systems and could represent relevant elements of innovation in this evolution.
{"title":"Introduction to the special issue on Optical Networks, Systems, and Technologies for Future Radio Access","authors":"Roberto Sabella;Luca Valcarenghi;Jun Shan Wey;Yuki Yoshida","doi":"10.1364/JOCN.582529","DOIUrl":"https://doi.org/10.1364/JOCN.582529","url":null,"abstract":"This special issue contains 13 papers, of which 5 are invited, relating to hot topics in the area of optical networks, systems, and technologies for future radio access. These topics are gaining increasing importance in mobile network evolutions and related radio systems and could represent relevant elements of innovation in this evolution.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 11","pages":"ONST1-ONST2"},"PeriodicalIF":4.3,"publicationDate":"2025-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11223152","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the Beyond 5G era, networks must deliver not only higher speed and larger capacity but also high reliability and support for massive simultaneous connections. To meet these requirements, we have proposed a smart mobile fronthaul (SMFH) architecture that actively utilizes high-frequency bands and combines analog radio-over-fiber (A-RoF) optical transmission with optically powered remote antennas. Unlike conventional point-to-point A-RoF links, SMFH is distinguished by enabling A-RoF-based mobile fronthaul networking through the insertion of networking functional devices—such as optical switches and optical couplers—into the A-RoF transmission section. In this paper, we construct an A-RoF-based SMFH networking testbed that integrates an open-source software 5G base-station system with an A-RoF transmission module, an optical switch, and an optical coupler in order to verify the networking capabilities of SMFH. Furthermore, we report successful experiments on the testbed that demonstrate key networking capabilities—dynamic serving-cell switching and simultaneous connectivity to multiple cells.
{"title":"Analog radio-over-fiber-based 5G smart mobile fronthaul networking testbed with an open-source software base-station system","authors":"Kojiro Nishimura;Ryuta Murakami;Yoshihiko Uematsu;Satoru Okamoto;Naoaki Yamanaka","doi":"10.1364/JOCN.566706","DOIUrl":"https://doi.org/10.1364/JOCN.566706","url":null,"abstract":"In the Beyond 5G era, networks must deliver not only higher speed and larger capacity but also high reliability and support for massive simultaneous connections. To meet these requirements, we have proposed a smart mobile fronthaul (SMFH) architecture that actively utilizes high-frequency bands and combines analog radio-over-fiber (A-RoF) optical transmission with optically powered remote antennas. Unlike conventional point-to-point A-RoF links, SMFH is distinguished by enabling A-RoF-based mobile fronthaul networking through the insertion of networking functional devices—such as optical switches and optical couplers—into the A-RoF transmission section. In this paper, we construct an A-RoF-based SMFH networking testbed that integrates an open-source software 5G base-station system with an A-RoF transmission module, an optical switch, and an optical coupler in order to verify the networking capabilities of SMFH. Furthermore, we report successful experiments on the testbed that demonstrate key networking capabilities—dynamic serving-cell switching and simultaneous connectivity to multiple cells.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 11","pages":"E144-E154"},"PeriodicalIF":4.3,"publicationDate":"2025-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11220984","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405355","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Vignesh Karunakaran;Ronald Romero Reyes;Behnam Shariati;Johannes Karl Fischer;Achim Autenrieth;Thomas Bauschert
With the dynamic nature of optical service provisioning and network topology reconfigurations, failure identification and management become complex, as the machine learning (ML) model is trained for a specific topology with pre-defined performance metrics. This paper proposes a hybrid ML framework for continuous monitoring and soft failure (SF) localization in a partially disaggregated optical network. The framework combines a distributed unsupervised machine learning approach for per-device monitoring and an inductive graph neural network (GNN) for SF localization. This allows the system to generalize across dynamic network conditions, including optical service reconfigurations and node additions or deletions. To support real-time data collection and provide data plane visibility in the management plane, this work proposes gNMI/gRPC-based telemetry streaming using a unified ONF-TAPI YANG data model, enabling vendor-neutral communication across multi-domain networks. The proposed telemetry streaming outperforms the existing solution by reducing traffic load by a factor of 78.4%, and the inductive GNN-based failure localization maintains an accuracy of 97.4% despite dynamic network reconfigurations.
{"title":"Dynamic network-aware soft failure localization using machine learning in optical networks","authors":"Vignesh Karunakaran;Ronald Romero Reyes;Behnam Shariati;Johannes Karl Fischer;Achim Autenrieth;Thomas Bauschert","doi":"10.1364/JOCN.564177","DOIUrl":"https://doi.org/10.1364/JOCN.564177","url":null,"abstract":"With the dynamic nature of optical service provisioning and network topology reconfigurations, failure identification and management become complex, as the machine learning (ML) model is trained for a specific topology with pre-defined performance metrics. This paper proposes a hybrid ML framework for continuous monitoring and soft failure (SF) localization in a partially disaggregated optical network. The framework combines a distributed unsupervised machine learning approach for per-device monitoring and an inductive graph neural network (GNN) for SF localization. This allows the system to generalize across dynamic network conditions, including optical service reconfigurations and node additions or deletions. To support real-time data collection and provide data plane visibility in the management plane, this work proposes gNMI/gRPC-based telemetry streaming using a unified ONF-TAPI YANG data model, enabling vendor-neutral communication across multi-domain networks. The proposed telemetry streaming outperforms the existing solution by reducing traffic load by a factor of 78.4%, and the inductive GNN-based failure localization maintains an accuracy of 97.4% despite dynamic network reconfigurations.","PeriodicalId":50103,"journal":{"name":"Journal of Optical Communications and Networking","volume":"17 11","pages":"1032-1046"},"PeriodicalIF":4.3,"publicationDate":"2025-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145405301","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}
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：