LLM-Powered Intent-Driven Configuration Generation for Multi-Vendor Networks

The diagnostic data from the MTX (Microwave Tokamak Experiment) are acquired and processed by an expandable, distributed, multivendor computer network. The system blends a variety of software into a coordinated, unified, and highly flexible design. Using modular software design techniques, a system stressing distributed processing, portability, and transparent data-access was created. Communication interfaces between modules and separated generic tasks from machine- and application-specific implementations were standardized. For flexible distributed processing, modular, portable software and an LLNL (Lawrence Livermore National Laboratory) facility that provides an interprocess communication system (IPCS) in the multivendor network was used. With transparent data access, any program can access data stored anywhere in the network without knowing the specific location. The computer hardware includes a DEC VAX cluster, HP workstations, and HP desktop computers. Commercial software is being used in addition to packages from MIT, ORNL (Oak Ridge National Laboratory) and LLNL. >

A major challenge in the development of multi-vendor HVDC networks are converter control interactions. While recent publications have reported interoperability issues such as persistent oscillations for first multi-vendor HVDC setups with AC-side coupling, multi-terminal HVDC networks are expected to face similar challenges. To investigate DC-side control interactions and mitigate possible interoperability issues, several methods based on the converters’ and DC network’s impedances have been proposed in literature. For DC network’s impedance modelling, most methods require detailed knowledge of all converters’ design and controls. However, in multi-vendor HVDC networks, converter control parameters are not expected to be shared due to proprietary reasons. Therefore, to facilitate impedance-based stability analyses in multi-vendor MTDC networks, methods that do not require the disclosure of the existing converter controls are needed. Here, detailed impedance measurements can be applied; however, they are time-consuming and require new measurement for a single configuration change. This paper proposes an equivalent impedance calculation method suitable for multi-vendor DC networks, which for available black-box models or converter impedance characteristics can be modularly applied for various network configurations, including different control settings and operating points, while significantly reducing the required time for obtaining an equivalent DC network impedance.

Inter-networking security in a multi-vendor heterogeneous network environment is a critical problem for IT professionals. This paper analyzes the security threats in a multi-vendor network environment and proposes a network security architecture for building a secure distributed network. It also discusses various information security packages including crypto-systems, network authentication, and security control software. The paper attempts to provide readers a systematic way for securing inter-network environment. >

The IETF Path Computation Element (PCE) framework is a prime example of a softwaredefined network virtualisation artifact, originally envisaged about a decade ago for constraintbased routing in the conventional best-effort Internet. Over the years, however, it evolved into the ultimate tool for multi-technology, multidomain, multi-layer, and multi-vendor networking. Carriers have already started to implement PCE-based solutions in their combined Internet IP/MPLS and Wavelength Switched Optical Networks (WSONs). However, significant development and testing is still needed to elevate the PCE concepts from theory to practice. This paper addresses standards, interoperability and deployment of coordinated computation and setup of multi-layer paths from a practical perspective and presents its first implementation in a multi-vendor, interoperable PCE testbed. In our testbed, we try to push the limits of multivendor interoperability, which current research mostly neglects, and thereby identify missing links in the current research and standards necessary to implement and deploy inter-operable multi-layer systems. We point to the key areas that the research and industrial communities will need to address in making the proposed solutions practical, which in turn will drive the upcoming efforts in software-defined networking with inter-operable PCE-based architectures.

We demonstrate the emergent integration and control of surviving multi-vendor optical systems after huge-disasters. Interconnecting surviving multi-vendor networks is not only feasible but also the most cost-efficient approach to the quick disaster recovery.

Network down-time can be very expensive to business customers like those in the financial world. Self-Healing Networks (SHNs) immediately detect service-affecting problems and respond to ensure service continuity to minimize the down-time experienced by a customer. Three approaches to SHNs are in use: route diversity, dualfed rings, and reconfigurable rings. Route diversity uses the protection switching capabilities integrated in the hardware of AT&T's lightwave systems (like FT Series G and DDM-1000 in asynchronous networks and FT-2000 and DDM-2000 in SONET networks) to provide instantaneous restoration hm performance degradation or facility failures by routing the Service and protection lines on physically diverse paths. Dud-fed rings also provide instantaneous restoration from performance degradation or facility failures on a DS1 or DS3 basis independent of the transport rate and format, and the vendor of the lightwave system. In reconfigurable rings, digital cross-connect systems (such as AT&T's DACS III-2000 and DACS IV2000) are used to reroute the tr& around failures. Reconfigurable rings provide SHNs in today's asynchronous network as well as in mixed async/SONEI' and multivendor networks by managing gateway vehicles. Operations of a multi-location multi-vendor network in the present and evolving SONET environment will be greatly facilitated through the use of a network management and control system such as DACScanm 2000 controller, both for day-today centralized provisioning and maintenance, and for restoration from network disasters. SHN capabilities become more important in the SONET network which is by definition fiber-based and canies higher bit rates than today. The SHN approaches must be easily deployed during the transition to SONET as well as in full SONET networks. AT&T's 2000 Product Family provides complete interworking with network elements currently deployed and maintains transparency for today's SHN features in the SONET network. SHN capabilities will be improved in the SONET network by leveraging off the SONET standard itself. Examples of additional capabilities that enhance the present self-healing network in the 2000 Product Family include faster restoral time, end to end performance monitoring, and enhanced fault isolation capabilities. Other capabilities, such as

Provably efficient security-aware service function tree composing and embedding in multi-vendor networks

Network telemetry based on data models is expected to become the standard mechanism for collecting operational data from network devices efficiently. But the wide variety of standard and proprietary data models along with the different implementations of telemetry protocols offered by network vendors become a barrier when monitoring heterogeneous network infrastructures. To facilitate the integration and sharing of context information related to model-driven telemetry, this work proposes a semantic network inventory that integrates new information models specifically developed to capture context information in a vendor-agnostic fashion using current standards defined for context management. To automate the integration of this context information within the network inventory, a reference architecture is designed. Finally, a prototype of the solution is implemented and validated through a case study that illustrates how the network inventory can ease the operation of model-driven telemetry in multi-vendor networks.

We demonstrate a multi-vendor network with various optical switching technologies and client devices emulating a real-world environment. End-to-end and automated provisioning is achieved using standard-based GMPLS control plane and a proprietary optical transmission control plane. (2 pages)

For the recovery of core optical transport networks after disaster, in this paper, we introduce an emergency optical network design problem. In this problem, we take account of both the interconnection of the surviving multi-vendor resources and the utilization of the emergency portable EDFAs. The surviving multi-vendor optical nodes [e.g., reconfigurable optical add/drop multiplexer (ROADM) or wavelength cross-connect (WXC)] and fiber links are the first available resources in disaster recovery. The portable EDFAs are the added emergency resources for replacing the optical nodes which are either destroyed or down due to the power outage in the disaster so as to satisfy the requirement of long distance optical signal transmission. For this design problem, we model an emergency network planning using the integer linear programming (ILP) formulation such that the locations for the emergency interconnection of the multi-vendor networks and for the placement of the portable EDFAs are optimally selected. Evaluations are conducted showing the benefits of the multi-vendor interconnection approach and the utilization of the emergency portable EDFAs.

Technical advances in multivendor network management capabilities allow customers to effectively manage their networks. Packaged offerings such as NetView® Extra simplify the ability to take advantage of these new capabilities. This paper describes the multivendor environment, customer network management requirements, IBM's initial approach to responding to these requirements, and enhancements needed to provide additional management offerings that automatically handle failures, including detection, bypass and recovery, vendor notification, and restoration of the repaired resource into service.

The increasing complexity of multi-vendor network infrastructures presents significant challenges in maintaining robust security. Traditional network security approaches are often insufficient to address the dynamic and sophisticated nature of modern cyber threats. This study proposes a conceptual model for network security automation, leveraging Artificial Intelligence (AI)-driven frameworks to enhance resilience across multi-vendor environments. The model integrates advanced AI techniques, including machine learning, predictive analytics, and natural language processing, to automate threat detection, response, and prevention. A central feature of the proposed framework is its ability to harmonize security protocols and policies across diverse vendor systems, enabling seamless interoperability and real-time threat intelligence sharing. The model incorporates automated anomaly detection to identify irregular network behaviors and a risk-based decision-making engine to prioritize and mitigate threats proactively. By employing AI, the model ensures adaptive learning, allowing the system to evolve with emerging threats and changes in network architecture. Key components of the framework include a centralized security orchestration layer, vendor-agnostic APIs, and a unified dashboard for real-time monitoring and analytics. This approach enhances operational efficiency by reducing manual intervention, accelerating incident response times, and minimizing false positives. Furthermore, the model emphasizes compliance with industry standards and regulatory frameworks, providing organizations with a robust foundation for secure multi-vendor network management. Preliminary findings suggest that adopting this AI-driven security automation model significantly improves threat resilience, operational scalability, and resource optimization in complex network environments. The study concludes by highlighting the potential of such frameworks to redefine network security practices, offering a transformative approach to managing risks in increasingly interconnected and heterogeneous infrastructures.

In the past, network management was primarily a centralized endeavor carried out by a virtual priesthood of technicians who stared at arcane command-line screens all day. Now management tasks are distributed to the most appropriate machine, and various tasks executing on different platforms can now be integrated. The centralized point-to-point connections used in the past have been replaced with large clouds of packet/cell switching networks or dual-purpose voice/data ISDN connections. In fact, opening voice circuits to the public has enabled many companies to implement combined voice/data networks. These have resulted in highly effective multi-vendor networks. The primary job of network management is to monitor and report on the status of the entire network. At NetCare® Network Management Services, we track the status of every component in the network, regardless of who the manufacturer is or what type of network it is operating. This paper describes our successful approach to multi-vendor network management.

Network automation is the current challenge that operators face to accelerate end-to-end service delivery and improve network operations in the context of rapidly growing capacity needs. Optical networks have long been built using vendors’ turn-key solutions, resulting in complexity for the operator to automate its multi-vendor networks in a consistent way. Questioning the vendors’ lock-in in disaggregating the optical transmission systems (i.e., splitting the system into parts, where each part can be provided by a different vendor) provides advantages (e.g., cost) but also additional complexity. Openness and standardization appear as key for these multi-vendor scenarios. The path is not easy, but it is not a matter of choice: opening and interoperability are obligations that we face because of the need for automation. The journey that we detail in this paper is the one that we think will be sustainable from our side. The opposite, i.e., continue building vendors’ silos and staking per-vendor specialized applications, is not sustainable. We propose a step-by-step approach, starting with the non-disaggregated situation, followed by partial and full disaggregation architectures to the last evolution towards data centric networking. We present and discuss implementations that Orange has been contributing to and identify some gaps the industry should address. We show that current works in the communities of open source, open initiatives, and standardization bodies are addressing all these steps, and in this respect can accelerate the deployment of automation solutions in current and future optical transport networks.

SNMPv1 is very successful due to its simplicity, flexibility, and extensibility. However, simplicity can also be a drawback. In SNMPv1, a manager can only retrieve one piece of information at a time, and an agent merely reports one error to its manager, though there may be several errors having occurred during the protocol operation. These two interactions enlarge the number of protocol exchanges between a manager and an agent. SNMPv2 permits the retrieval of large data blocks, but the errors can only be handled one by one. These errors continually occur in a multivendor network. This paper proposes an error handling method in SNMPv2 protocol operations. In this method, all errors within a received PDU (Protocol Data Unit) can be detected by the agent simultaneously.