SDN Adjacency Detection via Performance Monitoring Orchestration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing physical adjacency detection methods in networks are limited by being either single-layer or proprietary, making them unsuitable for multi-vendor, multi-layer deployments, and often disrupt traffic to function effectively.
Innovation Solution
A method and system that use Software Defined Networking (SDN) to detect physical interconnections between network devices by orchestrating changes in performance monitoring data, such as modifying transmit power with a unique fingerprint, allowing for in-service detection without traffic disruption across multiple layers and vendors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional adjacency detection methods (TTI, LLDP, proprietary solutions) are used, then detection capability is achieved at a single layer or vendor, but the system cannot operate in multi-vendor, multi-layer networks
Solution Approach 1:
The patent implements a universal adjacency detection mechanism that operates across multiple network layers (Layer 1 optical, Layer 2 Ethernet, Layer 3 IP) and is compatible with multi-vendor equipment. The system uses a standardized performance monitoring interface that can detect adjacencies regardless of the specific vendor or layer, making the detection capability universally applicable across heterogeneous network environments.
Solution Approach 2:
The patent introduces an intermediary performance monitoring system that acts as a mediator between different network layers and vendors. This intermediary layer captures performance monitoring data from various sources (optical signals, Ethernet frames, IP packets) and processes them through a unified detection algorithm, enabling cross-layer and cross-vendor adjacency detection without requiring vendor-specific implementations.
2Measurement precision
If laser connections are turned on and off to detect adjacency, then detection accuracy is improved, but traffic is disrupted and the method is not in-service
Solution Approach 1:
The patent enables continuous adjacency detection by utilizing ongoing performance monitoring data that is already being collected during normal network operation. Instead of interrupting traffic to perform detection, the system continuously analyzes performance monitoring parameters (such as optical power, signal quality, frame error rates) that are naturally generated during active data transmission, allowing in-service detection without traffic disruption.
Solution Approach 2:
The system leverages the network's own operational data for detection purposes. Performance monitoring data that is already being collected for network management and troubleshooting is repurposed for adjacency detection, eliminating the need for separate detection activities that would disrupt traffic. The network's normal operation itself provides the detection signals.
3Adaptability or versatility
If performance monitoring data is orchestrated with unique fingerprints, then vendor-independent detection is achieved, but system complexity increases
Solution Approach 1:
The patent achieves vendor-independent detection by monitoring changes in performance monitoring parameters rather than relying on vendor-specific identifiers. The system detects adjacency by observing characteristic patterns in parameters such as optical power levels, signal quality metrics, and error rates that change in predictable ways during normal operation, making the detection vendor-agnostic while avoiding complex orchestration of unique fingerprints.
Data Source
AI summary
Adjacency detection systems and methods configured to detect physical interconnection between two devices in a network include communicating data with one or more controllers and data sources associated with the network; for a given port in the network to detect a corresponding port physically connected thereto, causing orchestration of changes in performance monitoring data for the given port over a period of time; and analyzing the data subsequent to the orchestration to detect the corresponding port. The systems and methods can operate over multiple network layers, with different vendor equipment, independent of network protocols, and in-service.