SDN Traffic Digest Agents for Packet Loss Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current monitoring tools in data centers are inadequate for quickly detecting packet losses with sufficient detail while maintaining low overhead, and there is no efficient solution for fast and safe recovery using alternate paths in software-defined networks (SDNs).
Innovation Solution
A system and method that captures traffic at source and destination domains, creates traffic digests with unique packet identifiers, compares them to detect losses, identifies alternative paths, and resynchronizes traffic using a quantum and analytics-enabled approach, with machine learning algorithms for policy generation and dynamic path management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If generic monitoring tools are used to capture all types of network events, then network event detection coverage is improved, but processing overhead increases and detection speed decreases
Solution Approach 1:
The patent segments network monitoring into multiple specialized agents (loss detection agents, performance monitoring agents, security agents) distributed across network nodes. Each agent focuses on specific event types, enabling parallel processing and faster detection while maintaining comprehensive coverage through coordinated monitoring.
Solution Approach 2:
The patent introduces a centralized controller as an intermediary that receives data from multiple specialized agents and coordinates their operations. This mediator consolidates detection results, manages agent communications, and orchestrates recovery actions, improving overall system efficiency and reducing redundant processing.
2Measurement precision
If detailed packet loss detection is implemented to identify individual lost packets, then measurement precision is improved, but processing overhead increases
Solution Approach 1:
The patent extracts only the essential identifying information from packets (such as packet identifiers, sequence numbers, or hash values) rather than processing complete packet contents. This extraction approach enables precise tracking of individual packets while significantly reducing the computational overhead associated with analyzing full packet data.
Solution Approach 2:
The patent creates simplified copies or representations of packet information (such as metadata, headers, or digest values) that contain sufficient identifying characteristics without requiring full packet processing. These copies enable efficient comparison and detection of lost packets with minimal processing requirements.
3Ease of operation
If SDN controllers are used for failure detection and path establishment, then network control capability is improved, but recovery time increases due to round-trip delays
Solution Approach 1:
The patent implements preliminary actions by pre-computing and caching alternative paths before failures occur. When a failure is detected, the system can immediately switch to pre-prepared paths without requiring time-consuming real-time computation or multiple round-trip communications with the controller, thus reducing recovery time while maintaining centralized control.
Solution Approach 2:
The patent introduces dynamic elements by enabling distributed agents to make local decisions about path switching based on real-time conditions, rather than requiring static, centralized control for every decision. This dynamic approach allows faster local responses to failures while the controller maintains overall network policy and coordination.
4Difficulty of detecting and measuring
If proprietary management interfaces are used for failure detection, then detection capability is improved, but system adaptability decreases
Solution Approach 1:
The patent implements universal agents and interfaces that can operate across multiple SDN controllers and network environments. The agents use standardized protocols and data formats that are compatible with different SDN implementations, enabling the system to detect failures and perform recovery across diverse network infrastructures without requiring proprietary solutions for each vendor.
Data Source
AI summary
A method includes capturing traffic at a source domain and a destination domain, creating a first traffic digest comprising unique packet identifiers representing traffic captured at the source domain, creating a second traffic digest comprising unique packet identifiers representing traffic captured at the destination domain, comparing the first traffic digest to the second traffic digest, if a loss is detected based on the comparing step, identify an alternative path between the source domain and the destination domain, switching processing to the alternative path, and resynchronizing traffic between the source domain and the destination domain on the alternative path.