Intelligent Network Topology Mapping via Probe Packet Parameter Variation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network topology mapping methods, such as traceroute, fail to discover all paths in networks that implement flow-based load sharing, as they do not vary packet parameters and thus cannot trigger flow-based load balancing, leading to incomplete network topology views.
Innovation Solution
The method involves using probe packets with intelligently varied flow parameters to exercise each possible load shared path, maintaining a list of flows reaching each hop, and selecting parameters to generate probe packets that trigger flow-based load sharing, reducing the number of probe packets needed to map network paths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional traceroute methods are used with fixed packet parameters, then the method is simple to implement, but it fails to discover all network paths in flow-based load sharing networks
Solution Approach 1:
The patent systematically varies packet parameters including Time To Live (TTL), protocol type, IP header fields, and upper layer ports to trigger flow-based load sharing mechanisms. This parameter variation enables the discovery of multiple network paths that would otherwise remain hidden when using traditional fixed-parameter traceroute methods.
Solution Approach 2:
The patent implements dynamic parameter selection where packet parameters are adaptively adjusted based on previous traceroute results and observed network behavior. This dynamic approach allows the system to efficiently explore the network topology by learning which parameter combinations reveal additional paths.
2Measurement precision
If multiple packet parameters are varied to trigger flow-based load sharing, then complete network topology can be discovered, but the number of probe packets increases significantly
Solution Approach 1:
The patent employs a two-phase approach: first using a limited set of probe packets with standard parameters to establish baseline topology, then selectively applying parameter variations only where needed to discover additional load-shared paths. This avoids the excessive action of uniformly varying all parameters for all probes.
Solution Approach 2:
The patent segments the topology discovery process into multiple phases: initial discovery phase, refinement phase, and verification phase. Each phase uses appropriately tailored packet parameters, allowing efficient progression from basic topology mapping to complete path discovery without overwhelming the network in any single phase.
3Measurement precision
If extensive probe packets are transmitted to discover all paths, then all network paths can be identified, but network traffic for discovery increases and may overwhelm the network
Solution Approach 1:
The patent implements periodic probing strategies where probe packets are transmitted at controlled intervals rather than continuously. This periodic action allows network traffic to settle between probe bursts, reducing the risk of overwhelming network infrastructure while still achieving complete topology discovery over time.
Solution Approach 2:
The patent performs preliminary topology discovery using minimal probe packets to establish baseline network paths and identify potential load-sharing points. This preliminary action informs subsequent probing strategies, allowing the system to focus probe traffic only on critical paths and avoid unnecessary network load.
4Ease of operation
If traditional traceroute is used without varying packet parameters, then the implementation is straightforward, but it cannot trigger flow-based load balancing to reveal multiple paths
Solution Approach 1:
The patent introduces parameter variation as an intermediary mechanism that mediates between simple traceroute implementation and complete path discovery. By systematically varying packet parameters, the system triggers flow-based load balancing mechanisms in the network, causing packets to take different paths and thereby revealing the complete topology without complicating the core traceroute functionality.
Data Source
AI summary
A method for intelligent network topology mapping includes identifying network paths between a source and a destination, wherein identifying the network paths includes, for a hop in network topology: selecting a number of probe packets for revealing an expected number of next hops reachable from the hop; maintaining a list of flows reaching the hop; selecting, from the list, flow parameters for each of the probe packets; generating the number of probe packets and including the selected flow parameters in the probe packets; transmitting the number of probe packets to the hop; receiving responses to the probe packets; and recording network addresses of next hops revealed by the responses. The method further includes generating, from the network addresses of next hops, a network topology map illustrating the network paths.