Dynamic Router Dead Interval for Faster Network Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network protocols, such as OSPF, take too long to recover from network element failures due to high bandwidth and large data loss during unstable conditions, necessitating faster failure detection and recovery mechanisms.
Innovation Solution
Dynamic variation of the router dead interval based on congestion levels, with each router maintaining a separate dead interval for neighboring routers, and using modified Hello messages to include CPU utilization, allowing for quicker declaration of congested routers as down and reducing recovery time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a fixed router dead interval is used in network protocols like OSPF, then the protocol maintains stability and avoids false alarms, but the recovery time from network element failures becomes too long
Solution Approach 1:
The patent applies dynamics by making the router dead interval variable rather than fixed. The dead interval dynamically adjusts based on measured network congestion levels: it increases when congestion is detected to avoid false alarms, and decreases when the network is stable to enable faster failure detection. This resolves the contradiction between stability and recovery speed by adapting the parameter to current network conditions.
Solution Approach 2:
The patent changes the parameter of the router dead interval from a static value to a dynamic value that varies with network congestion. By monitoring congestion metrics and adjusting the dead interval accordingly, the system achieves both fast failure detection (when congestion is low) and false alarm prevention (when congestion is high), thus resolving the contradiction between recovery time and network stability.
2Loss of time
If the router dead interval is reduced to detect failures faster, then recovery time improves, but false alarms increase during congested periods
Solution Approach 1:
The patent implements feedback by continuously monitoring network congestion levels and using this information to adjust the router dead interval. When congestion is detected, the dead interval is increased to prevent false alarms; when the network is stable, it is decreased to enable faster failure detection. This feedback mechanism resolves the contradiction by making the detection speed adaptive to current network conditions.
Solution Approach 2:
The patent makes the dead interval dynamic rather than static, allowing it to respond to changing network conditions. By adjusting the interval based on real-time congestion measurements, the system achieves fast failure detection during stable periods while avoiding false alarms during congested periods, thus resolving the contradiction between detection speed and reliability.
3Device complexity
If a single dead interval is used for all neighboring routers, then the system is simpler to manage, but it cannot adapt to different congestion levels of different neighbors
Solution Approach 1:
The patent applies segmentation by dividing the single dead interval into multiple separate dead intervals, one for each neighboring router. Each neighbor-specific interval can be independently adjusted based on that neighbor's congestion level, allowing the system to adapt to different network conditions for different neighbors while maintaining manageable complexity through automated measurement and adjustment.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
An Internet Protocol router device (10) for use in a data network comprising a plurality of nodes (10, 20, 30), wherein packets of data are sent from one node in the network to another on a hop by hop basis. The router includes: a data store (14) for storing a routing table indicating the next hop destination for a received packet on the basis of the ultimate destination specified by the received packet; a digital processor (13) operable to maintain the routing table on the basis of communications received from neighbouring router devices (20, 30) within the network; and a timer (15) for timing the length of time elapsed since last receiving a communication, or a communication of a type which is one of a subset of the possible types of communication, from each of one or more of the router device's neighbours; wherein the data store (14) is further operable to store in respect of each of said one or more of the router device's neighbours a router dead interval; and the digital processor (13) is further operable to compare said length of time elapsed with the corresponding router dead interval in respect of each of said one or more of the router device's neighbours and to ascertain that the corresponding router device is no longer operating correctly if the comparison indicates that a length of time equal to or greater than the router dead interval has elapsed without receiving a said communication, or a said communication of a type which is one of a subset of the possible types of communication, from said neighbouring router device; the router device being characterised in that the digital processor (13) is further operable to dynamically vary the or each router dead interval in accordance with one or more properties of the network.