Discrete-Event Network Simulation for Extended Failure Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network simulators fail to account for the complexity of network failures over extended periods, which can affect the network's ability to meet service level objectives, particularly due to device failures, fiber damage, and large-scale disasters.
Innovation Solution
A discrete event network simulation (DENS) system that models failure events and repair events using stochastic processes, such as exponential distributions for failure times and fixed repair times, to simulate network behavior and generate metrics on traffic flow, incorporating network topology and traffic engineering techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional network simulators are used for short-term packet-level simulation, then simulation speed is fast, but the ability to account for network failure complexity over extended periods is insufficient
Solution Approach 1:
The simulation system is divided into two distinct components: a traditional packet-level simulator for short-term high-speed simulation, and a new discrete event simulator for long-term failure modeling. Each component handles specific time scales and failure types, allowing the system to achieve comprehensive reliability modeling without requiring a single complex simulator to handle all scenarios.
Solution Approach 2:
The patent introduces a new dimension of simulation by adding temporal scale as a differentiating factor. Instead of trying to improve the existing short-term simulator, a separate discrete event simulation layer is added that operates on extended time scales, modeling failures, repairs, and network evolution over months or years rather than seconds or minutes.
2Measurement precision
If discrete event simulation is implemented to model extended network failures, then network risk estimation accuracy improves, but computational time and resources increase
Solution Approach 1:
The discrete event simulator pre-generates failure scenarios, repair events, and network state transitions before actual risk assessment is needed. By pre-computing failure patterns and their impacts on network topology and traffic flow, the system establishes baseline risk metrics that can be quickly queried without requiring lengthy simulations each time risk assessment is needed.
Solution Approach 2:
The system creates simplified models and representations of complex network failure scenarios. Instead of simulating every individual packet and device state change in detail, the discrete event simulator uses aggregated models that capture essential failure patterns and their impacts, reducing computational requirements while maintaining risk estimation accuracy.
3Reliability
If comprehensive failure modeling is implemented including device failures, fiber damage, and natural disasters, then network reliability assessment improves, but simulation complexity and data requirements increase
Solution Approach 1:
The failure modeling approach applies different levels of detail and complexity to different parts of the network model. Critical network components and failure modes that have significant impact on reliability are modeled with high detail, while less critical elements use simplified models. This allows comprehensive coverage of failure types without uniformly increasing complexity across the entire system.
Data Source
AI summary
Aspects of the disclosure provide for discrete event network simulation (DENS) including failure modeling, network simulation, and metric reporting. A failure modeler can generate and model expected failure events. A network simulator can implement and execute the simulation processes with the failure events and calculate the flow availability of the network during the simulation processes. A report generator can generate various metrics from the simulation results.