Network Availability Prediction Using Recurrent Neural Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network monitoring approaches are reactive, leading to prolonged downtime as they require a network device to be deficient before generating an alert, resulting in manual and time-consuming troubleshooting processes to restore availability.
Innovation Solution
A computer-implemented method using a trained recurrent neural network (RNN) to predict forward-looking attribute values for network devices, enabling preemptive troubleshooting and remediation actions based on predicted network availability data points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If reactive monitoring tools are used to detect network device deficiencies, then network device availability can be monitored, but the response time to restore availability is prolonged due to manual troubleshooting processes
Solution Approach 1:
The system performs preliminary actions by training machine learning models on historical network data to predict future device deficiencies before they occur. This enables preemptive remediation actions to be taken before actual failures happen, eliminating the need for reactive manual troubleshooting and significantly reducing restoration time.
Solution Approach 2:
The system implements self-service by automating the entire monitoring and remediation process through machine learning models that automatically detect patterns, predict failures, and trigger remediation actions without human intervention. This eliminates manual troubleshooting processes and enables the system to self-correct issues, drastically reducing the time to restore network availability.
2Ease of repair
If manual troubleshooting processes are used to restore network availability, then remediation can be performed, but the process is tedious and time-consuming
Solution Approach 1:
The system automates the remediation process by using machine learning models to automatically analyze network data, predict deficiencies, and execute remediation actions without human intervention. This eliminates the tedious manual troubleshooting process entirely, making repair easier and faster by enabling the system to self-diagnose and self-correct issues autonomously.
Solution Approach 2:
The system implements continuous feedback loops where machine learning models constantly monitor network device performance, compare actual data against predicted patterns, and automatically adjust remediation strategies based on real-time conditions. This automated feedback mechanism eliminates manual troubleshooting while ensuring appropriate remediation actions are taken, reducing both the complexity and time of the repair process.
3Measurement precision
If traditional monitoring approaches are used, then network device attributes can be tracked, but preemptive actions cannot be taken before degradation occurs
Solution Approach 1:
The system performs preliminary actions by using machine learning models to analyze historical network attribute data and predict future device deficiencies before they manifest as actual failures. This enables preemptive remediation actions to be taken while the network is still operational, maintaining higher reliability by preventing degradation rather than merely tracking it.
Solution Approach 2:
The machine learning models serve as an intermediary between raw network attribute data and remediation actions. Instead of directly responding to actual failures, the models predict future states and enable preemptive interventions, bridging the gap between current monitoring capabilities and future failure prevention, thereby improving network availability maintenance.
Data Source
AI summary
In various embodiments, a prediction subsystem automatically predicts a level of network availability of a device network. The prediction subsystem computes a set of predicted attribute values for a set of devices attributes associated with the device network based on a trained recurrent neural network (RNN) and set(s) of past attribute values for the set of device attributes. The prediction subsystem then performs classification operation(s) based on the set of predicted attribute values and one or more machine-learned classification criteria. The result of the classification operation(s) is a network availability data point that predicts a level of network availability of the device network. Preemptive action(s) are subsequently performed on the device network based on the network availability data point. By performing the preemptive action(s), the amount of time during which network availability is below a given level can be substantially reduced compared to prior art, reactive approaches.