Predictive Learning Machine for LLN SLA Violation Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Low Power and Lossy Networks (LLNs) face challenges such as lossy links, low bandwidth, and complex network management due to a large number of nodes, making it difficult to predict network behavior using traditional methods, especially in the context of the Internet of Things (IoT) where classic approaches are inefficient and require extensive human processing.
Innovation Solution
The implementation of learning machines that use machine learning algorithms to predict service level agreements (SLAs) in LLNs by estimating network traffic parameters and performance metrics, allowing for proactive management and adaptation of routing topologies without requiring explicit node specification, using a predictive learning machine-based approach that dynamically determines necessary network characteristics and paths for SLA prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional network management approaches are used in LLNs, then manual processing and extensive probing are required, but this leads to high complexity and inefficiency in predicting network behavior
Solution Approach 1:
The patent implements self-service by deploying learning machines at network nodes that autonomously predict network behavior and SLA compliance without requiring manual intervention or extensive external probing. The nodes automatically collect local traffic data, train models, and make predictions, enabling the network to manage itself intelligently at scale.
Solution Approach 2:
The patent replaces mechanical/manual network management approaches with machine learning-based automated systems. Instead of human operators manually analyzing network behavior or running extensive probing sequences, learning machines automatically process network data and predict performance metrics, substituting mechanical processes with intelligent computational systems.
2Measurement precision
If extensive probing is performed to monitor SLA compliance, then measurement precision improves, but this increases network overhead and loss of time
Solution Approach 1:
The patent applies preliminary action by training learning machines offline or in advance using historical network data. Once trained, these models can quickly predict SLA compliance for new traffic patterns without requiring extensive real-time probing. The heavy computational work is performed beforehand, enabling fast predictions during actual network operation.
Solution Approach 2:
The patent uses partial action by leveraging locally available traffic data at network nodes instead of requiring complete network-wide probing. The learning machines process only the subset of data needed for prediction (local traffic statistics), avoiding the excessive time and resources required for comprehensive network probing while maintaining sufficient prediction accuracy.
3Adaptability or versatility
If classic routing protocols are used in LLNs, then implementation is straightforward, but they cannot adapt to changing network conditions and large numbers of nodes
Solution Approach 1:
The patent implements dynamics by integrating learning machines that continuously adapt routing decisions based on changing network conditions. Instead of static routing tables, the system dynamically adjusts routing policies by predicting future network state and SLA compliance, allowing the network to respond flexibly to varying traffic patterns, link conditions, and node failures.
Solution Approach 2:
The patent introduces learning machines as intermediaries between classic routing protocols and network traffic. These learning machines act as intelligent layers that predict network behavior and provide guidance to routing protocols, enabling adaptation to changing conditions without completely replacing proven routing mechanisms. The intermediary translates complex predictions into actionable routing decisions.
Data Source
AI summary
In one embodiment, a request to make a prediction regarding one or more service level agreements (SLAs) in a network is received. A network traffic parameter and an SLA requirement associated with the network traffic parameter according to the one or more SLAs are also determined. In addition, a performance metric associated with traffic in the network that corresponds to the determined network traffic parameter is estimated. It may then be predicted whether the SLA requirement would be satisfied based on the estimated performance metric.