AI Network Traffic Classification With Selective Model Synthesis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing neural network processing engines for machine learning face challenges in synthesizing complementary inference results from multiple models with different architectures and datasets, making it difficult to effectively classify communication network traffic and channel impairments.
Innovation Solution
A system combining single-class and multi-class artificial intelligence models with fuzzy logic operations to evaluate and synthesize inference results, using edge devices and cloud-based nodes for efficient classification of network traffic and channel impairments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple different neural network models with different architectures and datasets are used for classification, then the accuracy and comprehensiveness of inference results improve, but the complexity of synthesizing and managing these models increases
Solution Approach 1:
The patent segments the classification task into multiple specialized neural network models, each targeting specific classes or aspects of the classification problem. These models are trained on different datasets with different architectures suited to their specific functions, allowing high accuracy in each segment while managing overall system complexity through modular organization.
Solution Approach 2:
The patent introduces an intermediary synthesis mechanism that combines the inference results from multiple specialized models. This intermediary layer manages the complexity of integrating diverse model outputs by providing a standardized interface and coordination logic, enabling accurate classification without requiring direct management of all model interactions.
2Reliability
If multiple models are evaluated for every data packet flow, then the classification reliability improves, but the computational resources and processing time increase
Solution Approach 1:
The patent applies partial action by selectively evaluating models based on the specific characteristics of each data packet flow. Rather than running all models on every packet, the system evaluates only the subset of models relevant to the particular classification task at hand, maintaining reliability for required classifications while improving overall processing throughput by avoiding unnecessary computations.
3Adaptability or versatility
If comprehensive classification of all network traffic types is performed, then the versatility of the system improves, but the computational resources required increase
Solution Approach 1:
The patent creates a universal classification system where a single framework can handle multiple types of network traffic classification through its collection of specialized models. This multi-functional approach allows the system to adapt to diverse classification needs without requiring separate dedicated systems for each traffic type, achieving versatility while managing resource consumption through selective model deployment.
Data Source
AI summary
A system for controlling network traffic or responding to communication channel impairment. The system includes a number of circuits configured to perform classification using a number of artificial intelligence models trained to provide an inference related to one class and an artificial intelligence model trained to provide an inference related to several classes. Models are connected within an architecture providing for selective execution of one or more of the individual models. Classification results are used to perform actions to affect flow of information in a communications system.