Multi-Layer Bayesian Network for Communication Abnormality Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods are unable to comprehensively construct a causal model that accounts for the relationship with multiple services provided by a communication network system, making it difficult to estimate the location or cause of abnormalities across various services.
Innovation Solution
A model construction apparatus that collects both network and service-related data to construct a multi-layer Bayesian network causal model, enabling the estimation of abnormal services within the communication network system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional rule-based methods are used to construct causal models, then the model construction is simple and fast, but the model cannot comprehensively account for multiple services provided by the communication network system
Solution Approach 1:
The patent segments the causal model into multiple layers: a first layer representing the communication network system and a second layer representing multiple services. This segmentation allows the model to comprehensively cover multiple services while maintaining manageable complexity through structured organization of different system aspects.
Solution Approach 2:
The patent adds a service layer dimension to the traditional network-centric causal model. By constructing a multi-layer Bayesian network that includes both network system nodes and service nodes, the model transitions from a single-dimension network view to a multi-dimensional view that simultaneously captures network infrastructure and service dependencies.
2Measurement precision
If a causal model is constructed to estimate abnormality location or cause, then the estimation accuracy is improved, but the model construction requires comprehensive data collection and complex Bayesian network setup
Solution Approach 1:
The patent divides data collection into two distinct streams: first observed data from the communication network system and second observed data from multiple services. This segmentation allows systematic collection of relevant data from different sources while organizing the complex Bayesian network into manageable layers with specific observation nodes for each data type.
Solution Approach 2:
The multi-layer Bayesian network structure serves multiple functions simultaneously: it models normal operational relationships, detects abnormalities, estimates abnormality locations, and identifies causes. The same structured framework handles both network-level and service-level analyses, reducing overall system complexity through unified multi-functional design.
Data Source
AI summary
A model construction apparatus according to an embodiment includes a first collection unit that collects pieces of first observed data related to a communication network system that is a target for estimation of a location or a cause of an abnormality; a second collection unit that collects pieces of second observed data related to a plurality of services provided by the communication network system; and a model construction unit that constructs a causal model for estimating the location or the cause of the abnormality and an abnormal service among the plurality of services, using the pieces of first observed data and the pieces of second observed data.