Monitoring system and method for intelligent ai network performance based on epirb

Abstract

This invention relates to an EPIRB-based intelligent AI network performance monitoring system and method. By collecting network data from multiple sources and utilizing artificial intelligence technology for analysis and prediction, it optimizes network operation and maintenance. The monitoring system includes a data collection module, an analysis module, a notification module, and a user interface. The data collection module collects environmental and network data; the analysis module, connected to the data collection module, analyzes and optimizes the data and operations required during network operation and maintenance, generating multiple analysis results. The notification module communicates with the analysis module based on the analysis results and sends corresponding notifications or maintenance instructions to the user interface. Thus, this invention utilizes generative AI models and machine learning to continuously improve accuracy in fault handling.

Claims

1. An EPIRB-based intelligent AI network performance monitoring system, which collects multiple network data from various sources and uses artificial intelligence technology for analysis and prediction to optimize network operation and maintenance, the monitoring system comprising: A data collection module includes a plurality of sensors and a data acquisition unit. The sensors are integrated into an Emergency Position Indicating Radio Beacon (EPIRB). Each sensor continuously monitors the surrounding environment to collect a plurality of environmental data. The data acquisition unit retrieves the network data from a telecommunications operator's management website via a web crawler. The network data includes a plurality of historical operation data and a plurality of real-time operation data from multiple end users. A database is built from the environmental data and the network data. An AI machine analysis module is communicatively connected to the data collection module. The AI ​​machine analysis module is used to analyze and optimize the data and operations required in the network operation and maintenance process to generate a plurality of analysis results. The AI ​​machine analysis module includes a preprocessing unit, which is communicatively connected to the data collection module, to perform preprocessing such as cleaning, filtering, or standardization on the collected environmental data and network data to form a plurality of preprocessed data. A feature extraction unit, connected to the preprocessing unit via an interface, extracts multiple feature values ​​from the preprocessed data; and a machine learning unit, connected to the feature extraction unit, analyzes and predicts data, network performance, and user behavior required for network operation and maintenance to obtain multiple analysis results. It compares the mobile network usage of multiple end users at a specific time, and when the usage is less than a specific value, performs intelligent analysis through the machine learning unit to generate corresponding optimization suggestions or operation instructions; a notification module, connected to the AI ​​machine analysis module, issues user classification, anomaly notifications, or fault maintenance instructions based on the analysis results; and a user interface, connected to the notification module, displays the analysis results, user classification, anomaly notifications, or fault maintenance instructions; wherein the historical operation data includes: multiple historical abnormal event data, multiple historical base station trajectory data, multiple service item data, multiple usage data, multiple historical DNS domain history data, and multiple historical test speed data. The data acquisition unit collects multiple historical base station trajectory data of these terminal users from the telecommunications operator's management website through the crawler program, stores and manages these historical base station trajectory data and the real-time operation data related to the base station trajectory, and the AI ​​machine analysis module extracts these historical base station trajectory data and performs calculations with these real-time operation data through a graph neural network, a convolutional neural network or a recurrent neural network to analyze whether there is a location analysis result of the same regional cluster, and transmits the location analysis result to the notification module.

2. The monitoring system as described in claim 1, wherein the data acquisition unit retrieves multiple historical data of abnormal events of the terminal users from the telecommunications operator's management website through the crawler program, stores and manages the historical data of abnormal events and the real-time operational data related to performance maintenance, the AI ​​machine analysis module retrieves the historical data of abnormal events and combines it with the real-time operational data to perform calculations through the convolutional neural network, analyzes the patterns or correlations in the historical data of abnormal events related to the real-time operational data, generates an abnormal event analysis result, and transmits the abnormal event analysis result to the notification module.

3. The monitoring system as described in claim 1, wherein the data acquisition unit retrieves service item data and usage data of the end users from the telecommunications operator's management website through the web crawler, stores and manages the service item data, usage data, and real-time operational data related to the service items and usage, and the AI ​​machine analysis module retrieves the service item data and usage data, and performs calculations with the real-time operational data through a decision tree or a random forest to analyze whether the end users use the same type of telecommunications service, and whether there are any service analysis results related to a specific mobile phone model, and transmits the service analysis results to the notification module.

4. The monitoring system as described in claim 1, wherein the data acquisition unit collects the historical DNS domain history data of the end users from the telecommunications operator's management website through the crawler program, stores and manages the historical DNS domain history data and the real-time operational data related to the DNS domain history data, the AI ​​machine analysis module acquires the historical DNS domain history data of the end users, and performs calculations with the real-time operational data through a time series analysis, an association rule learning or a clustering algorithm to analyze whether the end users experience the same network problems or performance anomalies when using the same application or accessing the same domain, and transmits the domain analysis result to the notification module.

5. The monitoring system as described in claim 1, wherein the data acquisition unit collects historical test speed data of the terminal users from the telecommunications operator's management website through the crawler program, stores and manages the historical test speed data and the real-time operating data related to the test speed, the AI ​​machine analysis module extracts the historical test speed data of the terminal users in the database, and performs calculations with the historical operating data through a regression analysis or a principal component analysis to analyze the network speed analysis results of the dependencies or similarities between different low-speed network connections, and transmits the network speed analysis results to the notification module.

6. A method for monitoring the performance of an intelligent AI network based on EPIRB, which optimizes network operation and maintenance by collecting multiple network data from various sources and using artificial intelligence technology for analysis and prediction. The monitoring method includes the following steps: collecting multiple environmental data points by continuously monitoring the surrounding environment and collecting such environmental data through multiple sensors integrated in an emergency location indication wireless telecommunications sign device; collecting multiple network data points by retrieving such network data from a telecommunications operator's management website through a web crawler program in a data acquisition unit, wherein such network data includes multiple historical operation data and multiple real-time operation data of multiple end users; establishing a database by integrating such environmental data and such network data to establish a database; and data preprocessing by transmitting such environmental data and such network data from the database to a preprocessing unit in an AI machine analysis module to clean, filter, or standardize the data to form multiple preprocessed data points. The feature extraction step involves sending the preprocessed data to a feature extraction unit to extract multiple feature values ​​from the preprocessed data. These features are used to analyze and optimize the feature parameters required during network operation and maintenance. The machine learning analysis step involves inputting these feature values ​​into a machine learning unit to analyze and predict network performance and user behavior. At a specific time, the mobile network usage of multiple terminal users is compared. When the usage is less than a specific value, intelligent analysis is performed to generate a corresponding analysis result. The notification and display of analysis results steps: Based on the analysis results of the machine learning unit, a notification module issues user classification, abnormal notification, or obstacle maintenance instructions; the user classification, abnormal notification, or obstacle maintenance instructions and the analysis results are displayed in the user interface; wherein the data acquisition unit collects multiple historical base station trajectory data of these terminal users from the telecommunications operator's management website through the crawler program, stores and manages these historical base station trajectory data and the real-time operation data related to the base station trajectory; the AI ​​machine analysis module extracts these historical base station trajectory data and performs calculations with these real-time operation data through a graph neural network, a convolutional neural network, or a recurrent neural network to analyze whether there is a location analysis result of the same regional cluster, and transmits the location analysis result to the notification module.

7. The monitoring method as described in claim 6, wherein the data preprocessing step further includes an interference signal filtering step, wherein the interference signal filtering step filters the environmental data and the network data through a low-pass filter, a high-pass filter or a band-pass filter.

8. The monitoring method as described in claim 6, wherein in the machine learning analysis step, a deep reinforcement learning algorithm is used to optimize network performance, and the model parameters are automatically adjusted according to changes in network performance and end-user behavior, thereby generating dynamic optimization suggestions or operation instructions.

9. The monitoring method as described in Request 6, wherein in the step of notifying and displaying analysis results, a multi-dimensional analysis technique is used to divide the analysis results into multiple levels and display them in the user interface, thereby helping users to make targeted maintenance and adjustments in different network environments.

Images