US-12625797-B2 - Multiple telecommunication endpoints system and testing method thereof based on AI decision

US12625797B2US 12625797 B2US12625797 B2US 12625797B2US-12625797-B2

Abstract

An AI decision based multiple telecommunication endpoints system is provided, including a telecommunication endpoint device and an artificial intelligence decision controller. The telecommunication endpoint device performs a communication test with the radio frequency communication device under test. The artificial intelligence decision controller is electrically connected to the telecommunication terminal device to control the communication test, and performs an efficiency analysis on the result of the communication test, and generates a decision instruction according to the result of the efficiency analysis. The artificial intelligence decision controller can centrally control, replace and analyze multiple artificial intelligence modules and generate a large number of telecommunication endpoints device signal sending and receiving behaviors, providing a telecommunication endpoints system with self-adaptive adjustment process, parameters, result analysis and reasonable cost as a test for the development phase, deployment phase, or maintenance phase of the telecommunications equipment product development process.

Inventors

En-Cheng Liou
Ta-Sung Lee
Kai-Ten Feng
Yu-Chien Lin
Chia-hung Lin

Assignees

NATIONAL YANG MING CHIAO TUNG UNIVERSITY

Dates

Publication Date: 20260512
Application Date: 20210804
Priority Date: 20200903

Claims (16)

1 . An artificial intelligence (AI) decision controller, comprising: a telecommunication endpoint simulator, being configured to set a test environment parameter of a telecommunication endpoint device, the telecommunication endpoint device generates a test signal corresponding to the test environment parameter; a plurality of AI modules, comprising a configuration of a recommendation AI module and a backup AI module, the configuration corresponding to the test environment parameter, wherein each of the AI modules generates a prediction signal corresponding to the test environment parameter respectively; a data analysis unit, being configured to perform an efficiency analysis on a result of a communication test by comparing a feedback signal and each of the prediction signal, the communication test being performed by the telecommunication endpoint device and at least one radio frequency (RF) communication device under test, the prediction signal corresponding to the configuration, the feedback signal being transmitted by the at least one RF communication device under test; a decision unit, being configured to generate a decision instruction according to a difference value of each of the prediction signal and the feedback signal so as to adjust the configuration; and a main controller, being electrically connected to the telecommunication endpoint simulator, the AI modules, the data analysis unit, and the decision unit.
2 . The AI decision controller of claim 1 , wherein at least one of the telecommunication endpoint simulator, the AI modules, data analysis unit, decision unit, and main controller is implemented by a processor.
3 . The AI decision controller of claim 1 , wherein the telecommunication endpoint device is one of a user equipment, an internet of things (IoT) device, a software radio device, and a system platform for transmitting a RF signal.
4 . The AI decision controller of claim 3 , wherein the telecommunication endpoint device comprises a control device, a transmission interface, a telecommunication signal module, a channel emulator, and an antenna.
5 . The AI decision controller of claim 1 , wherein the test environment parameter of the telecommunication endpoint device comprises at least a time synchronization signal and a location signal.
6 . The AI decision controller of claim 1 , wherein the decision instruction, generated by the decision unit, comprises a weight value for adjusting the AI modules.
7 . The AI decision controller of claim 6 , wherein when the difference value of the recommendation AI module is less than a threshold value, the data analysis unit maintains the configuration.
8 . The AI decision controller of claim 6 , wherein when the difference value of the recommendation AI module is greater than a threshold value and the difference value of the backup AI module is less than the threshold value, the data analysis unit increases the weight value of the backup AI module through the decision instruction.
9 . The AI decision controller of claim 6 , wherein when the difference value of the recommendation AI module is greater than a threshold value and less than the difference value of the backup AI module, the main controller performs a machine learning training or a deep learning training on the recommendation AI module by transmitting the feedback signal.
10 . An artificial intelligence (AI) decision based multiple telecommunication endpoints system, comprising: a telecommunication endpoint device being configured to perform a communication test with at least one radio frequency (RF) communication device under test; and an AI decision controller being electrically connected to the telecommunication endpoint device to control the communication test, performing an efficiency analysis on a result of the communication test, and generating a decision instruction according to a result of the efficiency analysis, the AI decision controller comprising: a telecommunication endpoint simulator, configured to set a test environment parameter of a telecommunication endpoint device, the telecommunication endpoint device generating a test signal corresponding to the test environment parameter; a plurality of AI modules which comprising a configuration of a recommendation AI module and a backup AI module, the configuration corresponding to the test environment parameter, wherein each of the AI modules generates a prediction signal corresponding to the test environment parameter respectively; a data analysis unit, being configured to perform the efficiency analysis by comparing a feedback signal and each of the prediction signal; a decision unit, being configured to generate the decision instruction according to a difference value of each of the prediction signal and the feedback signal so as to adjust the configuration; and a main controller, being electrically connected to the telecommunication endpoint simulator, the AI modules, the data analysis unit, and the decision unit.
11 . The AI decision based multiple telecommunication endpoints system of claim 10 , wherein the at least one of the telecommunication endpoint simulator, the AI modules, data analysis unit, decision unit, and main controller is implemented by a processor.
12 . The AI decision based multiple telecommunication endpoints system of claim 10 , wherein the test environment parameter of the telecommunication endpoint device comprises at least a time synchronization signal and a location signal.
13 . The AI decision based multiple telecommunication endpoints system of claim 10 , wherein the decision instruction, generated by the decision unit, comprises a weight value for adjusting the AI modules.
14 . The AI decision based multiple telecommunication endpoints system of claim 13 , wherein when the difference value of the recommendation AI module is less than a threshold value, the data analysis unit maintains the configuration.
15 . The AI decision based multiple telecommunication endpoints system of claim 13 , wherein when the difference value of the recommendation AI module is greater than a threshold value and the difference value of the backup AI module is less than the threshold value, the data analysis unit increases the weight value of the backup AI module through the decision instruction.
16 . The AI decision based multiple telecommunication endpoints system of claim 13 , wherein when the difference value of the recommendation AI module is greater than a threshold value and less than the difference value of the backup AI module, the main controller performs a machine learning training or a deep learning training on the recommendation AI module by transmitting the feedback signal.

Description

CROSS-REFERENCE TO RELATED APPLICATION This application claims priority to Taiwan Application Serial Number 109130297, filed Sep. 3, 2020, which is herein incorporated by reference in its entirety. BACKGROUND Field of Invention The present invention relates to a multiple telecommunication system and a testing method thereof. More particularly, the multiple telecommunication system and the testing method of present invention are based on an artificial intelligence decision. Description of Related Art The cost of the test equipment for testing current mobile network remains high, and the conventional test equipment is difficult to integrate different test since it mainly focuses on a single function. For example, channel emulator for performing channel simulation on a single mobile terminal device or wireless signal does not discuss simulation of multiple mobile terminal systems, or using components such as an anechoic chamber to generate and synthesize channel analog signals does not discuss the integration of multiple terminal device systems with AI modules. Moreover, when testing a wireless device or a single base station, the common system under test of the overall base station and core network is ignored. Further, signals which are generated by a test system are often generated by a single signal source and then transmitted to multiple radio frequency elements. This does not discuss centralized definition of control signals, and then generate test signal as a main mode. According to above descriptions, the problems which exist in conventional technologies can be summarized into three categories. The first problem is that during the development and verification of mobile communication systems, the test equipment can only deploy a single device since traditional test systems and tools are designed and developed mainly for hardware. However, there is a huge difference in the number of users of the telecommunication system in actual commercial use, so the cost of the research and development stage remains high, or it is unable to provide a large number of research and development verification at the research and development stage, and usually rely on manpower to control and manage the test methods. The second problem is that at the equipment development or introduction stage, a lot of testing tools and testing personnel are required in laboratory testing and field testing, so errors will inevitably be caused by humans in the operation, management problems will happen, and results that are sufficient verifiable or quick adjustment may not always be obtained. The third problem is that when the mobile communication system develops telecommunications public or private network application services, the functional problems are usually eliminated, and the stability of performance, the experience of application quality, and the impact on existing network services still need to evaluate. It is difficult for the telecommunications network service industry to use existing testing tools and services to directly simulate and confirm the effectiveness of the development of relevant new-type application services and the feasibility of introduction and evaluation. SUMMARY The present disclosure provides multiple telecommunication endpoint system based on artificial intelligence (AI) decision, providing a set of highly intelligent multiple telecommunication endpoint system solutions for the mobile communication industry supply chain. The multiple telecommunication endpoint system solutions can simulate behavior of connecting with a large number of user signals by using multiple AI modules which can be replaced and controlled centralized in the same environment. Specifically, the present disclosure provides an artificial intelligence (AI) decision controller which comprises a telecommunication endpoint simulator, a plurality of AI modules, a data analysis unit, a decision unit, and a main controller. The telecommunication endpoint simulator is configured to set a test environment parameter of a telecommunication endpoint device and generates a test signal corresponding to the test environment parameter. The AI modules comprise a configuration of a recommendation AI module and a backup AI module, the configuration corresponding to the test environment parameter. Each of the AI modules generates a prediction signal corresponding to the test environment parameter respectively. The data analysis unit, being configured to perform an efficiency analysis on a result of a communication test by comparing a feedback signal and each of the prediction signal. The communication test is performed by the telecommunication endpoint device and at least one radio frequency (RF) communication device under test. The prediction signal corresponding to the configuration, and the feedback signal is transmitted by the at least one RF communication device under test. The decision unit is configured to generate a decision instruction according to a di