Systems and Methods for Reliability Assessment of Automated Decisions in Cellular Communication Networks

Abstract

The invention relates to a system and method for evaluating the reliability of network management decisions generated by artificial intelligence, machine learning, rule-based systems or other automated decision mechanisms in cellular communication networks. The system consists of the Decision Generation Module (10), Decision and Data Analysis Module (20), Decision Reliability Analysis Module (30), Decision Implementation Module (40), Network Element (50) and Network Performance Monitoring Module (60). The automated network decisions generated by the Decision Generation Module (10) are analyzed by the Decision and Data Analysis Module (20) together with the network data used in generating the decision. Then, the Decision Reliability Analysis Module (30) performs a reliability assessment considering technical criteria such as data quality, data currency, past application success, current network status, etc. Based on the assessment result, the Decision Implementation Module (40) decides whether to implement the decision, postpone it, or submit it for operator approval. After the decision is implemented, the reliability assessment parameters are dynamically updated based on the feedback data obtained by the Network Performance Monitoring Module (60). This ensures that automatically generated network decisions are validated before implementation, reducing the likelihood of implementing erroneous decisions, increasing network service continuity, and achieving an adaptive and closed-loop decision validation mechanism that improves itself by learning from implementation results.

Description

1 TARIFF Reliability of Automated Decisions in Cellular Communication Networks System and Method for Evaluation Technical Area The invention improves the reliability of automated decision-making mechanisms in cellular communication networks. It relates to a system and method for evaluation. State of the Art Today, energy management and handover optimization are crucial in cellular communication networks. Many processes, such as radio resource management, traffic forecasting, and fault management, are powered by artificial intelligence and machine learning. This is accomplished through learning or other automated decision-making mechanisms. This 15 Decisions made by systems often directly affect network elements. is being implemented. In current practices, the reliability of the decisions made is assessed in terms of technical criteria. There is no independent verification mechanism evaluating the decision. 20 The currency of the data used in its creation, data integrity, and the similarity of the decision to past practices. the success rate it demonstrated in these situations and the potential impact of the decision on network performance They are often not considered together. As a result, automated decisions that are flawed or have low reliability are directly linked to 25 This can be implemented, but it leads to a decrease in service quality and unnecessary optimization. this can negatively impact user experience and lead to inefficient use of network resources. This can lead to its use. On the other hand, in existing systems, why is the automated decision made reliable or 30 A technical assessment output indicating that it is unsafe is also rarely provided. This Therefore, network operators need manual control in the implementation of critical decisions. They are hearing, and this reduces the effectiveness of autonomous grid management. Therefore, multiple technical checks are performed before automatically generated network decisions are implemented. enabling reliability assessment using criteria, resulting in a reliability outcome. 2 according to which the implementation of the decision can be delayed or submitted for operator approval. An integrated decision verification system is needed. Purpose of the Invention The main purpose of the invention is to automatically generate signals in cellular communication networks. Instead of directly implementing the decisions, they are technically assessed for reliability before implementation. The aim is to present a decision validation system and methodology that enables this assessment. Another objective of the invention is to make the decision generated by the automated decision-making mechanism a 10-fold increase in the value of the decision. The timeliness and completeness of the data used in its creation, and the success of similar past decisions. the rate, current network performance indicators, and the potential impact of the decision on the network, etc. The goal is to provide a system and method for evaluation using numerous technical criteria. Another purpose of the invention is to make the decision based on the results of the reliability assessment created. direct implementation, postponement, or submission to the network operator's approval The goal is to establish a system and method that provides; energy management, base station sleep mode. management, handover optimization, radio resource management, traffic optimization, and similar services. The goal is to develop a system and method that can be used in automated network management applications; 20 by evaluating data quality, decision history, and network performance together. performing reliability analysis, directly identifying erroneous or low-reliability decisions. preventing implementation, directing critical decisions to operator approval when necessary. Autonomous systems that provide, improve network service continuity and user experience. 25 A general approach that can be applied to 4G, 5G, 6G, and future automated network management infrastructures. The goal is to provide a system and method that offers a decision verification infrastructure. Figures that will help understand the invention. Figure 1 shows the general architecture of the system that is the subject of the invention. 30 Explanation of Part References 10. Decision Making Module 20. Decision and Data Analysis Module 30. Decision Reliability Analysis Module 35 40. Decision Implementation Module 3 50. Network Element 60. Network Performance Monitoring Module Detailed Description of the Invention This detailed description outlines the preferred applications of the invention only for a better understanding of the subject. This is provided to ensure understanding and to limit the scope of protection of the invention. It is not of that quality. The invention improves the reliability of automated decision-making mechanisms in cellular communication networks. It relates to a system and method for evaluation, especially for 4G, 5G, and next-generation GPUs. Artificial intelligence, machine learning, and rule-based technologies are used in 6G mobile communication networks. Network management generated by systems and other automated decision-making mechanisms. a decision that ensures its implementation is verified in terms of technical criteria It is used within the scope of the verification system. The invention relates to base station management, energy optimization, sleep mode management, and handover. optimization, traffic management, radio resource management and similar automated network It is designed to be used in optimization processes. The system is created decisions are based on data quality, data timeliness, past application results, and current network status. and by evaluating them in terms of other technical parameters, only 20 were determined to be reliable. It ensures the implementation of decisions. The invention relates to mobile communications operators, private mobile networks (Private 5G), and O-RAN-based systems. autonomous communication infrastructures and the expected widespread adoption of autonomous networks in the future. Intelligent decision verification mechanisms that can be implemented in network management systems 25 It can be used within this scope. In a preferred application of the system and method that is the subject of the invention, the Decision Production Module (10) An automated decision is made regarding the cellular communication network. The word The subject of the decision is artificial intelligence, machine learning, rule-based systems or other automated 30 Decisions can be generated using decision-making mechanisms, and the generated decision is Decision and Data Analysis. It is transmitted to module (20). Decision and Data Analysis Module (20), together with the decision created, in the creation of the decision It analyzes the network data used. This includes data from base stations and 35 other sources. traffic load from network elements, number of active users, resource utilization rate, 4 Verification is done by evaluating alarm information, performance indicators, and similar technical data. Analysis outputs to be used in the process are being generated. The obtained analysis results are transferred to the Decision Reliability Analysis Module (30). Decision Reliability Analysis Module (30) assesses the reliability of the decision based on data quality, timeliness of the data, and past 5 the success rate of similar decisions, the current network condition, and the decision's network performance. It evaluates this by taking into account technical criteria such as its potential effects. The feasibility of implementing the decision is determined as a result of the evaluation. The resulting reliability assessment is sent to the Decision Implementation Module (40). 10 Decision Implementation Module (40), depending on the reliability assessment result, the decision It ensures that the Network Element (50) is applied automatically. Reliability If the level remains below the determined threshold value, the decision may be postponed. or it can be submitted for operator approval. Following the implementation of the decision, the Network Performance Monitoring Module (60) monitors the network Performance indicators obtained from employees are continuously monitored. Performance data is transferred back to the Decision and Data Analysis Module (20) and then It generates feedback data to be used in validation processes. In a preferred application, the data obtained by the Network Performance Monitoring Module (60) Decision Reliability Analysis Module using performance data and application results (30) The evaluation parameters are updated dynamically. Thus, the system updates the previous by considering the success rates of the decisions, we are continuously improving the verification process. an adaptive decision validation mechanism capable of adapting to changing network conditions 25 It constitutes. Thanks to this structure, the reliability of automatically generated network decisions is based on technical criteria. This is evaluated using the network, reducing the likelihood of implementing erroneous decisions. Service continuity is increased and autonomous network management systems are made more reliable. 30 The system also provides feedback from the application results. Adaptive and closed-loop systems that continuously update validation parameters using data. It offers a (closed-loop) decision verification mechanism.

Claims

REQUESTS 1. Reliability of automated decision-making mechanisms in cellular communication networks. It is a system for evaluation, and its characteristic feature is; 5 using artificial intelligence, machine learning or other automated decision-making mechanisms Decision generation module (10) which forms the decisions regarding network management. analyze the decision that was made and the network data used in making that decision a decision and data analysis module that prepares the necessary information for the verification process. (20), The reliability of the decision depends on data quality, the success rate of similar past decisions, and the current 10 reliability by evaluating network status and technical performance indicators. Decision reliability analysis module (30) which forms the result, Based on the reliability assessment result, the implementation, postponement or Decision implementation module (40) which enables submission to operator approval. Network 15, which is the base station or related network component where the verified decision is implemented. element (50), By monitoring the KPIs, alarms, and performance results generated after the decision is implemented, feedback can be provided. transmitting to the decision and data analysis module (20) for the purpose of feeding and the system's closed loop Network performance monitoring module (60) that enables (closed-loop) operation It includes. 20

Images