Abstract: Systems, methods, and devices that relate to efficient incident management using multi-dimensional correlation across different layers of the network are disclosed. In one example aspect, a method for wireless communication includes receiving reporting information about a network incident associated with multiple network nodes in a network, and identifying a network node or a connection between two network nodes as a root cause for the network incident based on a multi-dimensional correlation that correlates one or more sets of horizontal information in respective layers of an Open Systems Interconnection (OSI) model and vertical information across multiple layers of the OSI model.

Abstract: The system receives multiple alarms indicating the issue associated with the network and obtains multiple categories associated with the multiple alarms. The category indicates a component associated with the network. Based on the multiple categories, the system creates a correlation signature associated with the multiple alarms. The system obtains historical data including a historical correlation signature that is the same as the correlation signature, a cause associated with the historical correlation signature, and an indication of accuracy associated with the cause. The system determines whether the indication of accuracy satisfies a first criterion. Upon determining that the indication of accuracy satisfies the first criterion, the system makes a prediction that the cause associated with the multiple alarms indicating the issue is the same as the cause associated with the historical correlation signature.

Abstract: The system receives multiple messages from multiple components of a wireless telecommunication network. The multiple messages indicate a state associated with the multiple components and include a first message generated by a first component among the multiple components and a second message generated by a second component among the multiple components. The system filters the multiple messages received based on content associated with a first message among the multiple messages to obtain multiple filtered messages. The content indicates importance associated with the first message. The system determines whether the first message and the second message are similar. Upon determining that the first message and the second message are similar, the system automatically creates an alarm representing both the first message and the second message, thereby reducing a number of alarms generated from the multiple filtered messages.

Abstract: The system generates metrics associated with a monitoring system of a wireless telecommunication network. The monitoring system analyzes importance of multiple messages generated by multiple components of the network and, based on the analysis, generates multiple alarms. The metrics include an event coherence metric, an event significance metric, an event knowledge metric, and an event quiescence metric. The event coherence metric indicates number of the multiple alarms including sufficient information to aid in issue resolution. The event significance metric indicates number of the multiple alarms resulting in an issue ticket creation. The event knowledge metric indicates number of the multiple alarms including documentation associated with the alarm. The event quiescence metric indicates number of the multiple alarms that have at most a predetermined number of alarms generated per issue.

Abstract: A system can obtain a service request message indicating an event that affects a user device of a telecommunications network. The system can store an indication of the event at a memory device that stores historical information of events and can process the event with a machine learning (ML) model that is trained based on the historical information to simulate the effect of the event on the telecommunications network. The system generates an output for display on a computing device. The output includes a recommendation and identifies an attribute of the event and a predicted value configured to reduce a workload or increase operational efficiency. The recommendation is configured to make the event actionable. The system receives feedback that a reviewer has acted on the recommendation. In response to the feedback, the system can re-train the ML model and configure the process for handling the service request.

Abstract: A service management system can obtain a service request message indicating an event that affects a user device of a telecommunications network. The system can store an indication of the event at a memory device that stores historical information of events and can process the event with a machine learning (ML) model that is trained based on the historical information to simulate the effect of the event on the telecommunications network. The system generates an output for display on a computing device. The output includes a recommendation and identifies an attribute of the event and a predicted value configured to reduce a workload or increase operational efficiency. The recommendation is configured to make the event actionable. The system receives feedback that a reviewer has acted on the recommendation. In response to the feedback, the system can re-train the ML model and configure the process for handling the service request.

Abstract: A service management system can obtain a service request message indicating an event that affects a user device of a telecommunications network. The system can store an indication of the event at a memory device that stores historical information of events and can process the event with a machine learning (ML) model that is trained based on the historical information to simulate the effect of the event on the telecommunications network. The system generates an output for display on a computing device. The output includes a recommendation and identifies an attribute of the event and a predicted value configured to reduce a workload or increase operational efficiency. The recommendation is configured to make the event actionable. The system receives feedback that a reviewer has acted on the recommendation. In response to the feedback, the system can re-train the ML model and configure the process for handling the service request.