Dimensioning KPI - Concept, Validation, Dashboard & Artificial Intelligence

Key Performance Indicator. "5G" with different types mentioned below.

Here as per 3GPP Standard KPIs are broadly classified into 5 main categories

ACCESSBILITY

INTEGRITY

UTILIZATION

RETAINABILITY

ENERGY EFFICIENCY

These are key Pointers which are broadly described as per Radio Access Technologies, Additionally different OEMs categorize the Product embarkation based on additional value Added KPIs which here I have marked them as BETA KPIs which certainly be Critical for many operators.

AVAILIBILITY

E2E QOS/QOE

Subscription/Notification for VNF.

KPI TEMPLATE        

As per 3GPP 28.554 specification there is predefined Template which is to be followed.

• Description[Mandatory]
• Formulae [As per 3GPP 28.553 Parameters with required Formulations are mentioned]
• KPI Object [Mandatory]- Network Slice Subnet, Sub Network, Network Slice, NRCellDU, NRCellDU

Example, Registration Success Rate-RSR

KPI CATEOGARIZATION        

Whereas with any specific tools Online/Offline we can use for extracting Operational data trends, or we can authorize with Long Running Stability Lab Setup for Validation of the key parameters are mentioned here.

Major KPI Trends        

Whereas many operators under the BETA KPI like to visualize the NSA stay time 5G-NR involvement and which has detailed formulation to be established by OEMs and may be subjectively Proprietary.

NSA Network Duration time KPI [Conditional]- Operator Specific         

As per sample example above from UE logs here UE was mostly involved in NSA data download activity and hence trend can be seen accordingly. The above given can be conflicting with Data KPI hence correct relation elaboration is required, formulation need to be well driven for such KPIs.

KPI Dashboard

For an optimization engineer starting to test/verify the KPI a lucrative dashboard looks, cover all key aspect from multiple factors which are mentioned above, this apply consolidated learning for patterns on field while Drive Test, Pre-validation, Post Validation and Network Monitoring* [Relies on Pre-prepared Dashboards].

Below given is the example which is prepared on sample data, no OEM or Operator data collected here with below plots.

Once data is available for analytics we have many methods one of the popular method is Power BI, simply parse the data and it will help displaying trends as per expectations.

As we talk about Data Analytics So here we go to world of Artificial Intelligence Let's Demystify AI Design and 3GPP Model to introduced for O&M and Management Data Analytics.

Artificial Intelligence        

As per 3gpp standards AI/ML is playing very vital role in orchestrating every components of CORE, RAN with Data Analytics.

Here Broadly the key aspect is MDA "Management Data Analytics"

The MDA provides a capability of processing and analyzing data related to network and service events and status including e.g. performance measurements, KPIs, Trace/MDT/RLF/RCEF reports, QoE reports, alarms, configuration data, network analytics data, and service experience data from AFs, etc. to provide analytics output, i.e. statistics or predictions,, root cause analysis issues, and may also include recommendations to enable necessary actions for network and service operations.

AI/ML MDA Fundamental        

Observation: The observation of the managed networks and services. It involves monitoring and collection of events, status and performance of the managed networks and services, and providing the observed/collected data.

Analytics: The data analytics for the managed networks and services. MDA plays the role of Analytics in the management loop. It prepares, processes and analyses the observed/collected data or time series of the observed/collected data related to the managed networks and services. MDA reports may contain root cause analysis of ongoing issues, predictions of potential issues and corresponding relevant causes and recommended actions for preventions, and/or prediction of network and/or service demands.

Decision: The decision making for the management actions for the managed networks and services. The management actions are decided based on the analytics reports (provided by MDA) and other management data (e.g. historical decisions made previously) if necessary. The decision may be made by the consumer of MDAS (in the closed management control loop), or by a human operator (in the case of open management loop). The decision may include e.g.?what actions to take, and when to take the actions.

Execution: The execution of the management actions according to the decisions. During the execution step, the actions are carried out to the managed networks and services, and the reports (e.g. notifications, logs) of the executed actions are provided.

As per standard Network Data Analytics Function (NWDAF) is been incorporated with CORE Design on 3GPP TS 24.501,

KEY ROLE FOR [NWDAF]        

• Data collection based on subscription to events provided by AMF, SMF, UPF, PCF, UDM, NSACF, AF (directly or via NEF) and OAM;
• Analytics and Data collection using the DCCF (Data Collection Coordination Function);
• Retrieval of information from data repositories (e.g. UDR via UDM for subscriber-related information or via NEF(PFDF) for PFD information).
• Storage and retrieval of information from ADRF (Analytics Data Repository Function).

Conclusively design with NWDAF, MDA can work in robust framework with minimal Manual interactions required and pro-effective data analysis.

Question asked to ChatGpt - Can AI Replace in Future the KPI Monitoring Jobs?...         

A lot is there to innovate why not to do in #Style.

#LearnInStyle.