Refer Intelligent Slicing flow for complete closed loop sequence. Existing ML MS (not officially exposed to ONAP) requires certain enhancements to satisfy the intent guarantee.
Overview
Functionality of ML block is divided into three parts.
- Model Training (Offline)
- Validation
- Prediction
Step1: ML Offline Training
This step requires a huge amount of historical data for the analysis. It requires both the input (PM metrics) and output data (intent).
PM file format
To fetch the Slice Profile (intent) from config DB, below API can be used.
GET: http://localhost:8080/api/sdnc-config-db/v4/profile-config/{nSSAI}
The response will be in the form of:
{
"dLThptPerSlice": 1,
"uLThptPerSlice": 2,
"maxNumberOfConns":300
}
Once the CPS integration for RAN slice allocation is completed, Config DB API needs to be replaced with the CPS API.
CPS API:
To be updated
Recommendation for each cell can be maximum of 5120 and the aggregation of maximumNoOfConnections recommended for a slice (sNSSAI) can be upto 110% (10 % buffer) of the intent.
Step2: Validation
Trained ML model can be validated against the dynamic data in the E2E Slicing set up.
- Configure four E2E Slices (shared or non-shared)
- Trigger PM data generation for those slices using the below RAN Simulator API
curl -X POST -H "Content-Type: text/plain" http://localhost:8081/ransim/api/GenerateIntelligentSlicingPmData -i
3. Run the ML model against the generated PM data and check whether it works as expected
4. When it is not required, stop the pm data generation as it produces huge amount of data that consumes more space.
curl -X POST -H "Content-Type: text/plain" http://localhost:8081/ransim/api/stopIntelligentSlicingPmData -i
Step3: Prediction
Trained ML model is now ready to predict and recommend suggestions for the cell level configurations (maxNumberOfConns) for a slice.
A new micro service will be introduced to do the data processing and prediction for the dynamic data.
Suggested Configuration from the ML MS should be in the form of below format