RIC APP ML

RIC APP ML Overview

The O-RAN SC Machine Learning (ML) Common Services provides ML tools, adapters to integrate with a radio access network (RAN) controller.

Using Acumos ML models in the RIC:

  • Goal is to support ML models in non-real time and near-real time RIC usecases.
    ** quickly import an Acumos model into RIC and adapt it into as an xApp (near-real time). ** deploy Acumos models as is into non-real time (mostly on ONAP side).
  • Priority is to get something working with minimal changes possible on ML models
    ** focus on performance in the later releases, since many ML models take some time to execute anyway.
  • Build a standard xApp/Acumos microservice adapter
    ** deployed along with the Acumos ML model in one Kubernetes pod.
  • Adapter speaks RMR protocol to RIC
    ** communicates with the Acumos ML model in the standard http / GRPC manner.
  • Configuration needed for each deployment
    ** to tell adapter how to speak with Acumos ML model. ** can be auto generated using ML model protobuf definition.
  • Consider writing custom RMR model runner
    ** for performance in near-real time RIC xApps in the following releases.

Release Notes

This document provides the release notes for the Amber Release of the Acumos xAPP adapter.

Version history

Date Ver. Author Comment
2019-11-14 0.0.1 Guy Jacobson First draft

Summary

The Amber release of the Acumos xAPP adapter contains the code needed to use an existing Acumos microservice as an O-RAN xAPP, by providing “glue” that listens and speaks RMR protocol and translates these into calls to the Acumos microservice, which is co-deployed in the same pod as the adapter.

Release Data

Project RAN Intelligent Controller
Repo/commit-ID ric-app/ml
Release designation Amber
Release date 2019-11-14
Purpose of the delivery open-source adapter between Acumos and xAPPs.

Components

  • AcumosXappAdapter/ contains the source code and other items of interest. Under that directory :
    • rmracumosadapter.py is source code for the adapter itself.
    • iris_sklearn.py is the source code for a generic Acumos model (iris classification).
    • config.json is a sample configuration file, needed to connect the Acumos model with the xAPP adapter during deployment.
    • Dokcerfile is the Dockerfile that builds the xAPP adapter microservice.
    • testdata.csv contains sample input data to test the iris_sklearn.py classifier

Limitations

  • This is a first release and needs some fixes to the Dockerfile function correctly, due to known problems with the build process to incorporate the required nng libraries.