Anomaly Detection Overview

Anomaly Detection (AD) is an Xapp in the Traffic Steering O-RAN use case, which perfrom the following tasks:

  1. Data will be inserted into influxDB when xApp starts. This will be removed in Future when data will be coming via KPIMON to influxDB.

  2. AD, which iterates per 10 mili-second, fetches UE information from databse and send prediction to Traffic Steering

  3. Traffic Steering send acknowldgement back to AD.

Expected Input

The AD Xapp expects input in following structure:

{ ‘du-id’ : 1003, ‘nrCellIdentity’ : “c3/B13”, ‘prb_usage’ : 23.0, ‘rsrp’ : 84.0, ‘rsrq’ : 65.0, ‘rssinr’:65.0, ‘targetTput’ : 0.1, ‘throughput’ : , ‘ue-id’ : “Waiting passenger 1”, ‘x’ : -556, ‘y’ : -1160, ‘measTimeStampRf’ : “2021-05-12T07:43:51.652” }

Expected Output

The AD Xapp should Detect Anomulous UE’s and send those UE’s information as a JSON message via RMR with the following structure:

{ ‘ue-id’ : “Waiting passenger 1”, ‘measTimeStampRf’ : “2021-05-12T07:43:51.652”, ‘du-id’ : 1003, ‘Degradation’: “RSRP RSSINR” }