Components

Experimentation xApps with CF mMIMO Emulator

The purpose of the Near-RT RIC with an emulated cell-free massive MIMO (CF mMIMO) RAN is to test innovative radio resource allocation strategies in an environment which is very close to the real deployment of CF mMIMO system. The provided solution supports: a wide range of configuration parameters, random traffic simulation, user equipment (UE) mobility, a wide range of reported key performance measurements (KPMs). The Radio resource allocation strategy controlled by the xApps, and short control time of 0.1 s.

Key functionalities include the test of innovative radio resource allocation strategies in an environment close to the real CF mMIMO system. The xApps can change the radio resource allocation strategies per report period and showcase the performance of the allocation strategies. The solution combines the radio resource allocation consists of 3 Steps:

Key Features

01
Scheduling
Determines the UEs, which will get AP-PRB resources in the current TTI. Current implementation of the scheduling algorithm has in-built the following radio resource allocation algorithms which can be triggered thorough the xApp onboard in the Near-RT RIC. In the proposed solutions two types of scheduling feature is inbuild to showcase the performance of CF mMIMO networks. Two types of scheduling mechanism are added in scheduling modules: Immediate - all UEs with data in the traffic buffer are considered for scheduling; Delayed - the UEs are not scheduled immediately but the scheduler waits until there is data in the buffer for more UEs to schedule them simultaneously but not later than configurable packet delay budget (PDB).
02
Clustering
Determines the serving APs per served UE in the current TTI o All APs - all APs serve all UEs, which is available for ZF precoding. Heuristic SE - for MRT precoding to optimise the sum spectral efficiency (SE). For each AP-PRB one served UE is selected, xApp defined - the xApp provides a clustering vector with serving APs for each UE. The clustering vector can be changed every reporting period.
03
Precoding
Determines the precoding used when transmitted from the APs.

Technologies Involved

The key technologies involved in the Cognitive Compute Continuum are related with the π-edge Platform. These are:

  • PRB allocation for CF mMIMO.

  • UE mobility-aware dynamic clustering.

  • Adaptive precoding determination.

Collaboration Process

There are two processes for collaboration:

  • Through the official 6G-BRICKS portal.

  • Reaching one of the contact points to create personal access to the testbed.

Contact Points

  • Md Arifur Rahman
  • Lead Researcher at IS-Wireless (Warsaw, Poland)
  • a.rahman@is-wireless.com

Documentation/Resources

Accessibility by WAH
Scroll to Top