Functional Split-Aware Optimal BBU Placement for 5G Cloud-RAN Over WDM Access/Aggregation Network

Muhammad Ahsan, Ashfaq Ahmed, Arafat Al-Dweik, Arsalan Ahmad

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


Fifth generation (5G) cloud-radio access network (C-RAN) aims at providing better performance and support to various applications with stringent data rate and latency requirements. In C-RAN, the processing units, known as the baseband units (BBUs), are segregated from the individual remote radio heads (RRHs) and moved to a convenient location to serve more than one RRH. This migration leads to an efficient resource allocation and cost-effective solution at the expense of a huge fronthaul traffic between the RRH and BBU hotel. The required fronthaul data rate largely depends on the employed functional split options. In this article, we propose a novel optimal BBU placement with a mixed functional split scheme to combat the fronthaul latency challenge while providing the network with greater flexibility and cost reduction. We introduce an integer linear programming (ILP)-based BBU placement problem with a mixed functional split selection approach to simultaneously minimize the number of BBU hotels and fibers, thereby reducing the network cost. Furthermore, a heuristic algorithm is proposed to solve the proposed model for large network scenarios. The obtained results show that an improvement of 25% and 50% is realized with the proposed scheme over the conventional fixed split option scheme for small and large network scenarios, respectively.

Original languageBritish English
Pages (from-to)122-133
Number of pages12
JournalIEEE Systems Journal
Issue number1
StatePublished - 1 Mar 2023


  • Baseband unit (BBU) placement
  • cloud-radio access network (C-RAN)
  • fifth generation (5G)
  • mixed functional split
  • wavelength division multiplexing (WDM) network


Dive into the research topics of 'Functional Split-Aware Optimal BBU Placement for 5G Cloud-RAN Over WDM Access/Aggregation Network'. Together they form a unique fingerprint.

Cite this