.png){kind=icon}
Return
Is the flexible functional split worthwhile?
.png){kind=icon}
PRE
AI
DOI：10.1016/j.simpat.2026.103358.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The virtualized Radio Access Network (vRAN) enables the disaggregation and flexible deployment of 5G RAN protocol layer functions across distributed geographical locations through different Functional Splits (FSs). The dynamic selection of the FS, known as Flexible Functional Split (FFS), is essential for meeting user requirements and reducing operational costs, while ensuring mobile network stability is preserved. In this paper, we propose a use case based on the varying traffic demand conditions between day and night setups and introduce a new FFS selection algorithm aimed at reducing energy consumption while meeting user requirements. Using the open-source full-stack ns-3 5G-LENA system-level simulator, we show that the proposed FFS algorithm adapts the selected FS to the MH capacity available to the macro-cell DU as this capacity varies across 4, 2 and 1 Gbps, sustaining the offered rate of 18.24 Mbps per user, except for a short transient period after each reconfiguration. Fixed FS 7.1 fails to sustain the offered rate (delivering only 11.42 Mbps), fixed FS 7.2 sustains it only at a mean latency of 184 ms, and FS 7.3 sustains the load at a higher night-time cost. FFS reduces the total night-time cost by 30.8%, with four FS reconfigurations per day/night cycle in 90% of the 20 evaluated seeds and five in the remaining 10%. We further compare the proposed algorithm against an adaptation of the split selection criterion from the literature, which tracks the capacity variations but does not reach the most centralized split during the night, achieving a 14.0% reduction in night-time cost.
Keywords:
Flexible Functional Split
Fronthaul/midhaul network
Virtualized 5G RAN
Journal
IF:
4.6
Papers:
2.6K
Citations:
4.8K
Organization
View more organizations
Cited Papers
No cited papers available