Dynamic programming-based reconfiguration algorithm for optimized 5GC refactoring

•We utilize the previous design of the quantitative indicators to evaluate the queuing delay and resource allocation cost. As a result, we can evaluate the performance in a more time-efficient manner.•We transform and reconstruct the previous optimization problem formulation with the two quantitativ...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	Future generation computer systems Ročník 176; s. 108214
Hlavní autori:	Chiang, Wei-Kuo, Huang, Yun-Fan, Chen, Feng-Ming
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	Elsevier B.V 01.03.2026
Predmet:	5G core network architecture Dynamic programming algorithm Network function virtualization Quantitative analysis Queuing delay Refactoring Queuing delay Dynamic programming algorithm Network function virtualization Refactoring 5G core network architecture Quantitative analysis
ISSN:	0167-739X
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Popis
Shrnutí:	•We utilize the previous design of the quantitative indicators to evaluate the queuing delay and resource allocation cost. As a result, we can evaluate the performance in a more time-efficient manner.•We transform and reconstruct the previous optimization problem formulation with the two quantitative indicators, MER and MUD, as the objective function instead of the queuing delay and resource allocation cost.•This article proposes a dynamic programming-based heuristic algorithm to derive a near-optimal 5G core network configuration (DPR-5GC) to shorten the execution time of the real-time dynamic reconfiguration.•The execution time of the heuristic algorithm for the DPR-5GC is about 0.026s, while the execution time of the GUROBI Optimizer for the optimization problem is about 5896 min; that is, the DP-based reconfiguration algorithm performs approximately thirteen million times faster than the GUROBI Optimizer. [Display omitted] The 5G core network should be reconfigured effectively according to time-varying traffic demand and resource availability. This article proposes a dynamic programming-based reconfiguration algorithm to optimize the reconfiguration of the 5G core network (5GC). With two quantitative indicators for the quantitative analysis, message exchange reduction (MER) and merging utilization degradation (MUD), we can evaluate the impacts of merging certain network functions. We transform and reconstruct the previous optimization problem formulation with the two quantitative indicators, MER and MUD, as the objective function instead of the queuing delay and resource allocation cost. Moreover, we proved that the reconstructed refactoring problem is an NP-hard by reducing the Knapsack problem to it. Therefore, we employed a dynamic programming-based algorithm for the Knapsack problem to design the reconfiguration algorithm, performing quantitative analysis of MER and MUD to derive the DP-based Reconfigured 5GC (DPR-5GC) architecture. The performance results show that the DPR-5GC outperforms the original 5GC and is comparable to the GUR-5GC. The dynamic programming approach can overcome the time-consuming calculation, obtain a near-optimal solution, and is feasible for dynamic scaling design.
ISSN:	0167-739X
DOI:	10.1016/j.future.2025.108214