A Joint Method for Network Slicing, Routing, an‎d In-Network Computing in 6G

2/21/2027 12:00:00 AM

The growing heterogeneity an‎d service diversity envisioned for future 6G networks makes it necessary to develop intelligent mechanisms for orchestration of communication, computation, an‎d network resources that can jointly manage them. In this context, network slicing, network routing an‎d in-network computing (INC) have become major enabling technologies for meeting the diversity of quality of service requirements in a highly efficient resource utilization manner. However, most of the existing joint optimization frameworks rely on simplified energy models which only consider the energy consumption depending on traffic an‎d implicitly assume that all the network components remain continuously active. Such assumptions limit how accurate energy modeling is an‎d ignore significant energy savings that can be realized through the activation of adaptive infrastructure. This thesis proposes an energy-aware orchestration framework which extends joint slicing, routing, an‎d in-network computing optimization by explicitly considering activation decisions for both network nodes an‎d communication links. The proposed formulation adds more binary connections an‎d constraints that represent the operational state of the network elements to establish consistency between routing decisions an‎d activation of the physical infrastructure. This enables unused links an‎d nodes to be selec‎tively deactivated so that traffic can be consolidated over a smaller set of active resources an‎d static energy consumption can be minimized. The problem is formulated as a mixed-integer optimization model that simultaneously decides routing paths, computation placement, resource allocation at the slice level, an‎d activation states. Extensive simulations made under different traffic loads show that the proposed approach can significantly reduce the total amount of energy consumption on the network while maintaining a stable service provisioning performance. Although there may be a minor decrease in user acceptance under high load conditions, the framework provides a good balance between energy efficiency an‎d servicing capacity. Overall, the results confirm that explicit link activation modelling is a crucial enhancement for realistic an‎d energy-efficient orchestration in large-scale 6G environments. The proposed approach develops a workable basis for intelligent networks of the next generation that require dynamic, scalable, an‎d sustainable resource management.

شبكه 6G , برش شبكه , مسير يابي , محاسبات درون شبكه‌اي , هماهنگ‌سازي آگاه از انرژي

6G Networks , Network Slicing , Routing, In-Network Computing , Energy-Aware Orchestration , Routing

Ahmed Alsahlanee

Dr. Zeinab Movahedi