Joint Network Slicing, Routing, and In-Network Computing for Energy-Efficient 6G

TL;DR

This work tackles the energy-efficient joint optimization of network slicing, routing, and in-network computing for 6G. It formulates the problem as a Mixed-Integer Linear Program (MILP) to maximize the number of accepted users $A$ while minimizing energy $B$, subject to end-to-end delay and resource isolation constraints, and proves $NP$-hardness via a reduction from the Generalized Assignment Problem $GAP$. The authors propose WF-JSRIN, a water-filling-based heuristic with polynomial-time complexity that jointly determines slice allocations $\\lambda_m$, path decisions $x_{u_m}^p$, and per-path allocations $w_{u_m}^{p,v}$ and $z_{u_m}^{p,e}$. A comparative evaluation against Random-JSRIN, Opt-IN, and Opt-C demonstrates near-optimal performance with substantially reduced runtime and energy/bandwidth consumption when in-network computing is leveraged. The results indicate practical viability for real deployments and motivate future work on mobility-aware optimization, IAB-integrated resource management, and quantum-network slicing integrations.

Abstract

To address the evolving landscape of next-generation mobile networks, characterized by an increasing number of connected users, surging traffic demands, and the continuous emergence of new services, a novel communication paradigm is essential. One promising candidate is the integration of network slicing and in-network computing, offering resource isolation, deterministic networking, enhanced resource efficiency, network expansion, and energy conservation. Although prior research has explored resource allocation within network slicing, routing, and in-network computing independently, a comprehensive investigation into their joint approach has been lacking. This paper tackles the joint problem of network slicing, path selection, and the allocation of in-network and cloud computing resources, aiming to maximize the number of accepted users while minimizing energy consumption. First, we introduce a Mixed-Integer Linear Programming (MILP) formulation of the problem and analyze its complexity, proving that the problem is NP-hard. Next, a Water Filling-based Joint Slicing, Routing, and In-Network Computing (WF-JSRIN) heuristic algorithm is proposed to solve it. Finally, a comparative analysis was conducted among WF-JSRIN, a random allocation technique, and two optimal approaches, namely Opt-IN (utilizing in-network computation) and Opt-C (solely relying on cloud node resources). The results emphasize WF-JSRIN's efficiency in delivering highly efficient near-optimal solutions with significantly reduced execution times, solidifying its suitability for practical real-world applications.

Figures (2)

• Figure 1: The conceptual overview of joint network slicing, routing, and in-network computing.
• Figure 2: The total cost and bandwidth usage vs. the number of users (A & B), and the number of accepted users vs. the link capacity (C).