Macro--Micro Decision-Making in 6G Networks: An Agent-Based Framework for the Resource-Fungibility Landscape Resource-Fungibility Landscape

Sayanti Ghosh, Indrakshi Dey, Nicola Marchetti

A defining feature of 6G networks is that performance depends not only on the quantity of available resources (e.g., spectrum, antennas, cache memory, compute, and fronthaul bandwidth) but also on their \emph{fungibility}, i.e., the ability of one resource to substitute for another under changing conditions. We argue that the fungibility landscape of a distributed 6G system is governed by two coupled decision scales: \emph{micro} decisions made locally by agents and \emph{macro} outcomes that emerge at the network level. Existing distributed-optimization approaches largely conflate these scales. To address this gap, we develop an agent-based-modeling (ABM) framework that separates macro and micro decisions through three operator-controllable macro choices, three micro hyperparameters, and three structural metrics. We establish six key results: (i) a two-timescale decomposition theorem, (ii) a structural-metric basis theorem, (iii) a macro--micro design rule with closed-form factorization of the emergent breakdown threshold, (iv) a fungibility--resilience monotonicity proposition, (v) a connectivity--substitutability duality theorem, and (vi) a multi-application generalization proposition. Numerical results visualize the macro fungibility landscape and the micro decision-sensitivity region for a representative 6G deployment.