The Energy Cost of Artificial Intelligence Lifecycle in Communication Networks

TL;DR

The paper introduces eCAL, a lifecycle‑level energy metric for AI in communication networks, addressing the gap left by metrics that isolate transmission, data center, or model energy. It models the AI lifecycle as a set of data‑manipulating components (Data Collection, Preprocessing, Training, Evaluation, Inference) aligned with OSI and MLOps, and derives energy expressions for each component, culminating in a unified end‑to‑end energy per bit metric. A case study shows that increasing inference frequency amortizes the upfront development energy, yielding a 2.73× improvement in energy per bit from 100 to 1000 inferences for a simple MLP, and the authors provide an open‑source, modular calculator to enable broader adoption. The framework supports detailed breakdowns across wireless and wired segments, enabling designers to optimize energy efficiency in AI‑enabled networks and informing future standards like O‑RAN. Overall, eCAL offers a practical, lifecycle‑aware tool for assessing and reducing the environmental footprint of AI in telecommunications.

Abstract

Artificial Intelligence (AI) is being incorporated in several optimization, scheduling, orchestration as well as in native communication network functions. This paradigm shift results in increased energy consumption, however, quantifying the end-to-end energy consumption of adding intelligence to communication systems remains an open challenge since conventional energy consumption metrics focus on either communication, computation infrastructure, or model development. To address this, we propose a new metric, the Energy Cost of AI Lifecycle (eCAL) of an AI model in a system. eCAL captures the energy consumption throughout the development, deployment and utilization of an AI-model providing intelligence in a communication network by (i) analyzing the complexity of data collection and manipulation in individual components and (ii) deriving overall and per-bit energy consumption. We show that as a trained AI model is used more frequently for inference, its energy cost per inference decreases, since the fixed training energy is amortized over a growing number of inferences. For a simple case study we show that eCAL for 100 inferences is 2.73 times higher than for 1000 inferences. Additionally, we have developed a modular and extendable open-source simulation tool to enable researchers, practitioners, and engineers to calculate the end-to-end energy cost with various configurations and across various systems, ensuring adaptability to diverse use cases.

Figures (13)

• Figure 1: Data manipulation components involved in the lifecycle of an AI model.
• Figure 2: Total transmitting bits ($B_{\mathrm{T}}$) versus retransmission rate with $\gamma_{l}=0.5$, $OH_{\mathrm{DP},l}=10\%$ and $OH_{\mathrm{CP},l}=5\%$.
• Figure 3: Energy consumption of data collection ($E_{\mathrm{DC}}$) (log scale) with its components and configurations, including transmission energy ($E_{\mathrm{T}}$), receiving energy ($E_{\mathrm{R}}$), and computational energy ($E_{\mathrm{C}}$) with $N_{\mathrm{S}}=256$, $RR_{\mathrm{DP},l}=RR_{\mathrm{CP},l}=1$, $OH_{\mathrm{DP},l}=10\%$, $OH_{\mathrm{CP},l}=5\%$, and $\gamma_{l}=0.5$.
• Figure 4: Proposed $E_{\mathrm{DC}}$ comparison with real-world measurement data.
• Figure 5: Energy consumption ($E_{\mathrm{pre}}$) of different sample size ($I_{\mathrm{S}}$) across different preprocessing techniques.