Green Bin Packing

2025-10-30

AI Summary

This paper addresses energy consumption and sustainability challenges arising from high-utilization servers in cloud computing. We propose the “green bin packing” online optimization model: extending classical bin packing by introducing a utilization threshold $G$ and a linear overcapacity cost $eta$, where exceeding $G$ incurs penalty proportional to the excess load. The objective is to minimize the sum of the number of open bins and total overcapacity cost. To our knowledge, this is the first formalization of such a linear overcapacity mechanism. We systematically analyze how the competitive ratios of classical algorithms—including FirstFit and Harmonic—exhibit phase-transition behavior as a function of the product $eta G$. Theoretically, we prove that when $eta G leq 1$, standard algorithms retain their original optimal competitive ratios; when $eta G > 1$, we design novel algorithm variants that strictly improve worst-case competitive ratios. Empirical evaluation confirms significant gains in resource utilization and cost-efficiency.

Technology Category

• Search and Optimization: Algorithm Configuration
• Machine Learning: Efficient ML / Green AI
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Economics, Online Markets and Human Computation: Sustainability of Web economics
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments

Abstract

The online bin packing problem and its variants are regularly used to model server allocation problems. Modern concerns surrounding sustainability and overcommitment in cloud computing motivate bin packing models that capture costs associated with highly utilized servers. In this work, we introduce the green bin packing problem, an online variant with a linear cost $β$ for filling above a fixed level $G$. For a given instance, the goal is to minimize the sum of the number of opened bins and the linear cost. We show that when $βG le 1$, classical online bin packing algorithms such as FirstFit or Harmonic perform well, and can achieve competitive ratios lower than in the classic setting. However, when $βG > 1$, new algorithmic solutions can improve both worst-case and typical performance. We introduce variants of classic online bin packing algorithms and establish theoretical bounds, as well as test their empirical performance.

Problem

• Modeling server allocation with sustainability costs
• Minimizing bin usage and linear overfill costs
• Improving algorithms for high-cost utilization scenarios

Innovation

• Introduces green bin packing with linear cost β
• Modifies classic algorithms for βG > 1 cases
• Tests theoretical bounds and empirical performance