Hierarchical Federated Learning Incentivization for Gas Usage Estimation

Has Sun,Xiaoli Tang,Chengyi Yang,Zhenpeng Yu,Xiuli Wang,Qijie Ding,Zengxiang Li,Han Yu

2023-07-01

Abstract:Accurately estimating gas usage is essential for the efficient functioning of gas distribution networks and saving operational costs. Traditional methods rely on centralized data processing, which poses privacy risks. Federated learning (FL) offers a solution to this problem by enabling local data processing on each participant, such as gas companies and heating stations. However, local training and communication overhead may discourage gas companies and heating stations from actively participating in the FL training process. To address this challenge, we propose a Hierarchical FL Incentive Mechanism for Gas Usage Estimation (HI-GAS), which has been testbedded in the ENN Group, one of the leading players in the natural gas and green energy industry. It is designed to support horizontal FL among gas companies, and vertical FL among each gas company and heating station within a hierarchical FL ecosystem, rewarding participants based on their contributions to FL. In addition, a hierarchical FL model aggregation approach is also proposed to improve the gas usage estimation performance by aggregating models at different levels of the hierarchy. The incentive scheme employs a multi-dimensional contribution-aware reward distribution function that combines the evaluation of data quality and model contribution to incentivize both gas companies and heating stations within their jurisdiction while maintaining fairness. Results of extensive experiments validate the effectiveness of the proposed mechanism.

Machine Learning,Information Theory

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper aims to address several key issues in gas usage estimation and proposes a Hierarchical Federated Learning Incentive Mechanism for Gas Usage Estimation (H I-GAS) to solve these problems. Specifically:

1. **Data Privacy Issue**:
- Traditional gas usage estimation methods rely on centralized data processing, which poses privacy risks. Federated Learning (FL) allows each participant to process data locally, thereby protecting data privacy.

2. **Data Island Issue**:
- The data owned by a single company may not be sufficient to effectively train a model, as data is typically owned by different organizations. Collaborative model training can utilize data resources from multiple organizations, improving model quality.

3. **Participation Incentive Issue**:
- In federated learning, data owners need to consume their own computational and communication resources for local training. Without sufficient economic compensation, self-interested participants may be unwilling to contribute resources to federated learning model training. Additionally, some unreliable participants may send low-quality updates, affecting the performance of the global model.

The paper proposes the H I-GAS mechanism, based on a hierarchical federated learning ecosystem, including Horizontal Federated Learning (HFL) and Vertical Federated Learning (VFL). The mechanism designs a multi-dimensional contribution-aware reward allocation function, considering data quality and model contribution, to incentivize gas companies and heating stations to actively participate in federated learning training and submit high-quality data. Experimental results validate the effectiveness of this mechanism.