Privacy-Preserving Distributed Learning for Residential Short-Term Load Forecasting

TL;DR

The paper tackles privacy in residential short-term load forecasting by introducing a fully decentralized learning framework that integrates secure aggregation with a Markovian switching topology. The Distributed Markovian Switching (DMS) algorithm enables random participation of agent subsets, reducing communication and improving resilience to poisoning and gradient-based attacks, while a Shamir-based secure aggregation protocol protects gradient information. Theoretical convergence is proven under standard assumptions, and empirical case studies on real CER data demonstrate comparable accuracy to centralized and federated baselines with significantly lower communication overhead and enhanced scalability and security. The approach offers a practical pathway to privacy-preserving, scalable load forecasting in distributed energy systems.

Abstract

In the realm of power systems, the increasing involvement of residential users in load forecasting applications has heightened concerns about data privacy. Specifically, the load data can inadvertently reveal the daily routines of residential users, thereby posing a risk to their property security. While federated learning (FL) has been employed to safeguard user privacy by enabling model training without the exchange of raw data, these FL models have shown vulnerabilities to emerging attack techniques, such as Deep Leakage from Gradients and poisoning attacks. To counteract these, we initially employ a Secure-Aggregation (SecAgg) algorithm that leverages multiparty computation cryptographic techniques to mitigate the risk of gradient leakage. However, the introduction of SecAgg necessitates the deployment of additional sub-center servers for executing the multiparty computation protocol, thereby escalating computational complexity and reducing system robustness, especially in scenarios where one or more sub-centers are unavailable. To address these challenges, we introduce a Markovian Switching-based distributed training framework, the convergence of which is substantiated through rigorous theoretical analysis. The Distributed Markovian Switching (DMS) topology shows strong robustness towards the poisoning attacks as well. Case studies employing real-world power system load data validate the efficacy of our proposed algorithm. It not only significantly minimizes communication complexity but also maintains accuracy levels comparable to traditional FL methods, thereby enhancing the scalability of our load forecasting algorithm.

Figures (12)

• Figure 1: Deep leakage from gradient attack.
• Figure 2: Training procedure of federated learning.
• Figure 3: Agent secret sharing a gradient $\phi$ with the parties.
• Figure 4: An illustrative framework of distributed learning.
• Figure 5: Communication topologies of different strategies.