Securing Federated Learning in Robot Swarms using Blockchain Technology

TL;DR

This paper demonstrates a decentralized federated learning approach for robot swarms using blockchain technology, ensuring secure model aggregation without central servers, and addresses robustness against malicious robots.

Contribution

It introduces a blockchain-based method for secure, decentralized federated learning in robot swarms, including protection mechanisms against malicious interference.

Findings

Blockchain enables secure model sharing in robot swarms.

Malfunctioning robots can disrupt training without protection.

Protection mechanisms improve robustness against malicious robots.

Abstract

Federated learning is a new approach to distributed machine learning that offers potential advantages such as reducing communication requirements and distributing the costs of training algorithms. Therefore, it could hold great promise in swarm robotics applications. However, federated learning usually requires a centralized server for the aggregation of the models. In this paper, we present a proof-of-concept implementation of federated learning in a robot swarm that does not compromise decentralization. To do so, we use blockchain technology to enable our robot swarm to securely synchronize a shared model that is the aggregation of the individual models without relying on a central server. We then show that introducing a single malfunctioning robot can, however, heavily disrupt the training process. To prevent such situations, we devise protection mechanisms that are implemented through secure and tamper-proof blockchain smart contracts. Our experiments are conducted in ARGoS, a physics-based simulator for swarm robotics, using the Ethereum blockchain protocol which is executed by each simulated robot.