Learner's Dilemma: IoT Devices Training Strategies in Collaborative Deep Learning

TL;DR

This paper models the training behavior of IoT devices in collaborative deep learning using game theory, proposing a fair strategy to enhance cooperation among devices while balancing accuracy and overhead.

Contribution

It introduces a game-theoretic framework to analyze IoT device training strategies in CDL and proposes a cluster-based fair solution to promote cooperation.

Findings

80% of devices are willing to cooperate in CDL

A novel game-theoretic model for IoT device training behavior

Proposed strategy improves cooperation fairness in real-world deployment

Abstract

With the growth of Internet of Things (IoT) and mo-bile edge computing, billions of smart devices are interconnected to develop applications used in various domains including smart homes, healthcare and smart manufacturing. Deep learning has been extensively utilized in various IoT applications which require huge amount of data for model training. Due to privacy requirements, smart IoT devices do not release data to a remote third party for their use. To overcome this problem, collaborative approach to deep learning, also known as Collaborative DeepLearning (CDL) has been largely employed in data-driven applications. This approach enables multiple edge IoT devices to train their models locally on mobile edge devices. In this paper,we address IoT device training problem in CDL by analyzing the behavior of mobile edge devices using a game-theoretic model,where each mobile edge device aims at maximizing the accuracy of its local model at the same time limiting the overhead of participating in CDL. We analyze the Nash Equilibrium in anN-player static game model. We further present a novel cluster-based fair strategy to approximately solve the CDL game to enforce mobile edge devices for cooperation. Our experimental results and evaluation analysis in a real-world smart home deployment show that 80% mobile edge devices are ready to cooperate in CDL, while 20% of them do not train their local models collaboratively.