Backdoor Attacks on Contrastive Continual Learning for IoT Systems

TL;DR

This paper analyzes how contrastive continual learning in IoT systems can be exploited by backdoor attacks, highlighting vulnerabilities and proposing a layered taxonomy for understanding and defending against such threats.

Contribution

It provides a formal analysis of embedding-level backdoor attacks on contrastive continual learning in IoT, introduces a taxonomy, and evaluates defense strategies under IoT-specific constraints.

Findings

CCL can be vulnerable to persistent backdoor attacks.

IoT constraints affect the effectiveness of defense strategies.

Contrastive learning's geometric properties can be exploited by attackers.

Abstract

The Internet of Things (IoT) systems increasingly depend on continual learning to adapt to non-stationary environments. These environments can include factors such as sensor drift, changing user behavior, device aging, and adversarial dynamics. Contrastive continual learning (CCL) combines contrastive representation learning with incremental adaptation, enabling robust feature reuse across tasks and domains. However, the geometric nature of contrastive objectives, when paired with replay-based rehearsal and stability-preserving regularization, introduces new security vulnerabilities. Notably, backdoor attacks can exploit embedding alignment and replay reinforcement, enabling the implantation of persistent malicious behaviors that endure through updates and deployment cycles. This paper provides a comprehensive analysis of backdoor attacks on CCL within IoT systems. We formalize the objectives of embedding-level attacks, examine persistence mechanisms unique to IoT deployments, and develop a layered taxonomy tailored to IoT. Additionally, we compare vulnerabilities across various learning paradigms and evaluate defense strategies under IoT constraints, including limited memory, edge computing, and federated aggregation. Our findings indicate that while CCL is effective for enhancing adaptive IoT intelligence, it may also elevate long-lived representation-level threats if not adequately secured.