On the Sensitivity of Firing Rate-Based Federated Spiking Neural Networks to Differential Privacy

TL;DR

This paper investigates how differential privacy mechanisms affect firing-rate signals in federated spiking neural networks, revealing trade-offs between privacy and learning stability in speech recognition tasks.

Contribution

It provides the first analysis of how DP-induced perturbations impact firing-rate-based federated neuromorphic learning, offering practical guidance for balancing privacy and performance.

Findings

Rate shifts increase with stricter privacy budgets.

Aggregation is attenuated under differential privacy.

Client ranking becomes unstable due to rate perturbations.

Abstract

Federated Neuromorphic Learning (FNL) enables energy-efficient and privacy-preserving learning on devices without centralizing data. However, real-world deployments require additional privacy mechanisms that can significantly alter training signals. This paper analyzes how Differential Privacy (DP) mechanisms, specifically gradient clipping and noise injection, perturb firing-rate statistics in Spiking Neural Networks (SNNs) and how these perturbations are propagated to rate-based FNL coordination. On a speech recognition task under non-IID settings, ablations across privacy budgets and clipping bounds reveal systematic rate shifts, attenuated aggregation, and ranking instability during client selection. Moreover, we relate these shifts to sparsity and memory indicators. Our findings provide actionable guidance for privacy-preserving FNL, specifically regarding the balance between privacy strength and rate-dependent coordination.