Robust Over-the-Air Computation with Type-Based Multiple Access

TL;DR

This work tackles robust aggregation in over-the-air computation by using Type-Based Multiple Access (TBMA), which encodes data across $L$ radio resources to form a histogram representation $\mathbf{r}$ at the receiver. A decoupled robust-estimation pipeline applies thresholding, percentile truncation, and local outlier compensation to produce $\hat{\mathbf{r}}$, from which the function is estimated as $f \approx \Psi(\hat{\mathbf{r}})$, enabling computation of diverse functions beyond the arithmetic mean. The framework supports Byzantine resilience and reduces the need for precise channel state information, with demonstrated gains in NMSE under attacks and practical viability in Federated Edge Learning (FEEL) on MNIST, where a robust TBMA system maintains performance close to the attacked-noiseless baseline. Overall, TBMA emerges as a scalable, secure aggregation primitive for next-generation wireless networks and FEEL deployments, capable of robust multi-function computation with energy and CSI efficiency advantages.

Abstract

This paper utilizes the properties of type-based multiple access (TBMA) to investigate its effectiveness as a robust approach for over-the-air computation (AirComp) in the presence of Byzantine attacks, this is, adversarial strategies where malicious nodes intentionally distort their transmissions to corrupt the aggregated result. Unlike classical direct aggregation (DA) AirComp, which aggregates data in the amplitude of the signals and are highly vulnerable to attacks, TBMA distributes data over multiple radio resources, enabling the receiver to construct a histogram representation of the transmitted data. This structure allows the integration of classical robust estimators and supports the computation of diverse functions beyond the arithmetic mean, which is not feasible with DA. Through extensive simulations, we demonstrate that robust TBMA significantly outperforms DA, maintaining high accuracy even under adversarial conditions, and showcases its applicability in federated learning (FEEL) scenarios. Additionally, TBMA reduces channel state information (CSI) requirements, lowers energy consumption, and enhances resiliency by leveraging the diversity of the transmitted data. These results establish TBMA as a scalable and robust solution for AirComp, paving the way for secure and efficient aggregation in next-generation networks.

Figures (2)

• Figure 1: for different techniques at $\text{SNR}=30$ dB (lines) and 5 dB (markers).
• Figure 2: Test accuracy across training for an -based system under Byzantine attacks.