TinyML Datasets as Enablers of 6G Edge Intelligence: Key Insights and Research Gaps
Abstract
Tiny Machine Learning (TinyML) enables intelligent inference on resource-constrained edge devices, positioning it as a cornerstone for sixth-generation (6G) wireless systems demanding ubiquitous, low-latency, distributed intelligence at unprecedented scale. However, TinyML development is significantly constrained by limited availability of representative datasets and absence of standardized benchmarking protocols aligned with 6G requirements. This survey analyzes 12 representative TinyML datasets across five critical application domains, selected for 6G relevance, edge compatibility, and validation potential. Experimental results demonstrate practical feasibility: 25.9×compression with
Fields of Study & Subjects
Advanced Wireless Communication Technologies
Wireless Signal Modulation Classification
IoT and Edge/Fog Computing
Publication details
- Journal / source
- IEEE Wireless Communications
- Publisher
- Institute of Electrical and Electronics Engineers (IEEE)
- Volume
- 33
- Issue
- 2
- Pages
- 40–49
- Publication date
- 2 February 2026