An Experimental Study of Split-Learning TinyML on Ultra-Low-Power Edge/IoT Nodes
It addresses memory and compute limitations for edge/IoT deployments, but is incremental as it benchmarks existing methods on new hardware.
This paper tackles the challenge of running deep learning on ultra-low-power edge/IoT nodes by implementing a split-learning TinyML testbed on ESP32-S3 boards, benchmarking it with a MobileNetV2 model and finding that ESP-NOW achieves the best round-trip latency of 3.7 seconds.
Running deep learning inference directly on ultra-low-power edge/IoT nodes has been limited by the tight memory and compute budgets of microcontrollers. Split learning (SL) addresses this limitation in which it executes part of the inference process on the sensor and off-loads the remainder to a companion device. In the context of constrained devices and the related impact of low-power, over-the-air transport protocols, the performance of split learning remains largely unexplored. TO the best of our knowledge, this paper presents the first end-to-end TinyML + SL testbed built on Espressif ESP32-S3 boards, designed to benchmark the over-the-air performance of split learning TinyML in edge/IoT environments. We benchmark the performance of a MobileNetV2 image recognition model, which is quantized to 8-bit integers, partitioned, and delivered to the nodes via over-the-air updates. The intermediate activations are exchanged through different wireless communication methods: ESP-NOW, BLE, and traditional UDP/IP and TCP/IP, enabling a head-to-head comparison on identical hardware. Measurements show that splitting the model after block_16_project_BN layer generates a 5.66 kB tensor that traverses the link in 3.2 ms, when UDP is used, achieving a steady-state round-trip latency of 5.8 s. ESP-NOW presents the most favorable RTT performance 3.7 s; BLE extends battery life further but increases latency beyond 10s.