Security-aware selective quantization and optimization framework (SASQOF) for resource-constrained ML-based IoT irrigation systems
Keywords:
Irrigation control, Selective quantization, Security-aware design, Sensor authentication, Edge machine learningAbstract
This study evaluates selective dynamic 8-bit integer (INT8) quantization of long short-term memory (LSTM) recurrent weights while retaining a 32-bit floating-point (FP32) linear output layer, together with authenticated sensor processing, using 10,257 observations from the Secured and Optimized IoT-Based Irrigation System (SOIBIS) experimental farm at Federal University Lokoja, Nigeria. Five seeded runs compare eight models under chronological and feature-group evaluations. The chronological test contains only label pattern 101, for which a majority baseline also achieves 100% exact-match accuracy. On the class-diverse feature-group test, selective quantization achieves 98.40% mean exact-match accuracy, compared with 100% for a decision tree and 84.68% for a majority baseline. Selective quantization reduces serialized model size by 64.9% and single-record host model-only latency by 50.7%, but increases 12-record model-only latency by approximately 41.1%. Mean authenticated single-record inference-path time is 0.441 ms on the host. All 1,200 specified functional packet checks met their expected outcomes. These are host-software findings, not evidence of ESP32 deployment, independently verified pump actuation, water savings, or agronomic superiority.
References
[1] E. O. Oladipe, S. E. Adewumi, T. Kolajo & J. B. Agbogun, ``Crop recommendation in precision agriculture: a systematic literature review of methods, trends, and challenges'', African Scientific Reports 5 (2026) 590. https://doi.org/10.46481/asr.2026.5.3.590.
[2] K. Taueatsoala, C. Daniels, A. J. Ramsunar, P. Bronkhorst & A. E. Ezugwu, ``TinyML-enabled IoT for sustainable precision irrigation'', arXiv preprint (2026), arXiv:2601.13054. Available online: https://arxiv.org/abs/2601.13054.
[3] J. Lin, W.-M. Chen, Y. Lin, J. Cohn, C. Gan & S. Han, ``MCUNet: Tiny deep learning on IoT devices'', Advances in Neural Information Processing Systems 33 (2020) 11711. Available online: https://proceedings.neurips.cc/paper/2020/hash/86c51678350f656dcc7f490a43946ee5-Abstract.html.
[4] R. Bhaumik, B. Chakraborty, W. Choi, A. Dutta, J. Govinden & Y. Shen, ``The committing security of MACs with applications to generic composition'', in Advances in Cryptology---CRYPTO 2024, Springer, Cham, Switzerland, 2024, pp. 425--462. https://doi.org/10.1007/978-3-031-68385-5_14.
[5] G. Selander, J. Mattsson, F. Palombini & L. Seitz, ``Object security for constrained RESTful environments (OSCORE)'', Internet Engineering Task Force, RFC 8613 (2019). https://doi.org/10.17487/RFC8613.
[6] M. I. Akazue, ``A fuzzy based intelligent irrigation system'', SAU Science-Tech Journal 1 (2016) 12. Available online: https://www.researchgate.net/publication/370495372_A_fuzzy-based_intelligent_irrigation_system.
[7] H.-I. Liu, M. Galindo, H. Xie, L.-K. Wong, H.-H. Shuai, Y.-H. Li & W.-H. Cheng, ``Lightweight deep learning for resource-constrained environments: a survey'', ACM Computing Surveys 56 (2024) 267. https://doi.org/10.1145/3657282.
[8] B. Jacob, S. Kligys, B. Chen, M. Zhu, M. Tang, A. Howard, H. Adam & D. Kalenichenko, ``Quantization and training of neural networks for efficient integer-arithmetic-only inference'', Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, USA, 2018, pp. 2704--2713. https://doi.org/10.1109/CVPR.2018.00286.
[9] PyTorch Contributors, ``quantize_dynamic'', PyTorch API documentation. Available online: https://docs.pytorch.org/docs/stable/generated/torch.ao.quantization.quantize_dynamic.html (accessed 16 October 2026).
[10] PyTorch Contributors, ``LSTM'', Dynamic Quantized Modules, PyTorch API documentation. Available online: https://docs.pytorch.org/docs/stable/generated/torch.ao.nn.quantized.dynamic.LSTM.html (accessed 16 October 2026).
[11] J. Li & R. Alvarez, ``On the quantization of recurrent neural networks'', arXiv preprint (2021), arXiv:2101.05453. Available online: https://arxiv.org/abs/2101.05453.
[12] S. Maniam, Y.-K. Tee, E. Memar, H. Y. Wong & M. Zaman, ``Smart irrigation management: IoT-based RNN-LSTM model for soil moisture prediction in precision agriculture'', Smart Agricultural Technology 13 (2026) 101866. https://doi.org/10.1016/j.atech.2026.101866.
[13] S. Han, H. Mao & W. J. Dally, ``Deep compression: compressing deep neural networks with pruning, trained quantization and Huffman coding'', International Conference on Learning Representations, 2016. Available online: https://arxiv.org/abs/1510.00149.
Published
How to Cite
Issue
Section
Copyright (c) 2026 Dauda Olorunkemi Isiaka, Afolayan Ayodele Obiniyi, Victoria Ifeoluwa Yemi-Peters, Joshua Babatunde Agbogun

This work is licensed under a Creative Commons Attribution 4.0 International License.

