Most Internet of Things (IoT) network simulators are packet-level discrete-event systems in which physical-layer (PHY) behavior is approximated through analytical interference rules and precomputed error models. While this enables scalable experiments, it can miss key waveform-level effects such as adjacent-channel leakage, cross-modulation interference between coexisting signals, and receiver imperfections, which are critical in heterogeneous sub-GHz ISM-band coexistence scenarios. This paper discusses these limitations and introduces IQSim, a simulation paradigm based on in-phase/quadrature (IQ) stream mixing. Instead of predicting packet outcomes from abstract collision models, IQSim maintains a shared complex baseband IQStream into which simulated transmissions are inserted as IQ waveforms after propagation processing, and then demodulated by software-based receivers or hardware gateways. We outline the end-to-end workflow, including online or offline waveform generation, IQ-domain propagation, waveform superposition, and delivery to gateways. We also report preliminary prototype results supporting the feasibility of real-time execution.
@article{arxiv.2604.06408,
title = {Towards Realistic Waveform-Level IoT Network Simulation via IQ Mixing},
author = {Alexis Delplace and Samer Lahoud and Kinda Khawam and Dominique Quadri},
journal= {arXiv preprint arXiv:2604.06408},
year = {2026}
}
Comments
Accepted for publication at the 2026 IEEE International Symposium on Spectrum Innovation (DySPAN 2026). 5 pages, 3 figures