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.

翻译：大多数物联网网络仿真器是基于数据包的离散事件系统，其中物理层行为通过解析干扰规则和预计算误差模型进行近似。尽管这能支持大规模实验，但可能遗漏关键的波形级效应，如邻道泄漏、共存信号间的交叉调制干扰以及接收机非理想特性——这些在异构亚千兆赫ISM频段共存场景中至关重要。本文探讨了上述局限性，并提出了IQSim——一种基于同相/正交（IQ）码流混合的仿真范式。不同于通过抽象碰撞模型预测数据包结果，IQSim维护一个共享的复基带IQ流，仿真传输信号经传播处理后以IQ波形形式插入其中，进而由基于软件的接收机或硬件网关完成解调。我们概述了端到端工作流程，包括在线或离线的波形生成、IQ域传播、波形叠加及向网关的交付。此外，我们汇报了初步原型验证结果，证明实时执行的可行性。