A Tutorial on SDR-Based NB-IoT PHY: Synchronization, Demodulation, and Validation

Low-Power Wide-Area Networks (LPWANs) have become fundamental to the Internet of Things (IoT), with NB-IoT (Narrowband Internet of Things) standing out due to its seamless integration with cellular infrastructure, enhanced coverage, and support for dense deployments. Despite its commercial proliferation, SDR-based physical layer (PHY) exploration for NB-IoT remains limited, particularly in addressing unique complexities such as narrowband signal processing, cellular-specific synchronization sequences, and stringent link budget requirements. This paper bridges this gap by presenting a comprehensive tutorial on SDR-based NB-IoT PHY implementation, focusing on three pillars: robust time-frequency synchronization under severe fading and interference, efficient channel estimation for coherent detection, and experimental performance validation in real-world scenarios. We introduce a first-of-its-kind end-to-end SDR implementation supporting both single-tone and multi-tone transmissions, leveraging commercial off-the-shelf (COTS) platforms. Our novel signal processing workflow achieves synchronization through NPSS-based auto-correlation and NSSS-driven cell-ID detection while incorporating CFO estimation and compensation to mitigate oscillator mismatches. For uplink processing, we detail preamble detection and demodulation, addressing coverage enhancement (CE) levels and adaptive subcarrier spacing configurations. Extensive experiments conducted in both indoor (LOS/NLOS) and outdoor environments demonstrate reliable performance, with Bit Error Rate (BER) and Block Error Rate (BLER) metrics validating resilience under varying repetition counts and propagation conditions. The tutorial offers actionable insights for optimizing PHY-layer design, validated against 3GPP specifications, and lays the foundation for next-generation NB-IoT systems in emerging applications, such as smart cities and industrial automation.