Abstract

The rise of the number of IoT devices in heavily congested wireless networks has exacerbated the problem of interference, further degrading the quality of performance and trust in the communication reliability of the wireless networks. DFH, as an interference mitigation solution, allows devices to switch frequency bands with overlaid real-time spectrum monitoring to DFH's burst-mode/spectral scan analysis. In this work, we focus on the scalability, low latency, and flexibility aspects of DFH that further enhance its viability in IoT networks. DFH is distinct from fixed approaches to frequency allocation, as the technique considers freed frequency bands that may be allocated or avoided for congestion, even though they are heavily attributed to noise, data fidelity, and packet loss are maintained. We propose an improved framework with interference pattern-point prediction based on spectrometric sensing and optimization of hopping sequence ML for DFH. The simulations demonstrate substantial gains in noise ratio, throughput, and energy efficiency of IoT networks implemented in DFH, particularly for very high densities of co-channel interfering devices. The proposed controller is also under constraints of computational power and energy consumed by horizon-derived IoT devices. This work highlights the efficiency of DFH as a lightweight and robust solution for maintaining uninterrupted, interference-free, broadband-qualified communication throughout large-scale IoT systems in smart cities, industrial automation, and healthcare monitoring.