DVB-T signal detection for TV White Space applications

Abstract

This thesis investigates advanced detection methods for Digital Video Broadcasting-Terrestrial (DVB-T) signals within TV White Space (TVWS) applications, driven by the growing demand for efficient spectrum utilization in response to rising wireless communication usage. TVWS, consisting of unused spectrum, opens an opportunity for dynamic secondary spectrum usage, provided that primary DVB-T transmission can be reliably detected and protected from interference.

The study focuses on the implementation of robust spectrum sensing techniques capable of accurately detecting DVB-T signals under challenging environmental conditions such as low Signal-to-Noise Ratio (SNR), multipath fading, and adjacent channel interference. Welch’s method for spectrum estimation, followed by window-based (WIBA) spectrum detection and Constant False Alarm Rate (CFAR) based thresholding approaches as Forward Consecutive Mean Excision (FCME) and Order Statistics method has been systematically evaluated and compared through simulations. The results demonstrate high detection accuracy and reliability, also under low SNR scenarios, highlighting the effectiveness of each method. This improved detection performance supports broader cognitive radio deployments and enables efficient dynamic spectrum access.