Subspace-Based SNR Estimator for Cognitive Radio and Link Adaptation

Signal-to-Noise Ratio (SNR) parameter represents the main metric to characterize the performance of signal reception. Determining this parameter is of major importance for a wide variety of communication techniques such as spectrum sensing in Cognitive Radio, Link Adaptation, and power allocation. In general, there are two kinds of SNR estimation techniques: Data-Aided (DA) and Blind Estimation (BE). By using Data-Aided estimation (DA), the receiver estimates the SNR based on prior information from the transmitter. On the other hand, by using Blind-Estimation (BE), the receiver does not have any prior-knowledge of transmission parameters. This technique is extremely used for scenarios where transmission parameters are unknown, a common situation on spectrum sensing for non-cooperative applications in Cognitive Radio. However, some reported BE algorithms have been developed exploiting specific properties of some modulation schemes, which also demands some prior knowledge of signal parameters. This work is focused on describing SNR estimation algorithms suitable for several digital and analog modulation schemes. We propose the Subspace-Based SNR estimator for spectrum sensing by using the Energy Detector and Link Adaptation applications. Comparative simulation results regarding estimator performance exhibit the high precision for several channel models. The applicability of this estimator for several analog and digital modulation schemes is also shown as well as proper performance for low SNR levels is obtained, in exchange for higher computational complexity.