Channel Quality Estimation for Cognitive Wireless Systems Using Software Defined Radios

Cognitive radio (CR) is considered an enabler of wireless technologies in safety-critical applications such as connected mobility and industrial automation, due to its high reliability and resource efficiency. Here, accurate and fast estimation of the quality of the radio channel is crucial to CR's adaptive capability. In this paper, we propose a channel quality estimation architecture that devises a novel quality metric based solely on physical layer (PHY) measurements. The proposed metric utilizes a decision tree of PHY sub-metrics, without decoding the packets at higher layers. We validate the proposed architecture using a software defined radio (SDR) experimental setup, and investigate real-time capability and hardware implementation complexity. Our measurements show a high correlation between the proposed metric and the application layer quality of service (QoS). Our measurements also show that by using efficient software implementation, complexity is substantially reduced, which enables real-time functionality.