Quality of Service Assurance for Shared Spectrum Systems

Effective duration of a spectrum opportunity is an important factor for delivering any desired level of Quality-of-Service (QoS) to the secondary users (SU) in shared bands. Previous research efforts addressing the opportunity duration primarily focus on QoS prediction by the SUs using sensing and/or database approach. However, these initiatives to predict the opportunity duration and achievable QoS have a number of issues, such as prediction errors and computational overhead. In this paper, we present a QoS Assurance (QoSA) approach, in place of QoS prediction, centered around the Spectrum Access System (SAS) to get an idea about the opportunity duration with minimal involvement from the primary user (PU). In the existing approaches, the PU is not involved in the SU operation to avoid exposure of sensitive operational information. To address this concern, we propose an "Opportunity Clustering" mechanism that reduces the likelihood of information exposure. This paper mainly focuses on the interference avoidance and achievable SU throughput improvement due to the QoSA approach. We present the QoSA approach and the SAS framework that incorporates the proposed approach. We also present different QoSA schemes to improve the achievable SU throughput. Simulation results validate the improvement offered by the QoSA approach in terms of interference avoidance and throughput achieved by the SU.