Output-to-spectrum assignment algorithm for a LTE Cognitive Radio filter bank

For mobile user equipments (UE) a careful power management is essential. Despite this fact, quite an amount of energy is wasted in todays UEs analog and digital frontends. Those are engineered for extracting the wanted signal from a spectral environment defined in the corresponding communication standards with their extremely tough requirements. In a real receiving process those requirements can typically be considered as dramatically less critical. Knowledge about the actual environmental spectral conditions allows to reconfigure both frontends to the actual needs and to save energy. In the recent past several approaches solving this issue for Long Term Evolution (LTE) systems have been discussed. The most sophisticated of those is a generic spectrum sensing technique, which we call LTE Cognitive Radio. One problem, which has to be treated in this system, is the assignment of each output in a dyadic filter bank architecture to each individual part of the observed spectrum. This paper focuses on an algorithm being able to map filter outputs to the correct part of the desired spectrum estimate.