Relaxing pre-selector filter selectivity requirements using cognitive RF front-end control

This paper proposes the use of a cognitive engine to control the local oscillator and sampling frequencies in a flexible receiver RF front-end. The analysis in this paper shows that this cognitive engine has the potential to relax selectivity requirement of the pre-selector filter in receiver RF front-end. The cognitive engine is designed by modeling the RF front-end in channelized spectrum domain. The paper also develops a new spectrum occupancy model to evaluate the performance of the approach. Theoretical analysis and simulations are also carried out using the developed model. The results show that the designed cognitive engine can enable a poorly selective receiver to behave similar to highly selective receiver. Furthermore, this paper analyzes the computational complexity of the designed cognitive engine.