System Level Design of RF Receivers Based on Non Linear Optimization and Power Consumption Models

Abstract

This work presents a method to optimally distribute the block specifications in an RF receiver in order to reduce power consumption. The parameters are Gain (G), Noise Figure (NF) and Input Third Order Intercept Point (IIP3). The method is based on setting the signal quality per block at the output; Signal to Noise Ratio (SNR) for noise and Signal to Noise plus Distortion Ration (SNDR) for linearity. Both are limited in order to fulfill the sensitivity and intermodulation tests of a given standard. Non linear power models can be used as the method is based on heuristics associated with non linear optimization. First, random valid sets are tested "A" times, while the best candidate is chosen as starting point for a non linear optimization with bounds based on interior point algorithm. The process is repeated "B" times, and the best candidate is chosen. To validate the method, a direct-conversion receiver was dimensioned for the Long Term Evolution (LTE) and Bluetooth Low Energy (B-LE) standards. Two power models were used, labeled PM1 and PM2. First the LTE is considered. When compared to predetermined signal quality distributions, the method reduced the power consumption by 65% and 41%, considering PM1 and PM2, respectively. Then the B-LE is chosen with power PM2. This model is linear and has an analytical minimum derived in the literature. The optimization achieved a precision of 0.2% to the analytical minimum using A=1000 and B=15.