Power Performance Analysis for a Small L-Band Total Power Radiometer

This paper presents the design and development of a portable, ultra-low power and compact total power L-band radiometer based on an ARM A8 embedded microprocessor, with focus on the digital receiver and signal processor. This study shows a significant tradeoff between power consumption and algorithm performance limiting the processing capability and the system flight time. The processing algorithm, based in C, was tested for different clock frequencies, and high-level software optimizations were applied to reduce the power down to 2 W. The total size of the radiometer, including the antenna array is 30cm x 30cm x 10cm allowing these systems to become portable, low cost, and be carried by small drones to scan a large area. The physical system, including the L-band antenna array is shown in Fig. 1 and Fig. 2.The system was tested with different integration times, and clock frequencies, and it was found that the power consumption varies inversely proportional with the integration time and the processing speed, but the measurement resolution was not significantly affected (∆T< 2 K). Moreover, this approach allows finding the optimal point of work between power consumption and algorithm performance for portable radiometers. Fig. 3 summarizes the values of power consumption at 10 MS/s for each high-level optimization at different clock frequencies and Fig. 4 shows significant reduction in power while the algorithm execution time increases, which unfavorably affect the scan velocity and processing speed of the system. For a radiometer flying at 8 m/s for 30 min, with a footprint of 60m x 60m, the power consumption is calculated for a sampling speed of 10 MS/s, and τ = 4 ms. For a clock speed of 300 MHz, the calculated energy consumption is 1.5 Wh, while scanning 200 footprints. If the clock speed is 1 GHz, the energy consumption is 2.2 Wh, while scanning 410 footprints.