Speaker
Description
Wavy dark matter candidates, such as axion and dark photon, convert to ordinary photon. Since the frequency of the conversion photon of a meV mass particle is in the radio wave range (O(1 GHz) - O(100 GHz)) and has a narrow frequency peak, spectroscopic search is effective.
Since neither its mass nor its coupling to standard model particles is known, broad coverage of the mass region is important to understand the nature of dark matter. Therefore, a broadband spectrometer is required while maintaining fine frequency resolution to detect the narrow peak. For example, the bandwidth is desired to be ~10 GHz. However, traditional spectrum analyzer typically has narrow bandwidth (a few MHz), or low time efficiency (~1%). Specialized spectrometer optimized for dark matter search is required.
We developed “dSpec”, which is a dead-time free spectrometer based on the fast Fourier transform (FFT) technique. The dSpec achieved a wide bandwidth (10 GHz) under the fine frequency resolution of 40 kHz.
We built all functions on an RFSoC which contains a CPU, FPGA, and high-speed DAC/ADCs on a single chip. To achieve the above specifications, we developed efficient FFT architecture. The architecture consists of parallelization for high throughput, memory-efficient circuit for large-scale FFT operation, and algorithm for computation reduction. Thus, it allows the FFT circuit to be implemented in a single FPGA.
We will present details of the dSpec including its performance evaluations.
