Microwave Photonic Direction-Finding Spectrometer

Abstract

In this paper, we propose an architecture wherein the radio-frequency (RF) field at the antenna array is upconverted to an optical carrier and passed through an array of optical fibers, after which an analog Fourier transform is taken in a free-space optical processor. Through the use of multiple temporal dispersion projections, implemented through varied-length optical fiber segments, the locations of RF sources in three-dimensional space spanning angle-of-arrival (AoA) and instantaneous frequency may be determined on a millisecond time scale using commercial computing hardware after detection by a charge-coupled device (CCD) camera. We present a mathematical formulation of the problem, followed by simulated and experimental results showing three-dimensional spatial-spectral localization through the solution to a system of equations brought forth through the use of Fourier optics to process the RF field. I. INTRODUCTIONP ROVISIONS for future wireless networks are driven by the need for ever-increasing data rates [1]-[3]. To address this, emerging wireless networks are increasing their carrier frequencies to the millimeter-wave (mmW) regime, where the hardware requirements in terms of cost, size, weight and power (C-SWAP) to perform the digital beam-forming process become quite demanding. Digital beamforming, along with digital beamspace processing techniques such as BLAST and MUSIC, allow for angle-of-arrival (AoA) and frequency determination and are extensively implemented in the field [4]-[6]. However, digital beamforming techniques generally rely upon recording of high-frequency signals, implemented through high-speed analog-to-digital converters (ADCs). To handle today's broadband data streams, these ADCs have become quite expensive and power-hungry; techniques such as frontend downconversion reduce the required sample rates while introducing local oscillator (LO) synchronization error [7]. Further, performing direction-finding and frequency measurement requires either matrix inversions or fast Fourier transforms, with both operations becoming increasingly complex as array size increases. Additional techniques such as RF lenses alleviate the computational cost of the beamforming step, but are bulky due to their necessary RF wavelength-scale footprint, and can be lossy depending upon the material system [8], [9]. Array windowing or subarray processing offers a reduction in