Abstract
—Frequency Diverse Array (FDA) radar uses small frequency offsets across antenna array elements to generate a range-angle dependent beampattern. A well-known limitation, however, is that the received echo is time-variant, preventing direct target localization. In this work, we experimentally demonstrate that time variance can be effectively cancelled after reception, yielding a stable, time-invariant range-angle map that accurately reflects the true target location. Implementation of this concept gives rise to some subtle experimental challenges which are caused by the hardware complexity of FDA radar. We implement the transceiver using two Universal Software Radio Peripherals (USRPs). This configuration causes several practical challenges, including random initial phase offsets across Rx channels and the limited instantaneous bandwidth available for realizing a short-range laboratory prototype. These challenges are addressed through offline processing of independently collected I/Q samples for each FDA frequency offset. This approach alleviates the hardware constraints but introduces new difficulties, such as target motion synchronization and maintaining positional accuracy, which we analyze and mitigate through careful calibration and measurement design. To validate the proposed methodology, we first conduct wired-channel experiments, where RF cables of varying lengths emulate controlled, time-invariant range-angle steering. We then perform free-space FDA radar measurements and compare the resulting range-angle maps with simulations, evaluating localization accuracy and Peak Side Lobe Level (PSLL). The results demonstrate the practical feasibility of low-complexity, short-range, time-invariant FDA radar prototyping.
Funding Information
  • National Research Foundation of Korea (NRF) through Korean Government (2022R1F1A1072478)

This publication has 48 references indexed in Scilit: