A Low-Complexity Target Swerling Type Estimation Method for Next-Generation OFDM-Based ISAC Systems

Integrating sensing and communications (ISAC) functions over a common waveform has emerged as an important application area for 5G-New Radio (5G-NR) and future 6G orthogonal frequency division multiplexing (OFDM) based systems. While major researches on ISAC focuses on target range, velocity or angle estimation, our work places special emphasis on accurately characterizing the radar cross section (RCS) fluctuation of a target identified by the Swerling type. In this study, a low-complexity, fast and robust Swerling type estimation method (even under low-SNR condition) is proposed by utilizing the normalized variance (NV) statistic calculated over returned signal from the target. The proposed study uses a quasi-monostatic system, with one antenna at the base station transmitting the ISAC waveform and another antenna receiving the backscattered signal from a single airborne target. Our simulation results, supported by theoretical derivations, show that the proposed NV-based approach achieves high estimation performance at low SNR by utilizing a sufficient number of OFDM symbols. Unlike complex methods based on deep learning (DL) or maximum likelihood estimation (MLE), which are widely used in the literature for target classification from radar data, our technique has significantly reduced computational complexity, making it an ideal solution for real-time ISAC applications. Simulation results show that even under low-SNR conditions, the classification accuracy exceeds 95% when a sufficient number of BPSK-OFDM symbols are processed. It has also been shown that this performance can be achieved not only for BPSK but also for other 5G-NR compatible modulation types such as QPSK and 16-QAM.