Learned Compression of SAR Phase-History Data: A Rate-Honest Feasibility Study on GOTCHA

合成孔径雷达（SAR）相位历史数据的学习型压缩：基于 GOTCHA 数据集的速率公平可行性研究

Abstract: On-board compression of synthetic aperture radar (SAR) phase history is bandwidth-critical, and block-adaptive quantization (BAQ) remains the operational standard. We test whether a small convolutional autoencoder, with its encoder on the sensor, can compete with BAQ on complex phase-history patches from the AFRL GOTCHA collection. Every method is charged for all transmitted bits, rates are reported in bits per complex sample (b/cs), and detection is scored by one-to-one matching of CA-CFAR detections.

摘要： 合成孔径雷达（SAR）相位历史的星载压缩对带宽要求极高，目前块自适应量化（BAQ）仍是行业标准。我们测试了一种小型卷积自动编码器（其编码器部署在传感器端）在处理来自 AFRL GOTCHA 数据集的复杂相位历史切片时，是否能与 BAQ 竞争。所有方法均按传输的总比特数计费，速率以每复采样点比特数（b/cs）报告，检测性能则通过 CA-CFAR 检测的一对一匹配进行评分。

The autoencoder (28,656 encoder parameters) loses at every rate. At 16 b/cs it reaches -2.87 dB NMSE, against -35.5 dB for 8-bit BAQ with $\pm 3\sigma$ clipping and -41.0 dB with a tuned clipping range. It also loses to a $16 \times 16$ block Karhunen-Loève transform (KLT), a local linear coder with a tenth of its encoder cost (-5.39 dB). Running the network in a companded Fourier domain helps, but its detection F1 remains bounded at 33%.

该自动编码器（拥有 28,656 个编码器参数）在所有速率下均表现不佳。在 16 b/cs 的速率下，其归一化均方误差（NMSE）仅达到 -2.87 dB，而采用 $\pm 3\sigma$ 截断的 8-bit BAQ 可达 -35.5 dB，经过调优截断范围的 BAQ 更可达 -41.0 dB。它甚至输给了 $16 \times 16$ 分块的卡尔胡恩-洛伊夫变换（KLT）——一种编码器成本仅为其十分之一的局部线性编码器（-5.39 dB）。在压扩傅里叶域运行该网络虽有改善，但其检测 F1 分数仍被限制在 33%。

The evidence points to this model, its normalization, and its objective, not to a fundamental limit of learned coding. Per patch, the data have modest lag-1 coherence ($|\rho| \approx 0.3$) and patch-specific spectral concentration. Two findings concern evaluation itself. First, 97% of CFAR crossings on raw $64 \times 64$ patches are border artifacts of the zero-padded detector. Second, on interior cells BAQ’s clipping range decides detection: 8-bit BAQ keeps 69% F1 with tuned clipping but 17% at $\pm 3\sigma$, and at 8 b/cs or less adaptive FFT thresholding preserves more detections than BAQ. We close with an evaluation protocol for learned radar compression.

证据表明，性能瓶颈在于该模型本身、其归一化方式及其目标函数，而非学习型编码的根本局限。在每个切片中，数据具有适度的滞后-1相干性（$|\rho| \approx 0.3$）以及切片特定的频谱集中度。关于评估本身，我们有两项发现：首先，原始 $64 \times 64$ 切片上 97% 的 CFAR 交叉点实际上是零填充检测器产生的边界伪影；其次，在内部单元上，BAQ 的截断范围决定了检测效果：8-bit BAQ 在调优截断下可保持 69% 的 F1 分数，但在 $\pm 3\sigma$ 下仅为 17%；此外，在 8 b/cs 或更低速率下，自适应 FFT 阈值处理比 BAQ 能保留更多的检测结果。最后，我们提出了一套用于学习型雷达压缩的评估协议。