The compressive sensing theory is implemented by means of analog-to-information converter (AIC) that provides a digitized and compressed output $y$ directly from its analog input signal $x$ . A reconstruction algorithm, in turn, is needed to obtain a reconstructed signal $\hat {x}$ that estimates $x$ from $y$ . This algorithm can be far away from the AIC, linked through a communication channel, and embedded in a remote machine, for instance. In general, $\hat {x}$ is evaluated in order to test the AIC itself and, in consequence, the performance of the reconstruction algorithm affects the test result of the AIC-under-test and, as a result, can invalidate the AIC test at all. Few state-of-the-art works have proposed the test methods that take into account only the AIC output $y$ . However, they take into consideration a single-sine test signal to follow the IEEE Standard 1241 for analog-to-digital converter testing. In this paper, two new AIC test methods are introduced in which only the AIC outputs, that is, only the compressed outputs $y$ , are taken in account. In the first one, multisine signals can be applied and, in the second one, four-parameters sine-wave fit methods, as recommended by the IEEE Standard 1241, can be applied. Experimental results were obtained for two different AIC implementations (namely, random demodulator and random modulation preintegrator) and the signal-to-noise and distortion ratio (SINAD) of both was measured directly from the AIC output proving that the proposed AIC test methods are effective.

中文翻译：

---

基于 IEEE-STD-1241 的新型模拟信息转换器测试方法

压缩感知理论是通过提供数字化和压缩输出的模数转换器 (AIC) 实现的 $y$ 直接来自其模拟输入信号 $x$ . 反过来，需要重建算法来获得重建的信号 $\hat {x}$ 估计 $x$ 从 $y$ . 例如，该算法可以远离 AIC，通过通信通道链接，并嵌入到远程机器中。一般来说， $\hat {x}$ 被评估是为了测试 AIC 本身，因此，重建算法的性能会影响被测 AIC 的测试结果，从而使 AIC 测试完全无效。很少有最先进的作品提出只考虑 AIC 输出的测试方法 $y$ . 但是，他们考虑了单正弦测试信号，以遵循 IEEE 标准 1241 进行模数转换器测试。本文介绍了两种新的 AIC 测试方法，其中只有 AIC 输出，即只有压缩输出 $y$ , 被考虑在内。在第一个中，可以应用多正弦信号，在第二个中，可以应用 IEEE 标准 1241 推荐的四参数正弦波拟合方法。获得了两种不同 AIC 实现（即随机解调器和随机调制预积分器）的实验结果，并且直接从 AIC 输出测量了两者的信噪比和失真比 (SINAD)，证明所提出的 AIC 测试方法是有效的.