The indirect test offers a cost-effective alternative for the specification test of analog integrated circuits (ICs). It infers high-cost specifications from low-cost measurements through a mapping function. Deep learning provides an efficient means for modeling this mapping. However, batch-to-batch variations in IC manufacturing often lead to distribution shifts, which degrade the accuracy of such data-driven techniques. To address this, we propose a transfer learning-based framework—Bayesian model calibration (BMC)—for the indirect test. During testing, a deep neural network (DNN) is trained with early data for estimation. When a new-batch arrives, a small number of samples are conventionally tested to form a calibration set. The original estimator is updated with the new dataset via a Bayesian loss function. This loss function integrates a prior term (reflecting the early data) and a likelihood term (reflecting the new data), enabling a principled tradeoff between prior knowledge and new observations. The estimator thereby adapts its parameters to improve accuracy on the new batch. Moreover, by adjusting the relative weights of the prior and likelihood terms, the extent of model update can be controlled to match the degree of distribution shift, achieving optimal calibration. The proposed approach is evaluated on a simulated operational amplifier and two commercial IC datasets to demonstrate its advantage. The BMC can improve accuracy on new batches with minimal additional samples. Moreover, by flexibly integrating the prior knowledge and new data, it outperforms baseline methods such as direct data combination (DC) and simple fine-tuning (SFT).

Automatic modulation classification (AMC) plays a vital role in Internet of Things (IoT) and Cognitive Radio (CR) to ensure communication security, but the AMC technology is suffering from low signal-to noise ratio (SNR), non-ideal defects in communication systems, and so forth. In order to suppress the impact of interferences on AMC and improve classification performance, we propose a novel multi-dimensional image-based AMC method: Multi-dimensional Image Information Fusion Network (MIIF-Net). Based on the analysis of signal model, this work focuses on suppressing noise and carrier frequency offset (CFO) to improve the classification accuracy of AMC. Motivated by multi-dimensional feature strategy, modulation signals are converted and encoded into multi-dimensional images to introduce multi-domain information and suppress CFO. From the perspective of image enhancement, a novel noise suppression method is proposed, which does not rely on any prior information about signals. In addition, residual network (ResNet) is utilized as encoder to extract distinctive feature in different dimensions, and channel attention mechanism (CAM) is employed to fuse multi-dimensional features. The multi-layer perceptron is utilized for modulation classification in MIIF-Net. MIIF-Net is evaluated in public dataset. Simulation results demonstrate that the proposed MIIF-Net achieves superior performance, especially under low SNR levels. MIIF-Net is further evaluated in practical acquired data. The experimental result shows that MIIF-Net exhibits the consistent advantage under low SNR levels.

The estimation model-based alternate test strategy for analog integration circuits (ICs) offers an effective way to reduce test costs. However, as a data-driven method, the estimation process of alternate test is invisible, leading to low test reliability. Hence, to address this problem, we propose an adaptive confidence interval-based alternate test (ACIT) to enhance the reliability of alternate test. Multiestimators are implemented to generate target parameters of each circuit synchronously. All estimations for the same sample are averaged to get the final result. The reliability of each final estimation is evaluated by comparing its adaptive confidence interval to the correlated parameter boundary. The proposed adaptive confidence interval is obtained from the variance of multioutputs and estimation-boundary distance. Estimations with confidence intervals crossing boundaries are identified as suspect results and returned to repeat testing by the conventional approach. The remaining results are classified as “pass” (entire confidence intervals within the qualified range) or “fail” (entire confidence intervals outside of the qualified range). Our approach is studied with simulation data and verified on commercial ICs. Results demonstrate that the ACIT can eliminate the misclassification circuits effectively by identifying unreliable estimations.

Machine learning-based alternate test of analog/mixed-signal integrated circuits (ICs) has been widely studied in the last decade, which has the benefits of simplifying test equipment and decreasing test costs. However, due to low reliability and accuracy, it is hard to adopt the alternate test technique in the industry. In this article, a model splitting approach (MDSP approach) is proposed to improve the reliability and accuracy of the alternate test. The machine learning-based estimation model is “split” into two models with “complementary” performance (a “positive” model and a “negative” model). The “positive” model generates estimations that are no smaller than label values, while the “negative” model outputs estimations that are no larger than label values. Estimations with excessive differences between two models are identified as suspected estimations with large errors and filtered out. The rest results of “complementary” models are averaged to generate the final estimations. By comparing estimations of two models, the estimations with large error are filtered out effectively, and the estimation accuracy is improved significantly by fusing the results of two estimators. The MDSP approach is investigated with data from the commercial analog-to-digital converter and operational amplifier (OP). Results demonstrated that the proposed approach can improve test reliability and accuracy significantly.

This paper presents a 48×32 elements dual-polarized transmit (Tx) phased array for Ku-band (13.5-14.5GHz) satellite communication (SATCOM). In order to achieve a low profile, the array uses a multilayer printed circuit board (PCB) with the antenna elements on the front side, the silicon Tx beamformer chips surface-mounted on the back side, and the beamformers' control and power supply routings are also integrated into the PCB. Each beamformer has 8 channels connected to the vertical and horizontal polarization of the surrounding 4 antenna elements, and the array can be circular or linear polarization by changing the phase of the channels. The total Tx phased array consists of 12 subarrays of 8×16 dual-polarized antenna elements, which is designed to reduce the cost of production. The antenna elements are arranged in a regular rectangle with a spacing of 10.5 mm, allowing a scanning range of ±60° in the operating band. The effective isotropic radiated power (EIRP) of the Tx array is higher than 82.2 dBm at 13.5-14.5 GHz, which allows for a good realization of communications between ground equipment and satellites.

This paper provides a sampling system noise decomposition strategy based on a visual geometry group (VGG) convolutional neural network and a long short-term memory (LSTM) prediction network. Three typical noise components in acquisition systems—Gaussian white noise, quantization noise, and device 1/f noise—are considered in our analysis. The proposed approach is separated into two parts. The first is the classification system, which adopts the independent component analysis (ICA) method to preprocess the samples before employing the VGG16 network to perform autonomous classification of noise components. The estimation of noise power is the second task, in which 14 features of the acquired signals are fed to the LSTM prediction network to estimate the power of each noise component. The simulation results demonstrate that the suggested method has reliable classification results for mixed signals comprising the three noise components and can estimate the power of each component accurately.