The design, analysis, implementation, and measurement of a compact analog complex cross-correlator for passive millimeter-wave imaging applications are presented in this paper. The correlator uses the “add and square” and “subtract and square” detection scheme by adopting the Schottky diodes and RF distribution network. Meanwhile, in order to realize the correlator with a compact size and low cost, commercially available surface mounting devices and multilayer PCB technique are employed. The relationship between the phase error of the correlator and the phase unbalance of the RF distribution network is also analyzed. Single-frequency test is applied to the correlator, and the results of the measurement reveal that the correlator operates well within the frequency range of 1.5-2.5 GHz. Moreover, a measurement system based on an actual imaging scenario is also developed to characterize the performance of the correlator when injected with correlated broadband noise signals, and the measurement results show that the correlator is well suited for imaging applications. The bandwidth of the current version correlator is around 1 GHz, but it could easily be changed by replacing distributed components.

The design and measurement of a power detector circuit based on the 0.15-μm GaAs pHEMT technology and its usage in complex cross correlation measurement are presented in this article. The six-port technology is employed for the calculation of the cross correlation, and the design equations from the requirements of the correlation measurement are provided to guide the power detector design. The relation between the signal-to-noise ratio (SNR) of the correlation results and the load resistor is deduced, which can restrict the resistor value choice for an acceptable SNR. By properly choosing the load resistor and the biasing of the transistor, a power detector circuit is designed with a suppressed output dc voltage variation resulting from the statistic feature of monolithic microwave integrated circuit (MMIC) production. A fabricated detector chip has been mounted onto a printed circuit board and has been characterized. The measurement results show a flat responsivity over 3.5-8 GHz and a 3-nV/√Hz output noise voltage spectrum. The complex cross correlation measurement is implemented using two power detector modules with carefully chosen bias currents. Across the frequency range of 3.5-8 GHz, the measured correlation circles have an amplitude imbalance of 3.5 dB and an axial ratio from 0.85 to 1.15. With an input power of -5 dBm, the measured SNR of the correlator output is above 26 dB.

Wi-Fi-based contactless activity recognition is of great importance to computer-human interaction, accounting for convenience concerns. However, it remains challenging to recognize activities from multiple users due to the multi-path distortion and disruption of Wi-Fi signals. In this paper, we propose a highly universal framework, namely WISDOM, for Wi-Fi based multi-user activity recognition. Specifically, we first leverage an existing model to identify the number of users from the input Wi-Fi signals. Then, we develop a sub-carrier correlation and inversion based sorting algorithm to extract the signal for each user. Finally, we design a neural network, i.e., WISDOM-Net, which is built on a bidirectional gated recurrent unit network incorporated with the attention mechanism and the one dimension convolutional neural network, to recognize the corresponding user activities. Experimental results show that our proposed WISDOM-Net outperforms the existing baselines on both the public and our own datasets. In particular, WISDOM-Net can reach an average recognition accuracy of up to 98.19% and 90.77% in 2-user and 3-user scenarios, respectively.

The design, implementation, measurements, and broadband calibration of a 16‐channel vector modulator (VM) module covering 1.5–2.4GHz band are presented in this study. Since the practical realisation of the VM suffers from the effects of non‐ideal characteristics, a broadband calibration technique is adopted which greatly improves the transmission of phase and amplitude accuracy of the VM over a wide frequency range. The measurement results of the calibrated VM are given to show the high calibration accuracy and efficiency of the proposed technique.

Online multi-object tracking (MOT) aims to maintain consistent target identities across video frames, yet it remains vulnerable to identity switches under occlusion and appearance variation. Many existing trackers rely on single-prototype exponential moving average (EMA) memory, which is efficient but prone to contamination, over-smoothing, and staleness. To address this issue, we propose Hierarchical Multi-Prototype Appearance Memory (HMP), a plug-and-play module for online MOT. HMP separates stable long-term identity anchors from short-term transitional evidence through a multi-prototype long-term memory and a short first-in-first-out (FIFO) queue. A unified joint reliability score governs memory writing and maintenance, and a frozen two-stage association strategy first performs stable primary matching and then allows conservative short-term recovery only on residual cases. Experiments on MOT17 and MOT20 show that HMP improves identity continuity while preserving competitive overall tracking quality. Controlled ablations further support the effectiveness of the proposed memory representation, reliability control, and staged evidence usage under fixed upstream modules.

In this paper, a calibration approach based on transfer function extraction for the Cartesian vector modulator (VM) is presented. Three kinds of VM models’the ideal VM model, the frequency-dependent VM model and the modified frequency-dependent VM model, are introduced in the proposed calibration approach. The calibration approach starts with an initialization of the transfer function of the modified frequency-dependent VM model. Then, the parameters of the transfer function are modified and extracted from the data of the measured transmission state (transmission amplitude and phase) of the actual VM by iteration, until the transmission state predicted by the extracted transfer function agrees well with the measured transmission state. Subsequently, the extracted transfer function of the modified frequency-dependent VM model is capable of describing the transmission characteristics of the actual VM, and the calibrated baseband control voltages for the desired transmission amplitude and phase of the actual VM are able to be obtained by using the extracted transfer function. An actual VM is used as an example to verify this method. By adopting the proposed method, the maximum amplitude and phase errors at different com-plex gain setpoints are reduced to 0.05 dB and 0.30 respectively after only two iteration steps. Since the actual VM is able to be accurately calibrated in only a few iteration steps, the results reveal that high accu-racy and efficiency can be obtained in this calibration technique, which is well suited for applications in-volving high-accuracy calibration, real-time calibration and multichannel VM system calibration.