Traditional data stream clustering methods lack online dimensionality reduction capabilities for high-dimensional data, leading to limited clustering performance. To address this issue,a Scalable Subspace Learning for Clustering Data Streams(S²LCStream) method is proposed. Firstly,this method establishes a projection relationship between historical data and new data through scalable subspace learning,projecting the new data into the subspace spanned by historical data to obtain its clustering assignment in real-time. Secondly,to maintain the accuracy of clustering assignments over time, the method performs consistency detection of data distribution on the continuously arriving data stream,capturing concept drifts and adjusting clustering assignments through a backtracking mechanism to adapt to dynamically changing data distributions. Finally,the proposed method is validated on multiple real-world datasets, demonstrating its efficiency in handling high-dimensional data streams. Specifically, S²LCStream maintains high clustering accuracy while efficiently handling concept drift.

For the problems of multi-person vital sign detection at the same distance and extraction of heartbeat signals under the interference of respiratory higher harmonics, an algorithm based on digital beamforming and orthogonal projection filter (OPF ) is proposed. After the radar measures multiple baseband signals, digital beamforming is first performed to obtain the vital sign signals of people by generating multiple beams pointing to the subjects in space. Then, using the OPF, the obtained vital sign signals are projected onto the null space of their respective respiratory higher harmonics to achieve the removal of respiratory higher harmonic interference and the real-time acquisition of heartbeat signals. The results of multi-person experiments show that the proposed algorithm can successfully separate the detection of multi-person vital sign signals and the extraction of heartbeat signals, and the accuracy rate of heartbeat frequency estimation is as high as 99.68%.

In satellite-to-ground communication,during the process of P-band signal crossing the atmosphere to reach the ground array,its direction of arrival(DOA) and polarization angle are changed by ionosphere effects. Therefore,there is an error between the estimated values obtained by the existing DOA-polarization estimation algorithm and the real values. In order to solve the problem,corrections are made from the aspects of DOA and polarization angle. Firstly, considering the ionospheric refraction effect, a signal propagation model of ionosphere is established,based on which a correction method for pitch angle error is proposed. Secondly,regarding the Faraday rotation(FR) effect of ionosphere,the influence of Faraday rotation angle (FRA) on signal polarization is analyzed,and two estimation methods of FRA based on scattering matrix are proposed according to the process of ground target scattering. Simulation results demonstrate that the proposed methods can accurately calculate the correction value of pitch angle and FRA of P-band signal. Under certain conditions,the pitch angle estimation accuracy is improved by about 0.1°,and the FRA estimation accuracy is improved by nearly 1°,so the correction of DOA and polarization angle is realized.

Deep learning (DL) is an effective method for achieving automatic modulation identification (AMI) technology. However，DL methods generally struggle to balance recognition accuracy and efficiency simultaneously. To address this，a lightweight AMI method based on enhanced multi-scale feature fusion is proposed. First，a lightweight multi-scale feature fusion module is designed，which efficiently extracts multi-scale features of modulation signals through a cross-scale convolutional structure，enhancing the model's ability to represent different signal features. Next，an adaptive feature enhancement module is constructed，combining depthwise separable convolution and attention mechanisms to adaptively learn channel weights of key features，highlighting important signal features while reducing interference from irrelevant ones. Finally，a differential balance classifier is designed to focus on recognizing subtle modulation patterns，enabling efficient classification. Experimental results show that the proposed method improves recognition accuracy by an average of 5.91%，reduces the number of parameters by approximately 8.5×10⁵，and decreases iteration time per sample by 0.0624 seconds. Compared with the advanced models，it achieves higher accuracy，faster speed，and fewer parameters.

Multi-dimensional telemetry data pattern mining holds significant importance for satellite status monitoring. However, the sheer volume of telemetry parameters and data poses a challenge in obtaining precise solutions within a short timeframe. To address this issue,the authors propose a matrix profile-based pattern mining approach that employs stochastic principles to search for approximate solutions,which can serve as surrogates for precise solutions within an acceptable error margin. Firstly, spectral analysis is performed on the multi-dimensional telemetry data to determine the template length based on the characteristic frequencies of the patterns. Subsequently,the Mueen's algorithm for similarity search(MASS) is iteratively applied in a stochastic manner to compute elements within the distance matrix. A crucial step involves zeroing out elements near the main diagonal to form the multi-dimensional distance matrix. Finally, the minimum values are extracted from each column to generate the multi-dimensional distance matrix profile(MDMP ) . On this profile, the locations of the maximum and minimum values correspond to the identified rare and frequent patterns, respectively. Experimental analysis indicates that when processing three-dimensional telemetry data containing 150000 sampling points, the proposed method, at a 1% mining depth,is able to constrain the positional error between the approximate and precise solutions within 400 sampling points.

To solve the problem of signal aliasing after sampling in software defined radio (SDR ) , an improved phase-adjusted filtering algorithm is proposed. Based on the second-order RF bandpass sampling front-end with adjustable time delay,an anti-aliasing filter that supports multi-segment filtering is designed, and multiple frequency segments can be flexibly set according to the actual needs,so as to achieve accurate filtering for different frequency bands. Through the simulation verification in MATLAB SIMULINK, the method has better suppression performance than similar filters(38 dB or more) ,and can effectively filter out the required signal without affecting the integrity of other signals,which simplifies the receiving front-end,has strong flexibility and adaptability, and can better support future communication technology and high-density communication connections.

The traditional antenna array wideband signal synthesis performance evaluation method has the problem of low signal synthesis performance evaluation accuracy due to the limited accuracy of the signal-to-noise ratio(SNR) estimation algorithm in the wideband and low SNR scenarios. For above problem, a wideband signal synthesis performance evaluation method of antenna array using power calculation is proposed. The method first simulates multiple intermediate frequency(IF) signals, applies time and phase delays to simulate the time delay and phase difference of the actual antenna received signals, and adds noise to each signal to simulate a low SNR environment. Then, the original and delayed signals are synchronously compensated until convergence using the antenna grouping algorithm to be evaluated. Finally, the synthesized power of the original signal after compensation is calculated and compared with the synthesized power of the ideal signal to obtain the synthesis loss. Simulation experiments results show that under the conditions of signal bandwidth of 250～500 MHz and SNR of -20～0 dB, the method has an improvement of about 1 dB in evaluation accuracy and 0.1 dB² in evaluation stability compared with the wideband signal synthesis performance evaluation method based on SNR, and the improvement effect is more significant with the decrease of signal bandwidth, and the improvement effect is more significant with the decrease of signal bandwidth.