Human activity recognition (HAR) is crucial for real-world applications such as healthcare, surveillance, and smart homes. Among sensing technologies, millimeter wave (mmWave) sensors stand out due to their contactless nature, high sensitivity, and ability to operate in low-light environments while preserving privacy. However, the scarcity of mmWave sensing data limits the generalizability of mmWave-based HAR systems. To address this, we propose mmAP, a data augmentation and pretraining framework that synthesizes a large mmWave dataset using human mesh data, followed by pretraining a robust and general mmWave heatmap encoder using a multi-modal masked autoencoder framework using the synthesized data. We enhance the model’s robustness with heatmap-specific data perturbations and perform task-specific fine-tuning on a small real-world dataset. The experiment results over the baseline demonstrate the effectiveness of the proposed mmAP framework.

The direction of arrival (DOA) and amplitude‐phase (AP) of source signal can be estimated through array signal processing, while the process of information acquisition has not been described by information theory yet. In this study, the authors propose a novel spatial information theory framework to describe the DOA and AP estimation process. Firstly, they give a sensor array system model, where the spatial information is defined mathematically. Furthermore, in a single source scenario, the derivations of theoretical expression and an asymptotic upper bound for DOA information are presented. It also proves that the AP information accords with Shannon channel capacity when the source follows complex‐Gaussian distribution. In addition, when in an uncorrelated multi‐sources scenario, the AP information is the sum of that of each source for Gaussian distributed sources. Finally, they propose a definition of entropy error to evaluate the information acquisition capability of a sensor array. Numerical results verify that the entropy error is consistent with mean square error and approaches Cramér–Rao bound in high signal‐to‐noise ratio. It also verifies the correctness of theoretical derivation and shows that the spatial information can be employed as an evaluation index to quantitatively analyse the system performance which has significant guidance.

Systematic Luby transform (SLT) codes over the additive white Gaussian noise (AWGN) multiple access channels (MACs) are analysed in this study. The authors introduce a novel optimisation model extending from the point‐to‐point AWGN channels to design the degree distributions over the AWGN MACs, which dramatically improve the bit‐error‐rate (BER) performance of SLT codes. First, they extend the Gaussian approximation analysis of SLT codes to the AWGN MACs to analyse the asymptotic BER performance where the symmetric condition of log‐likelihood ratio message can no longer be assumed. Second, based on the asymptotic BER analysis over the AWGN MACs, they extend the existing optimisation model from the point‐to‐point AWGN channels to the AWGN MACs. Simulation results show that the degree distributions obtained in this study outperform other common degree distributions (optimised for single‐user AWGN channels) over the two‐user AWGN MACs.

In this paper, we study systematic Luby transform (SLT) codes over binary input additive white Gaussian noise (BIAWGN) channels. To improve the bit-error-rate (BER) performance of SLT codes on the BIAWGN channel, this paper presents a novel optimization design of degree distributions. First, we apply the Gaussian approximation (GA) to analyze the asymptotic BER performance of SLT codes and calculate overhead thresholds for successful decoding. Second, we derive an approximate closed-form expression for lower bound on BER by applying the GA and further simplify the expression in low and high signal-to-noise ratio regimes, respectively. Third, we adopt the conventional linear programming (CLP) constrained by the GA to optimize the degree distribution of SLT codes. The objective of CLP is to minimize the average degree of the degree distribution that may result in bad error floor. To improve the error floor caused by CLP, we put forth a novel optimization model constrained by the lower bound on BER, and using the model we can design the degree distribution more flexibly with any desired BER. Simulation results show that the proposed degree distribution can provide better BER performance of SLT codes over the BIAWGN channel.