Polar codes are closer to the Shannon limit with lower complexity in coding and decoding. As traditional decoding techniques suffer from high latency and low throughput, with the development of deep learning technology, some deep learning-based decoding methods have been proposed to solve these problems. Usually, the deep neural network is treated as a black box and learns to map the polar codes with noise to the original information code directly. In fact, it is difficult for the network to distinguish between valid and interfering information, which leads to limited BER performance. In this paper, a deep residual network based on information refinement (DIR-NET) is proposed for decoding polar-coded short packets. The proposed method works to fully distinguish the effective and interference information in the codewords, thus obtaining a lower bit error rate. To achieve this goal, we design a two-stage decoding network, including a denoising subnetwork and decoding subnetwork. This structure can further improve the accuracy of the decoding method. Furthermore, we construct the whole network solely on the basis of the attention mechanism. It has a stronger information extraction ability than the traditional neural network structure. Benefiting from cascaded attention modules, information can be filtered and refined step-by-step, thus obtaining a low bit error rate. The simulation results show that DIR-Net outperforms existing decoding methods in terms of BER performance under both AWGN channels and flat fading channels.

In the sensor array signal reception system, improving the accuracy of weak-signal detection is crucial. Traditional fixed-step algorithms struggle to balance the convergence rate (CR) and low steady-state error (SSE) owing to their inherent trade-off limitations. To address this limitation, we propose a novel variable-step-size least-mean-square (VSS-LMS) algorithm based on a modified versoria function, specifically redesigned to enhance curvature characteristics. This approach establishes dynamic coupling between error statistics and step-size factors through nonlinear mapping. It derives closed-loop equations governing parameters ($\alpha$, $\beta$, and $\gamma$) relative to the smoothed instantaneous error correlation function. Consequently, an adaptive feedback system is constructed to achieve real-time adjustment through optimal step-size generation. The optimal parameters ($\alpha$, $\beta$, and $\gamma$) are determined through empirical enumeration and analysis of their impact on algorithmic performance. Comparative evaluations against established VSS-LMS algorithms confirm performance: the proposed algorithm accelerates convergence while maintaining a low SSE, and exhibits robust signal recovery capabilities under low-SNR conditions with diverse interference types.

Index-modulated orthogonal frequency division multiplexing (OFDM-IM) has been recognized as a promising multicarrier transmission scheme due to its flexibility and favorable bit error rate (BER) performance. However, for future wireless communication systems requiring high reliability, high spectral efficiency, and low complexity, existing OFDM-IM schemes still face challenges in simultaneously improving spectral efficiency, maintaining diversity gain, and controlling detection complexity at the receiver. To address these issues, this paper proposes a joint-mode and repeated-index modulation-based coordinate interleaved OFDM scheme (MRIM-CI-OFDM). Building upon the shared subcarrier activation pattern (SAP) and coordinate interleaving structure, the proposed scheme introduces cross-cluster mode-pair indexing, enabling information bits to be jointly carried by the SAP domain, mode domain, and constellation symbol domain. This design enhances spectral efficiency while preserving the diversity advantages of coordinate interleaving. Furthermore, a rotated multi-mode constellation construction method based on inter-constellation minimum product distance is developed to improve mode separability. By exploiting the equivalent real-valued orthogonal structure introduced by coordinate interleaving, low-complexity maximum likelihood (ML) and three-stage Max-Log detectors are constructed. Simulation results demonstrate that the proposed low-complexity detectors achieve near-ML detection performance. Additionally, at a spectral efficiency of 1.25 bps/Hz, MRIM-CI-OFDM achieves approximately 3 dB SNR gain over the coordinate-interleaved/repeated-index benchmarks and more than 5 dB gain over conventional OFDM-IM.

UAV communication, leveraging high mobility and flexible deployment, is gradually becoming an important component of 6G integrated air–ground networks. With the expansion of aerial services, air–ground spectrum resources are increasingly scarce, and spectrum sharing and opportunistic access have become key technologies for improving spectrum utilization. Spectrum sensing is the prerequisite for UAVs to perform dynamic access and avoid causing interference to primary users. However, in air–ground links, the channel time variability caused by Doppler effects, carrier frequency offset, and Rician fading can weaken feature separability, making it difficult for deep learning-based spectrum sensing methods to maintain reliable detection in complex environments. In this paper, a dual-stream hybrid-attention spectrum sensing method (DSHA) is proposed, which represents the received signal simultaneously as a time-domain I/Q sequence and an STFT time-frequency map to extract complementary features and employs a hybrid attention mechanism to model key intra-branch dependencies and achieve inter-branch interaction and fusion. Furthermore, a noise-consistent paired training strategy is introduced to mitigate the bias induced by noise randomness, thereby enhancing weak-signal discrimination capability. Simulation results show that under different false-alarm constraints, the proposed method achieves higher detection probability in low-SNR scenarios as well as under fading and CFO perturbations. In addition, compared with multiple typical baselines, DSHA exhibits better robustness and generalization; under Rician channels, its detection probability is improved by about 28.6% over the best baseline.