Deep learning based channel prediction has garnered significant attention to mitigate channel aging in high-mobility multiple-input multiple-output (MIMO) systems. However, existing channel prediction methods extract the temporal correlations from the channel sequences estimated at uniform pilots, which require dense pilot configuration to mitigate Doppler aliasing in high-mobility scenarios and incur substantial estimation overhead. To tackle this problem, we propose a channel prediction method based on continuous-time transformer with the non-uniform pilot pattern, thereby enabling accurate prediction across arbitrary time scales with only a small number of pilots. Specifically, we first design the non-uniform pilot pattern based on Chebyshev polynomial roots and then prove its optimality under Doppler-dominated channel variations with relatively stable user velocity, wherein a subset of pilots are densely configured to provide a finer resolution of Doppler phase estimation. To adapt to the non-uniform pattern, a continuous-time transformer is further proposed, which integrates the superior feature extraction capability of transformer with the continuous-time modeling strength of neural ordinary differential equation (ODE) for flexibly processing the estimated channel sequences with non-uniform time scales. More concretely, the attention mechanism is extended to the continuous-time domain by incorporating neural ODE, while a high-frequency temporal encoding is designed to fit rapidly time-varying channels. Besides, an element-wise prediction mechanism is proposed to efficiently capture temporal correlations and prevent overfitting. Simulation results demonstrate that our proposed method can realize accurate continuous-time channel prediction in high-mobility scenarios, and significantly outperforms existing channel prediction methods.

For multiple-input multiple-output systems, channel prediction is crucial for mitigating channel aging in mobile scenarios. The existing channel prediction schemes typically require strictly equal sampling intervals of historical and predicted channel sequences, which imposes enormous pilot overhead in high-mobility scenarios with frequent channel estimation. To tackle this problem, we investigate the super-resolution channel prediction, where the future channel sequence is predicted at a finer temporal resolution without additional channel estimation. Specifically, we theoretically analyze the physics process underlying super-resolution channel prediction to show that the measurement of Doppler phase rotation faces the challenging issue of phase ambiguity in high-mobility and high-frequency scenarios. To address this issue, a deep learning-based dual-band fusion approach is proposed to adaptively integrate the low-frequency information for accurate Doppler phase measurement. To realize accurate channel prediction at a finer temporal resolution, we propose the physics feature-inspired neural ordinary differential equation with modulated-periodic-based multi-layer perceptron for effectively learning the dynamics of fast time-varying channels. Simulation results verify that our proposed scheme outperforms existing channel prediction schemes and it maintains robust performance in high-mobility scenarios.

Benefiting from a large amount of unallocated bandwidth, millimeter-wave (mmWave) communications have been regarded as one of the most promising technologies. To overcome high pathloss of mmWave signals, the beamforming technique plays a fundamental role. In recent years, with the success of deep learning (DL), DL-based beam prediction methods have been widely studied to reduce the training overhead of traditional beam scanning methods. In this paper, a novel DL-based low-overhead beam prediction scheme is proposed, which is motivated by two important observations: (1) The optimal beam prediction is difficult for non-line of sight (NLOS) scenario, which limits the overall prediction accuracy. (2) On the contrary, the optimal beam can be precisely predicted with low computational costs under line of sight (LOS) scenario. Therefore, we propose a flexible network architecture, namely multi-stage network (MSN), to conduct the optimal beam prediction. Firstly, MSN contains multiple branches with gradually increasing computational complexity, and each branch carries with a classifier, which enables the MSN to have the capability of adaptively and dynamically allocating computational resources. Meanwhile, to combine the advantages of convolutional neural network (CNN) and transformer for feature extraction in MSN, we design joint CNN and transformer (JCT) module and its simplified module, namely Ghost-JCT. Secondly, we propose two pre-training strategies to effectively improve the performance of classifiers without additional computational costs. Finally, we propose confidence-based and Markov-based classifier selection strategies, which could select the appropriate classifier to strike a balance between accuracy and computational complexity. Simulation results demonstrate that MSN enjoys significant superiority in terms of computational complexity and prediction accuracy compared to its traditional counterparts.

The latest TypeII codebook selects partial strongest angular-delay ports for the feedback of downlink channel state information (CSI), whereas its performance is limited due to the deficiency of utilizing the correlations among the port coefficients. To tackle this issue, we propose a tailored autoencoder named TypeII-CsiNet to effectively integrate the TypeII codebook with deep learning, wherein three novel designs are developed for sufficiently boosting the sum rate performance. Firstly, a dedicated pre-processing module is designed to sort the selected ports for reserving the correlations of their corresponding coefficients. Secondly, a position-filling layer is developed in the decoder to fill the feedback coefficients into their ports in the recovered CSI matrix, so that the corresponding angular-delay-domain structure is adequately leveraged to enhance the reconstruction accuracy. Thirdly, a two-stage loss function is proposed to improve the sum rate performance while avoiding the trapping in local optimums during model training. Simulation results verify that our proposed TypeII-CsiNet outperforms the TypeII codebook and existing deep learning benchmarks.

Deep learning based channel state information (CSI) acquisition and feedback in frequency division duplex systems have drawn much attention in the beyond fifth-generation (B5G) wireless systems. In this paper, we focus on exploiting the CSI codebook in B5G wireless standards with deep learning to enhance the performance of CSI acquisition and feedback. Specifically, the angular-delay-domain partial reciprocity between uplink and downlink channels is considered, and part of angular-delay-domain ports are selected for measuring and feeding back the downlink CSI, where the performance of the conventional deep learning methods is limited due to the deficiency of sparse structures. To address this issue, we propose the new paradigm of adopting deep learning to improve the performance of CSI codebook. Firstly, considering the relatively low signal-to-noise ratio of uplink channels, deep learning is utilized to refine the selection of the dominant angular-delay-domain ports, where the focal loss is harnessed to solve the class imbalance problem. Secondly, we propose to reconstruct the downlink CSI by way of deep learning based on the feedback of CSI codebook at the base station, where the information of sparse structures can be effectively leveraged. Finally, a weighted shortcut module is designed to facilitate the accurate reconstruction, and a two-stage loss function with the combination of the mean squared error and sum rate is proposed for adapting to actual multi-user scenarios. Simulation results demonstrate that our proposed angular-delay-domain port selection and CSI reconstruction paradigm can improve the sum rate performance by more than 10% compared with the standard CSI codebook and traditional deep learning benchmarks.

In non-cooperative communications, malicious electromagnetic interference attacks communication systems and causes higher probability of communication disruption. In order to address the challenges posed by electromagnetic interference, the wireless interference recognition technique has emerged, which identifies the interference signals without priori information. In recent years, the success of deep learning (DL) has sparked interest in introducing DL in the field of wireless interference recognition. However, most DL-based interference identification methods improve accuracy by dramatically increasing network sizes while ignoring the important effect of network inputs. For this reason, we extensively investigate the impact of different signal transformation forms of interference (called signal modalities) on performance. The artificial features of the interference signal are also utilized as one of the refined modalities, which breaks the inherent concept that artificial features are only used in the methods of feature extraction. Convolution and transformer are combined in the extraction of different modal features. In order to reduce the complexity of transformer, a dual transformer module (DTM) is proposed. Furthermore, to overcome the imbalance of modal optimization during the training process, an adaptive gradient modulation (AGM) strategy is proposed, which leads to better convergence for the multimodal training. Finally, modal information selection mechanism (MISM) selects the most appropriate modalities for each input sample, which saves computational costs. Extensive experiments demonstrate that combining multiple interference modalities is more effective than trying different networks.

The Type-II codebook in Release 17 (R17) exploits the angular-delay-domain partial reciprocity between uplink and downlink channels to select part of angular-delay-domain ports for measuring and feeding back the downlink channel state information (CSI), where the performance of existing deep learning enhanced CSI feedback methods is limited due to the deficiency of sparse structures. To address this issue, we propose two new perspectives of adopting deep learning to improve the R17 Type-II codebook. Firstly, considering the low signal-to-noise ratio of uplink channels, deep learning is utilized to accurately select the dominant angular-delay-domain ports, where the focal loss is harnessed to solve the class imbalance problem. Secondly, we propose to adopt deep learning to reconstruct the downlink CSI based on the feedback of the R17 Type-II codebook at the base station, where the information of sparse structures can be effectively leveraged. Besides, a weighted shortcut module is designed to facilitate the accurate reconstruction. Simulation results demonstrate that our proposed methods could improve the sum rate performance compared with its traditional R17 Type-II codebook and deep learning benchmarks.

Multiple-input multiple-output (MIMO) and low-density parity check (LDPC) codes are two of the fundamental technologies in the fifth-generation (5G) networks, where an efficient power allocation scheme is desired to minimize the bit error rate (BER) of the LDPC-coded MIMO system. However, the conventional power allocation methods do not take into account the constraint of modulation and coding scheme (MCS), which may degrade the BER performance. To solve this issue, we propose a deep learning based method to predict the efficient power allocation scheme in coded MIMO systems. Specifically, a neural network is built to learn the complex BER-SNR function to derive the power allocation ratio between the parallel MIMO streams, where the training label is acquired based on the exhaustive searching algorithm. Simulation results show that our proposed method could achieve better BER performance than its conventional counterparts.

In this paper, motivated by the inter-base station (BS) channel dependence due to the shared wireless environment, we propose to fuse sub-6 GHz channel information and mmWave low-overhead measurement to predict the optimal mmWave beam in heterogeneous networks (HetNets) and reduce the overhead of both mmWave BS selection and beam training. Moreover, deep learning is adopted to extract the complex dependence between sub-6 GHz and mmWave channels for achieving high prediction accuracy. Specifically, we propose to leverage a few user equipment (UE)-specific high-quality mmWave wide beams predicted by the sub-6 GHz channel state information (CSI) as the mmWave low-overhead measurement. In order to adapt to different confidences of the mmWave wide beam prediction for diverse UE, the sum-probability criterion is proposed to flexibly adjust the number of measured wide beams. Besides, to fully fuse the diversified features extracted from the sub-6 GHz CSI and mmWave wide beams, the attention mechanism is further exploited to adaptively weight the features for improving the prediction accuracy. Simulation results show that our proposed scheme achieves higher beamforming gain while imposing smaller mmWave measurement overhead over the conventional deep learning based schemes.

In this paper, an adaptive millimeter-wave (mmWave) beam tracking scheme is proposed to flexibly adjust the angular range of beam tracking based on the user-specific speeds for reducing the tracking overhead, where deep learning is exploited to accurately extract the user movement features. Specifically, long short-term memory network is utilized to predict the possible optimal beams according to the received signals of previous beam tracking. Based on the predicted probabilities, the sum-probability criterion is proposed to track the subset of maximum-probability beams whose sum-probability is larger than the predefined threshold, where the beam with the highest received power is selected as the optimal one. Considering the limited number of received beam tracking signals, a two-stage training strategy is further proposed to stabilize the model optimization. Simulation results demonstrate that our proposed scheme could effectively reduce the overhead of beam tracking in guarantee of high beamforming gains, compared with the conventional schemes.

Huge overhead of beam training poses a significant challenge to mmWave communications. To address this issue, beam tracking has been widely investigated whereas existing methods are hard to handle serious multipath interference and non-stationary scenarios. Inspired by the spatial similarity between low-frequency and mmWave channels in non-standalone architectures, this paper proposes to utilize prior low-frequency information to predict the optimal mmWave beam, where deep learning is adopted to enhance the prediction accuracy. Specifically, periodically estimated low-frequency channel state information (CSI) is applied to track the movement of user equipment, and timing offset indicator is proposed to indicate the instant of mmWave beam training relative to low-frequency CSI estimation. Meanwhile, long-short term memory networks based dedicated models are designed to implement the prediction. Simulation results show that our proposed scheme can achieve higher beamforming gain than the conventional methods while requiring little overhead of mmWave beam training.

Huge overhead of beam training imposes a significant challenge in millimeter-wave (mmWave) wireless communications. To address this issue, in this paper, we propose a wide beam based training approach to calibrate the narrow beam direction according to the channel power leakage. To handle the complex nonlinear properties of the channel power leakage, deep learning is utilized to predict the optimal narrow beam directly. Specifically, three deep learning assisted calibrated beam training schemes are proposed. The first scheme adopts convolution neural network to implement the prediction based on the instantaneous received signals of wide beam training. We also perform the additional narrow beam training based on the predicted probabilities for further beam direction calibrations. However, the first scheme only depends on one wide beam training, which lacks the robustness to noise. To tackle this problem, the second scheme adopts long-short term memory (LSTM) network for tracking the movement of users and calibrating the beam direction according to the received signals of prior beam training, in order to enhance the robustness to noise. To further reduce the overhead of wide beam training, our third scheme, an adaptive beam training strategy, selects partial wide beams to be trained based on the prior received signals. Two criteria, namely, optimal neighboring criterion and maximum probability criterion, are designed for the selection. Furthermore, to handle mobile scenarios, auxiliary LSTM is introduced to calibrate the directions of the selected wide beams more precisely. Simulation results demonstrate that our proposed schemes achieve significantly higher beamforming gain with smaller beam training overhead compared with the conventional and existing deep-learning based counterparts.

The low-frequency and mmWave links usually co-exist in the next generation wireless terminals, where the low-frequency link is always on and the mmWave link becomes active when high rate transmission is required. Since low-frequency and mmWave channels have spatial similarities, it is feasible to utilize low-frequency channel information to reduce the beam training overhead in mmWave communications. In this paper, we propose a deep learning assisted beam prediction scheme using out-of-band information extracted from low-frequency channel state information (CSI). To overcome the inaccuracy in estimating spatial features due to small number of antennas in low-frequency band, deep learning is introduced to extract robust channel features and increase the prediction accuracy. Moreover, dedicated pre-processing algorithm and network architecture are derived to improve the performance. Simulation results demonstrate that the proposed scheme is robust to various CSI matrix sizes and signal-to-noise ratio. By exploiting low-frequency CSI, it could successfully predict the optimal beam direction to facilitate the initial beam training in mmWave communications with over 94% accuracy in line-of-sight scenarios, which can reduce the overhead of beam training significantly.

Three-dimensional visible light positioning (3D-VLP) is capable of achieving superior locating accuracy in comparison with other existing positioning techniques, such as global positioning system (GPS) and Wi-Fi-based method, which draws much attention from the researchers. In this paper, a novel 3D-VLP scheme using regression neural network is proposed to provide accurate and real-time positioning service. In the proposed method, the angle of arrival (AOA) vectors corresponding to the light-emitting diodes (LEDs) are obtained by the image sensor of the receiver and then fed into a regression neural network, which directly outputs the positioning results. Simulations are carried out to validate the superiority of the proposed method. It’s observed that, in spite of the inevitable quantization error in the positioning process, the mean positioning error is still as accurate as 1.1 cm. In addition, the proposed positioning method is more robust to camera’s height, and takes only 0.27ms to calculate the position, which could be used for real-time locating.