High-precision and low-latency wireless localization is a key objective for future networks. The majority of existing localization methods rely on multi-antenna arrays (MAAs) to estimate directions of arrival (DoAs), but the size and cost of MAAs limit their application in portable electronic devices. Virtual antenna arrays (VAAs), constructed from signals received at different positions by a moving single-antenna receiver, offer a promising alternative. However, VAA-based localization is challenged by the local oscillator frequency offset (LOFO) in transceivers, the irregular array geometry caused by the receiver’s movement, and the non-line-of-sight (NLoS) environments due to physical blockage. To address the above challenges, this paper proposes a VAA-based online localization approach. Specifically, we employ a manifold separation technique based on the Jacobi-Anger expansion to maintain a uniform linear array-like channel representation for the irregular VAA geometry. This enables us to transform the joint estimation of multi-path DoAs, times of arrival (ToAs), and LOFO into a modified two-dimensional atomic norm minimization problem, which is solved using an alternating convex search algorithm. Additionally, we introduce an unscented Kalman filter-based simultaneous localization and mapping algorithm for real-time position tracking in NLoS environments. Extensive simulations validate the effectiveness of our approaches, showing a 70.15% reduction in DoA estimation error and a 54.02% reduction in localization error compared to benchmark methods.

In this paper, we propose a Variational Bayesian Autoencoder (VBA)-based channel state information (CSI) compression and feedback scheme for massive multiple-input multiple-output (MIMO) systems. The proposed scheme incorporates the model-assisted knowledge of low-dimensional feedback features and the sparsity of channel to achieve enhanced compression efficiency. We also design a CsiVBA architecture that outputs distributions of the feedback features and the channel at the encoder and decoder, respectively, which facilitates a Bayesian training formulation exploiting the underlying channel sparsity. In addition, we also propose a low-complexity training scheme for new networks of different bit rates, significantly reducing the retraining cost for new compression requirements. Simulation results show that the proposed scheme achieves better rate-distortion trade-offs than the state-of-the-art solutions.

In this article, we propose an online training framework for deep neural network (DNN)-based mmWave massive multiple-input multiple-output (MIMO) channel estimation (CE) in Internet-of-Things (IoT) systems with nonlinear amplifier distortions. The DNN-based channel estimator is trained online in the IoT device based on real-time received pilot measurements from the base station (BS) without knowledge of the true channels, and can simultaneously generate CE in real time. To realize this, we first propose three axioms for a legitimate online loss function under known nonlinearity, based on which we develop a channel model-free online training algorithm with convergence analysis. For unknown nonlinearity, we propose a two-stage DNN structure with nonlinear modules, for which the DNN-based CE and nonlinear functions can be jointly trained online based on real-time received pilots. Simulation results show that the proposed solution achieves better CE accuracy than traditional compressive sensing (CS) algorithms while enjoying a much faster computational efficiency. In addition, the proposed method is robust to various nonlinear channel model mismatches and is able to track the change of the nonlinear channel model.

In this paper, we propose a two-stage deep neural network (DNN)-based channel estimator for massive multiple-input multiple-output (MIMO) systems with nonlinear amplifier distortions. The proposed two-stage structure is able to jointly learn a DNN-based channel estimator and the nonlinear transfer functions online based on real-time received pilot measurements, while generating channel estimation (CE) simultaneously. This is realized by a careful design of an online loss function that does not depend on ground truth channels while taking into account the unknown nonlinear distortions. Simulation shows that the proposed scheme outperforms the traditional compressive-sensing (CS) algorithms in terms of CE accuracy, and enjoys a much faster computational time during channel inferencing. The proposed scheme is also robust to various model mismatches and can adapt to the change of the underlying channel model.

In this paper, we propose an online deep learning (DL)-based channel estimation (CE) for massive multiple-input multiple-output (MIMO) systems with limited pilots, where the training stage can be implemented online based on real-time received pilot signals. This is realized by introducing a sparsifying loss function that is model-free and does not need ground truth labeled channel data. Simulation results show that the proposed online DL-based scheme achieves comparable channel estimation performance to traditional model-based compressive sensing (CS) algorithms while enjoying a much faster computation during the channel inferencing stage. In addition, the proposed scheme is also robust to various model mismatches and is able to track the change of the underlying propagation environment.