Wireless resource allocation is a critical component in modern communication systems, and deep neural networks (DNNs) have shown great promise in addressing this challenge. However, the conventional DNNs assume that testing data follows the same distribution as that of the training data, which is incongruent with the dynamic nature of real-world wireless environments. This paper introduces a new training algorithm designed specifically for dynamic wireless environments where channel distribution exhibits variability. This method helps DNNs adapt to new environments while preserving previously learned information. The proposed approach distinguishes itself by updating the DNN parameters in the null space of the low-rank covariance of previous data, which reduces memory needs and boosts training efficiency. Additionally, to counter the problem of DNNs hitting their model capacity during continuous adaptation, a selective forgetting mechanism is proposed. This mechanism allows DNNs to discard the unimportant knowledge over time, freeing up model capacity for more effective adaptation. The effectiveness of the algorithm is validated by integrating it with graph neural networks and multilayer perceptrons for weighted sum-rate maximization. Through a comprehensive evaluation that includes synthetic and ray-tracing-based datasets, superior performance is demonstrated compared to existing methods.

In continual learning, networks confront a trade-off between stability and plasticity when trained on a sequence of tasks. To bolster plasticity without sacrificing stability, we propose a novel training algorithm called LRFR. This approach optimizes network parameters in the null space of the past tasks’ feature representation matrix to guarantee the stability. Concurrently, we judiciously select only a subset of neurons in each layer of the network while training individual tasks to learn the past tasks’ feature representation matrix in low-rank. This increases the null space dimension when designing network parameters for subsequent tasks, thereby enhancing the plasticity. Using CIFAR-100 and TinyImageNet as benchmark datasets for continual learning, the proposed approach consistently outperforms state-of-the-art methods.

In mobile edge computing (MEC) systems, the wireless channel condition is a critical factor affecting both the communication power consumption and computation rate of the offloading tasks. This paper exploits the idea of cooperative transmission and employing reconfigurable intelligent surface (RIS) in MEC to improve the channel condition and maximize computation efficiency (CE). The resulting problem couples various wireless resources in both uplink and downlink, which calls for the joint design of the user association, receive/downlink beamforming vectors, transmit power of users, task partition strategies for local computing and offloading, and uplink/downlink phase shifts at the RIS. To tackle the challenges brought by the combinatorial optimization problem, the group sparsity structure of the beamforming vectors determined by user association is exploited. Furthermore, while the CE does not explicitly depend on the downlink phase shifts, instead of simply finding a feasible solution, we exploit the hidden relationship between them and convert this relationship into an explicit form for optimization. Then the resulting problem is solved via the alternating maximization framework, and the nonconvexity of each subproblem is handled individually. Simulation results show that cooperative transmission and RIS deployment can significantly improve the CE and demonstrate the importance of optimizing the downlink phase shifts with an explicit form.

In this paper, simultaneously transmitting and reflecting (STAR) reconfigurable intelligent surface (RIS) is investigated in the multi-user mobile edge computing (MEC) system to improve the computation rate. Compared with traditional RIS-aided MEC, STAR-RIS extends the service coverage from half-space to full-space and provides new flexibility for improving the computation rate for end users. However, the STAR-RIS-aided MEC system design is a challenging problem due to the non-smooth and non-convex binary amplitude coefficients with coupled phase shifters. To fill this gap, this paper formulates a computation rate maximization problem via the joint design of the STAR-RIS phase shifts, reflection and transmission amplitude coefficients, the receive beamforming vectors, and energy partition strategies for local computing and offloading. To tackle the discontinuity caused by binary variables, we propose an efficient smoothing-based method to decrease convergence error, in contrast to the conventional penalty-based method, which brings many undesired stationary points and local optima. Furthermore, a fast iterative algorithm is proposed to obtain a stationary point for the joint optimization problem, with each subproblem solved by a low-complexity algorithm, making the proposed design scalable to a massive number of users and STAR-RIS elements. Simulation results validate the strength of the proposed smoothing-based method and show that the proposed fast iterative algorithm achieves a higher computation rate than the conventional method while saving the computation time by at least an order of magnitude. Moreover, the resultant STAR-RIS-aided MEC system significantly improves the computation rate compared to other baseline schemes with conventional reflect-only/transmit-only RIS.

In this paper, we present a deep learning based trainable approximate message passing algorithm (TAMP) for signal detection in massive multiple-input multiple-output (MIMO) systems. The TAMP network consists of a preprocessing layer and a fixed number of detection layers, where the preprocessing layer is designed by using a standard fully connected layer, and the structure of each detection layer is derived by unfolding each iteration of the iterative GAMP algorithm. In addition, the proposed TAMP includes trainable parameters controlling prior mean and variance of minimum mean squared error (MMSE) denoiser. The parameters are trained by standard deep learning techniques. We evaluate the signal detection performance of the proposed TAMP under Rayleigh-fading and spatial correlated MIMO channels. Furthermore, we compare TAMP with existing state-of-the-art iterative message passing-based detection algorithms and deep learning based detection algorithms. Computer experiments show that TAMP is applicable under both Rayleigh-fading and spatial correlated channel in massive MIMO systems. Moreover, comparison results demonstrate that TAMP can achieve better detection accuracy and faster convergence.

This paper describes general framework of photo-realistic 3D visualization of urban building with digital map, Lidar point and close-range image in photogrammetry community. First, Vanish Point (VP) geometry deduced from classic photogrammetry co-linear equation is used to obtain camera Interior Of Parameter (IOP) and Exterior Of Parameter (EOP) using groups of mutually orthogonal space lines appeared in buildings; Next, "line photogrammetry" is introduced to precisely reconstruct building geometrical model. To this end, coarse building geometry model is generated from digital map and Lidar poin, and further refined with constrains from abundant image line features; After that, texture for each building facet is well reconstructed by applying photogrammetry differential rectification to eliminate texture distortion and detecting texture elements to remove occlusion as well; Finally, two experiments are presented by integrating digital map with ground-level still image and low-altitude aerial video image strips for photo-realistic 3D visualization of building respectively. Contributions are also concluded.