In this paper, we propose an unmanned aerial vehicle (UAV) and bird recognition scheme with signal processing and deep learning for integrated sensing and communications (ISAC) system. We first provide the basic scene of low-altitude targets monitoring, and formulate the motion equations and echo signals for UAVs and birds. Next, we extract the centralized micro-Doppler (cmD) spectrum and the high resolution range profile (HRRP) of the low-altitude target from the echo signals. Then we design a dual feature fusion enabled low-altitude target recognition network with convolutional neural network (CNN), which employs both the images of cmD spectrum and HRRP as inputs to jointly distinguish between UAV and bird. Meanwhile, we generate 237600 cmD and HRRP image samples to train, validate, and evaluate the designed low-altitude target recognition network. The proposed scheme is termed asAirGuard, whose effectiveness has been demonstrated by simulation results.

In this paper, we design the wideband hybrid-analog-digital (HAD) beamforming for integrated sensing and communication (ISAC) systems. Specifically, we incorporate the phase shifters (PSs) and true-time delay lines (TTDs) to combat the wideband beam squint effect, which are able to provide frequency-dependent phase shift in the analog beamforming stage. The fully-digital (FD) beamformers with guaranteed sensing and communication signal-to-interference-plus-noise ratios (SINRs) are first designed. Then, the HAD beamforming is formulated as a least squares (LS) problem to approximate the designed FD beamformers with constant-modulus constraints, whose main challenges are the complicated objective function and the non-convex constraints. To tackle these issues, we majorize the objective function to decouple the optimization variables. Then, the beamformer for PSs can be solved with a closed-form solution, whereas the beamformer for TTDs can be obtained by a simple grid-search. Finally, we adjust the PS and TTD beamformers by the Riemannian conjugate gradient method (RCGM) to improve the performance. Simulation results demonstrate the superior performance of the proposed algorithm over the conventional algorithm.

In this paper, we present a novel multi-hop link scheduling framework that utilizes the vision perception from cameras of the road-side unit (RSU) as well as cameras of the vehicle to support the large-capacity and reliable transmission of high-speed dynamic vehicle network. Specifically, we propose a vision based link state identification method to determine whether the communications links among RSU and different vehicles are blocked or connected. We firstly utilize the 3D detection technique to obtain the vehicle spatial distribution in surrounding environment. Then, the geometric calculation is adopted to accurately analyze the link states between RSU and different vehicles. Moreover, we design an environmental statistical information based low-complexity link scheduling method, and utilize the joint statistical distribution of the residual transmission distance and the residual multi-hop latency to optimize the total transmission latency. Simulation results show that the proposed vision based link state identification method significantly outperforms the exiting methods, and the proposed link scheduling method can approximately achieve the optimal performance as that from the exhaustive search method but with much less computation overhead.

In this paper, we propose an environment sensing-aided beam prediction model for smart factory that can be transferred from given environments to a new environment. In particular, we first design a pre-training model that predicts the optimal beam by sensing the present environmental information. When encountering a new environment, it generally requires collecting a large amount of new training data to retrain the model, whose cost severely impedes the application of the designed pre-training model. Therefore, we next design a transfer learning strategy that fine-tunes the pre-trained model by limited labeled data of the new environment. Simulation results show that when the pre-trained model is fine-tuned by 30% of labeled data from the new environment, the Top-10 beam prediction accuracy reaches 94%. Moreover, compared with the way to completely re-training the prediction model, the amount of training data and the time cost of the proposed transfer learning strategy reduce 70% and 75% respectively.

In low-altitude economy (LAE), the widespread use of various types of unmanned aerial vehicles (UAVs) could provide convenience and enhance efficiency. However, the existence of unauthorized or illegal UAVs would pose significant challenges to urban privacy and security. In this paper, we propose a networked integrated sensing and communications (ISAC) based UAV tracking and handover scheme towards LAE. We define avirtual sensing cell (VSC)where oneprimary base station (PBS)transmits sensing signals, while both the PBS and twosecondary base stations (SBS)receive echoes. Since the echoes contain the clutter of static environment, each base station (BS) would first filter out the clutter and then estimate the UAV’s horizontal angle, elevation angle, distance, and radial velocity with the multiple signal classification (MUSIC) algorithm. Next, we employ the centralized extended Kalman filter (EKF) to fuse the estimations from the three BSs and leverage the one-step prediction results of the EKF to distinguish and track multiple UAVs. When the UAV flies within the coverage of a VSC, we design aPBS handoverstrategy to select the optimal BS from three BSs as the new PBS in real-time. Moreover, we propose aVSC handoverstrategy to track the UAV continuously when it flies from one VSC to another. Simulation results demonstrate the effectiveness of the proposed scheme and provide valuable reference for UAV tracking and handover in LAE.

Integrated sensing and communications (ISAC) has been deemed as a key technology for the sixth generation (6G) wireless communications systems. In this paper, we explore the inherent clustered nature of wireless users and design a multi-user based environment reconstruction scheme. Specifically, we first select users based on the estimation precision of channel’s multipath, including the line-of-sight (LOS) and the non-line-of-sight (NLOS) paths, to enhance the accuracy of environment reconstruction. Then, we develop a fusion strategy that merges communications signalling with camera image to increase the accuracy and robustness of environment reconstruction. The simulation results demonstrate that the proposed algorithm can achieve a remarkable sensing accuracy of centimeter level, which is about 17 times better than the scheme without user selection. Meanwhile, the fusion of communications data and vision data leads to a threefold accuracy improvement over the image only method, especially under challenging weather conditions like raining and snowing.