With mobility and flexibility in non-terrestrial environments, unmanned aerial vehicles (UAVs) have shown the potential for emerging applications, such as remote surveillance and mobile base stations. This work raises the emerging need to chain UAV relays in support ofon-demand connectivityto faraway users. However, the ideal isotropic antennas are too simplified to ensure connectivity among UAVs. Since antennas used in realistic communication systems including UAVs are non-isotropic, and the angle of radiations (AoR) between a pair of transmitter and receiver significantly influences received signal strengths (RSSs). Therefore, this work takes the AoR into account to deploy UAV relays in a 3D environment. Not only positions but also the headings of UAVs are modeled in the AoR-based deployment problem to minimize the number of UAVs in the relay chain, while the end-to-end RSSs are guaranteed. We prove the NP-hardness of the AoR-based deployment problem. Then, theAngle-of- Radiation-based Deployment (AIM) algorithmis proposed to solve the problem. The extensive simulation results show that the AIM algorithm reduces the number of UAVs used in the relay chain by 59.6% compared to the baselines when the heading of each UAV is well-adjusted. Also, the proposed AIM algorithm outperforms the baselines by 61.2% in terms of the average number of UAVs and demonstrates adaptability to various terrains.

Recently, unmanned aerial vehicles (UAVs) have shown potential for a variety of emerging applications with their high mobility and versatility. To conduct the tasks and receive control and command signals, the seamless connectivity for the UAVs is crucial. While the aerial connection can be provided by the existing ground cellular network, the down-tilted antennas of the ground base stations (GBSs) and the beam lobes generated by the antenna pattern lead to inevitable null areas in the airspace, making it challenging to ensure connectivity and safety in the sky. In this paper, we leverage an unmanned aerial relay (UAR) to extend the signal coverage of existing GBSs, enabling connectivity for aerial user equipments (AUEs) in aerial null areas. We design an efficient algorithm, Connectivity-Aware UAR (CARE-U), optimizing the UAR's trajectory to serve AUEs with pre-scheduled tasks. By simulations in both synthetic and realworld settings, the connected time ratio of the AUEs improves up to 66.3 % compared to that without the UAR. Besides, CARE-U outperforms the approach that modifies trajectories of AUEs, showing the great compatibility of CARE-U.

Fall is a major threat to stroke survivors with the problems of gait and balance disorders in the rehabilitation phase following severe consequences on quality of life and a heavy burden to their families. Many solutions have been proposed to assess fall risk for elders based on inertial sensor-based signals, however, there still exists a great challenge of transferring them from elderly populations to the stroke-survivors populations as gait disorder patterns are significant difference between elders and stroke survivors. In this study, we conduct a pilot study to collect inertial sensor-based signals from stroke survivors when they performed the timed up and go test, and build an automatic fall risk assessment model with the architecture of Siamese network, with a merit of mitigating the problem of small sample size. Specifically, the proposed automatic fall risk assessment model consists of two parallel convolutional neural networks, each of which is composed of three convolutional layers, two max-pooling layers, and three fully connected layers. To utilize the space relation among accelerator-based and gyroscope-based signals, two-dimensional discrete wavelet transform extracts image-like features, wavelet coefficients, from inertial sensor-based signals as the input. Experimental results show that the proposed fall risk assessment model has achieved a promising results, which outperform cutting-edge methods with a big margin. The proposed fall risk assessment model with low computational complexity and limited memory consuming can be deployed on an embedded system to provide fall risk assessment service for stroke survivors in point-of-care environments or community settings.