Search

In federated learning (FL), devices contribute to the global training by uploading their local model updates via wireless channels. Due to limited computation and communication resources, device scheduling is crucial to the convergence rate of FL. In this paper, we propose a joint device scheduling and resource allocation policy to maximize the model accuracy within a given total training time budget for latency constrained wireless FL. A lower bound on the reciprocal of the training performance loss, in terms of the number of training rounds and the number of scheduled devices per round, is derived. Based on the bound, the accuracy maximization problem is solved by decoupling it into two sub-problems. First, given the scheduled devices, the optimal bandwidth allocation suggests allocating more bandwidth to the devices with worse channel conditions or weaker computation capabilities. Then, a greedy device scheduling algorithm is introduced, which in each step selects the device consuming the least updating time obtained by the optimal bandwidth allocation, until the lower bound begins to increase, meaning that scheduling more devices will degrade the model accuracy. Experiments show that the proposed policy outperforms state-of-the-art scheduling policies under extensive settings of data distributions and cell radius., Comment: submitted to IEEE Trans. Wireless Communications

In this paper, we present a user-complaint prediction system for mobile access networks based on network monitoring data. By applying machine-learning models, the proposed system can relate user complaints to network performance indicators, alarm reports in a data-driven fashion, and predict the complaint events in a fine-grained spatial area within a specific time window. The proposed system harnesses several special designs to deal with the specialty in complaint prediction; complaint bursts are extracted using linear filtering and threshold detection to reduce the noisy fluctuation in raw complaint events. A fuzzy gridding method is also proposed to resolve the inaccuracy in verbally described complaint locations. Furthermore, we combine up-sampling with down-sampling to combat the severe skewness towards negative samples. The proposed system is evaluated using a real dataset collected from a major Chinese mobile operator, in which, events due to complaint bursts account approximately for only 0.3% of all recorded events. Results show that our system can detect 30% of complaint bursts 3 h ahead with more than 80% precision. This will achieve a corresponding proportion of quality of experience improvement if all predicted complaint events can be handled in advance through proper network maintenance.

In this paper, the graduation design course in the digital media art major of College of Sino-Korea Multimedia Design subordinate to Shanghai University of Engineering Science is taken as an example to expound four key factors of construction of the graduation design course, together with exploration in and reflection on the teaching and instruction modes of such course.

With the popularity of smart phones, the advent of 4G network era, and the establishment of the “Internet + education” ecosystem, the mobile terminal construction witnesses the high-speed development, especially the mobile education campus APP on teaching services centered on college students. This kind of mobile interactive application provides the teachers and students with a more convenient resource sharing and campus life of higher education. In this paper, the teaching service requirements of the School are analyzed, and the needs of the target users are explored. Based on the interactive process in conformity with the user's behaviors, the user interface information architecture of the APP is constructed. Moreover, the visual design is created in accordance with the APP design requirements of iOS system, in order to design the interface prototype of APP “i Sino-Korea” launched by Sino-Korean School of Multimedia Design of Shanghai University of Engineering Science.

Combining the Guangdong LNG station project, the gas station consumption instrument air equipment were analyzed, and the calculation of gas station instrument air system gas storage tank capacity, according to the Guangdong LNG station, the special circumstances, calculate separate switching valve gas storage jars. Through the analysis of the calculation process and result, relevant design wind instrument system is put forward.

In this paper a novel robust controller is investigated for unmanned helicopter. Based on signal compensation theory, robust controllers are designed to enhance both rotational and translational tracking performance, as well as robust performance over parameter uncertainty and external disturbance. The designed robust controllers are time invariant and easy to realize. The full controllers are implemented on onboard avionics of a small scale unmanned helicopter THeli260 developed by Tsinghua University. Real flight experiments are successfully carried out, which shows advantage of designed robust controller.

Recently, the flexible pressure sensor has attracted significant attentions as its high performance and wide applications, including wearable electronics, environmental monitoring device, and human-machine interfaces device and robotics. In this paper, silver nanowires (AgNWs) as the electrode materials were synthesized by polyol process, and the laboratory-made silver nanoparticles (AgNPs) as well as commercial carbon nanotubes (CNTs) were also chosen as the electrode materials in comparison. The flexible plastic, PI, PET, and PE films were chosen as the dielectric layer. The flexible pressure sensor was prepared with a simple, low-cost, and low energy-consuming method as sandwiching the dielectric layer between two opposing electrodes. The effects of the different electrode materials and the dielectric layer on the properties of the flexible pressure sensor were investigated. Results showed that the sensor with larger dielectric constant PI film and easy deformation AgNW/PDMS electrode had a greater sensitivity (1. 284 kPa−1), fast response time, and high stability. The applied testing showed that the prepared flexible pressure sensor could detect tiny pressure signals and has the potential to be used in the electronic skins filed.

As a novel cross-subject, digital media art is the combination of technology and art. The program design course is an important part of computer teaching course system. Considering the characteristics of the major and knowledge of the students, The Basic Course of Animation Programming is chosen as an introductory programming course. This paper proposes the reform of practicing on the content of course and teaching method and it also explores the measures combining teaching and ability training so as to raise the students' interest in learning and improve their programming knowledge and methods of quick understanding and application.

Due to the ground effect, it is difficult for quadrotor to take off or land precisely. This paper presents a new ground effect model of quadrotor based on the Cheeseman's model for helicopter, and a ground effect compensator is introduced into the control system. Meanwhile by combining the proportional-integral-derivative (PID) control method and the robust compensation technique, a robust altitude controller is designed to keep the height of quadrotor and land it at an accurate time. Experimental results demonstrate the effectiveness of our ground effect model and the control method.

We consider multiple base station (BS) cooperative transmission in downlink cellular networks to improve the spectral efficiency and the system capacity. Grouping BSs into clusters is a practical solution to realize BSs cooperation and reduce system complexity. However, it still suffers from inter-cluster interference, especially for the cluster-edge users. In this paper, clustering and scheduling are jointly considered to deal with the problem. The clusters are formed dynamically from users' point of view to minimize the inter-cluster interference, and are allowed to be overlapped. Accordingly, coordinated precoding scheme is designed to manage the intra-cluster interference. A greedy scheduling algorithm is proposed jointly with dynamic clustering. Simulations show that the proposed joint algorithm provides impressive average throughput gain over the non-joint ones, and the user fairness is improved significantly.