Search

Low Earth orbit Satellite Networks (LSNs) have been advocated as a key infrastructure for truly global coverage in the forthcoming 6G. This paper presents our initial measurement results and observations on the end-to-end network characteristics of Starlink, arguably the largest LSN constellation to date. Our findings confirm that LSNs are a promising solution towards ubiquitous Internet coverage over the Earth; yet, we also find that the users of Starlink experience much more dynamics in throughput and latency than terrestrial network users, and even frequent outages. Its user experiences are heavily affected by environmental factors such as terrain, solar storms, rain, clouds, and temperature, so is the power consumption. We further analyze Starlink's current bent-pipe relay strategy and its limits, particularly for cross-ocean routes. We have also explored its mobility and portability potentials, and extended our experiments from urban cities to wild remote areas that are facing distinct practical and cultural challenges., 12 pages, 20 figures, to be published in IEEE INFOCOM 2023

Strigolactones (SLs) are a class of phytohormones playing diverse roles in plant growth and development, yet the limited access to SLs is largely impeding SL-based foundational investigations and applications. Here, we developed Escherichia coli–Saccharomyces cerevisiae consortia to establish a microbial biosynthetic platform for the synthesis of various SLs, including carlactone, carlactonoic acid, 5-deoxystrigol (5DS; 6.65 ± 1.71 μg/liter), 4-deoxyorobanchol (3.46 ± 0.28 μg/liter), and orobanchol (OB; 19.36 ± 5.20 μg/liter). The SL-producing platform enabled us to conduct functional identification of CYP722Cs from various plants as either OB or 5DS synthase. It also allowed us to quantitatively compare known variants of plant SL biosynthetic enzymes in the microbial system. The titer of 5DS was further enhanced through pathway engineering to 47.3 μg/liter. This work provides a unique platform for investigating SL biosynthesis and evolution and lays the foundation for developing SL microbial production process.

Federated learning (FL) has recently emerged as a promising distributed machine learning (ML) paradigm. Practical needs of the "right to be forgotten" and countering data poisoning attacks call for efficient techniques that can remove, or unlearn, specific training data from the trained FL model. Existing unlearning techniques in the context of ML, however, are no longer in effect for FL, mainly due to the inherent distinction in the way how FL and ML learn from data. Therefore, how to enable efficient data removal from FL models remains largely under-explored. In this paper, we take the first step to fill this gap by presenting FedEraser, the first federated unlearning methodology that can eliminate the influence of a federated client's data on the global FL model while significantly reducing the time used for constructing the unlearned FL model.The basic idea of FedEraser is to trade the central server's storage for unlearned model's construction time, where FedEraser reconstructs the unlearned model by leveraging the historical parameter updates of federated clients that have been retained at the central server during the training process of FL. A novel calibration method is further developed to calibrate the retained updates, which are further used to promptly construct the unlearned model, yielding a significant speed-up to the reconstruction of the unlearned model while maintaining the model efficacy. Experiments on four realistic datasets demonstrate the effectiveness of FedEraser, with an expected speed-up of $4\times$ compared with retraining from the scratch. We envision our work as an early step in FL towards compliance with legal and ethical criteria in a fair and transparent manner., To be published in Proc.of 29th IEEE/ACM IWQoS, 2021

Recent years have witnessed the rapid growth of crowdsourced multimedia services, such as text-based Twitter, image-based Flickr, and video streaming-based Twitch and YouTube live events. Empowered by today's rich tools for multimedia generation/distribution, as well as the growing prevalence of high-speed network and smart devices, most of the multimedia contents are crowdsourced from amateur users, rather than from commercial and professional content providers, and can be easily accessed by other users in a timely manner. Since cloud computing offers reliable, elastic and cost-effective resource allocation, it has been adopted by many multimedia service providers as the underlying infrastructure. In this thesis, we formulate the cloud resource allocation in crowdsourced multimedia services as a standard network utility maximization (NUM) problem with coupled constraints, in which real-time user interaction is a fundamental issue, and develop distributed solutions based on dual composition. We further propose practical improvements for the content generation and big data processing of crowdsourced multimedia services in a cloud environment. Crowdsourced multimedia services also rely on convenient mobile Internet access, since mobile users occupy a large portion of both content generators and content consumers. The rich multimedia content, especially images and videos, put significant pressure on the infrastructure of state-of-the-art cellular networks. Device-to-device (D2D) communicationthat smartly explores local wireless resources has been suggested as a complement of great potential to support proximity-based applications. In this thesis, we jointly consider resource allocation and power control with heterogeneous quality of service (QoS) requirements from diverse multimedia applications.

The risk of individual citizen’s investment in education specifically refers to the benefit changeability of individual investment in education. Risks always go with uncertainty. By adopting the method of quantile regression estimation and taking the Chinese urban citizens as samples, the author makes a positivist study of the risk of Chinese urban citizens’ individual investment in education. The result indicates that the individual citizen’s investment in education often suffers obvious benefit changeability, i.e., the risk and benefit of investment in education shows a positive correlation. It is not realistic to reduce the inequality of salaries by means of investing more money for receiving higher-level education. Therefore, relevant policy makers should pay due attention to the establishment of the mechanism for sharing the risk of investment in education and the system of social insurance.

In recent years, the formulation "people-run economy" [minying jingji] has emerged with increasing frequency in all sorts of documents and literature in the field of economic studies and in the news media. In the beginning, people started to use the term, "people-run economy," to refer to economies that were under the nonpublic system of ownership; however, in reality, it has proved to be a very vague and ambiguous concept, and one that has caused a certain degree of confusion both in theory and practice. Its unscientific character has been manifested in the following areas: First, the formulation "people-run economy" fails to reflect accurately or precisely the nature and character of the variety of types and forms of economies included within the scope of the term with regard to the ownership systems to which they would belong, and, therefore, the use of the term merely serves to blur people's lines of vision.

Based on the mathematical model of electric power steering (EPS) system, an assist characteristic transfer function of EPS system is established in this paper. The steering wheel angle is taken as input signal and the steering column angle as output signal in the model. According to biological immune feedback mechanism, a control algorithm for steering column angle of EPS based on immune PID is proposed in this paper. Compared with the genetic PID algorithm, the simulation results indicate that the immune PID control algorithm can effectively improve response speed of EPS, shorten time up to the steady state value, enhance the efficiency of steering column angle tracking the steering wheel angle, and make EPS to have good performance of power steering.

The polysilicon sensor is one of the most prospective pressure sensors because of its good performance at relatively high temperature. In this paper we put forward a technique to compensate its negative sensitivity temperature drift that prohibits its successful utilization in wide range of temperature. The technique is inexpensive and easy to practice in applications.