Deployment of a cloud pipeline for real-time visual inspection using fast streaming high-definition images

We investigate the challenges of building an end-to-end cloud pipeline for real-time intelligent visual inspection system for use in automotive manufacturing. Current methods of visual detection in automotive assembly are highly labor intensive, and thus prone to errors. An automated process is sought that can operate within the real-time constraints of the assembly line and can reduce errors. Components of the cloud pipeline include capture of a large set of high-definition images from a camera setup at the assembly location, transfer and storage of the images as needed, execution of object detection, and notification to a human operator when a fault is detected. The end-to-end execution must complete within a fixed time frame before the next car arrives in the assembly line. In this article, we report the design, development, and experimental evaluation of the tradeoffs of performance, accuracy, and scalability for a cloud system.     

Effects of driver age and experience in abrupt-onset hazards

Novice drivers and older drivers are found to have the highest crash risk among all drivers and this has motivated many research studies into various aspects of novice and older drivers. Although age-related declines were expected, studies did not find older drivers to respond slower to hazards. This study examined the hazard detection and response latencies of 14 young novice drivers, 14 young experienced drivers, and 12 older experienced drivers, to abrupt-onset hazards. Older drivers were found to take longer times before fixating on an abrupt-onset road hazard but appeared to have insignificantly faster reaction times after the initial fixation. Hence, the overall response latency did not suggest any age effects. Older drivers also scanned the roadway less as compared to their younger counterparts. No effects of experience were found. The findings provided insight on age-related declines in hazard detection whose effects have been masked by other components of hazard response.     

Varna-based optimization:a novel method for capacitated controller placement problem in SDN

Recently,software defined networking(SDN)is a promising paradigm shift that decouples the control plane from the data plane.It can centrally monitor and control the network through softwarization,i.e.,controller.Multiple controllers are a necessity of current SDN based WAN.Placing multiple controllers in an optimum way is known as controller placement problem(CPP).Earlier,solutions of CPP only concentrated on propagation latency but overlooked the capacity of controllers and the dynamic load on switches,which is a significant factor in real networks.In this paper,we develop a novel optimization algorithm named varna-based optimization(VBO)and use it to solve CPP.To the best of our knowledge,this is the first attempt to minimize the total average latency of SDN along with the implementation of TLBO and Jaya algorithms to solve CPP for all twelve possible scenarios.Our experimental results show that TLBO outperforms PSO,and VBO outperforms TLBO and Jaya algorithms in all scenarios for all topologies.     

Particle filter with one-step randomly delayed measurements and unknown latency probability

In this paper, a new particle filter is proposed to solve the nonlinear and non-Gaussian filtering problem when measurements are randomly delayed by one sampling time and the latency probability of the delay is unknown. In the proposed method, particles and their weights are updated in Bayesian filtering framework by considering the randomly delayed measurement model, and the latency probability is identified by maximum likelihood criterion. The superior performance of the proposed particle filter as compared with existing methods and the effectiveness of the proposed identification method of latency probability are both illustrated in two numerical examples concerning univariate non-stationary growth model and bearing only tracking.     

An improved technology for eliminating nondeterministic latency in the L1 trigger system

Gamma Ray Array Detector(GRAD) is one of external target facility subsystems in the Cooling Storage Ring on the Heavy Ion Research Facility at Lanzhou(HIRFL-CSR).The trigger subsystem of GRAD is required to make a fast L1 trigger decision with a fixed latency for the data acquisition.Because the hit signals from the detector are asynchronous with the local clock of the trigger system,a nondeterministic latency(the value changes between zero and one clock period) is generated when the synchronous receivers of the conventional trigger system process the hit signals.In this paper,an improved trigger logic based on a field-programmable gate array is developed,and comprised of zero-delay broadening circuits as receivers and an improved adding circuit designed for the new receivers.Software simulation and experimental measurement have been conducted.Comparison with the conventional trigger logic,the improved trigger logic has the advantage of eliminating the nondeterministic latency and reducing the total processing latency.     

面向按序执行处理器的预执行机制设计空间探索

面向按序执行处理器开展预执行机制的设计空间探索,并对预执行机制的优化效果随Cache容量和访存延时的变化趋势进行了量化分析.实验结果表明,对于按序执行处理器,保存并复用预执行期间的有效结果和在预执行访存指令之间进行数据传递都能够有效地提升处理器性能,前者还能够有效地降低能耗开销.将两者相结合使用,在平均情况下将基础处理器的性能提升24.07%,而能耗仅增加4.93%.进一步发现,在Cache容量较大的情况下,预执行仍然能够带来较大幅度的性能提升.并且,随着访存延时的增加,预执行在提高按序执行处理器性能和能效性方面的优势都将更加显著.     

射频识别芯片设计中时钟树功耗的优化与实现

UHF RFID是一款超高频射频识别标签芯片,该芯片采用无源供电方式,对于无源标签而言,工作距离是一个非常重要的指标,这个工作距离与芯片灵敏度有关,而灵敏度又要求功耗要低,因此低功耗设计成为RFID芯片研发过程中的主要突破点。在RFID芯片中的功耗主要有模拟射频前端电路,存储器,数字逻辑三部分,而在数字逻辑电路中时钟树上的功耗会占逻辑功耗不小的部分。本文着重从降低数字逻辑时钟树功耗方面阐述了一款基于ISO18000-6Type C协议的UHF RFID标签基带处理器的的优化和实现。     

New Latency Model for Dynamic Frequency Scaling on Network-on-Chip

Modulating both the clock frequency and supply voltage of the network-on-chip （NoC） during runtime can reduce the power consumption and heat flux, but will lead to the increase of the latency of NoC. It is necessary to find a tradeoff between power consumption and communication latency. So we propose an analytical latency model which can show us the relationship of them. The proposed model to analyze latency is based on the M/G/1 queuing model, which is suitable for dynamic frequency scaling. The experiment results show that the accuracy of this model is more than 90%.     

一种基于UDP电子政务消息传输的新技术

为了提高小消息、大并发、高吞吐量、高实时消息传输的高效性和可靠性,提出了一种基于UDP（User Datagram Protocol,用户数据报协议）的电子政务消息传输的新技术.该技术采用了一种消息刚好满足使用（Just meet）的思想,通过可靠UDP技术、低延迟处理协议等信息处理技术,尽量剔除数据传输中没有用的多余信息,来保证信息传输处理的低延迟和可靠性.应用结果表明,该技术提高了消息传输效率和可靠性,消息的处理能力可以高达每秒百万（条）级,处理延迟可控制在50ms左右.     

无线接入网关终端保持接入设计与实现

分析了WiMax网络架构标准中现有的终端接入方案的缺陷,提出并实现了一种在终端保持的接入方案,即终端出现异常并在短时间内重新接入时,接入网关保留其断开连接前的相关信息,保留网络层连接,仅需与终端重建链路层连接,从而大大减少了接入时延,保证了上层业务的连续性。