The application of Vehicle-mounted self-organizing help control system is a promising direction of intelligent transportation. The efficiency and security have greatly improved by information sharing among vehicles th...
详细信息
In the past decade, internet of things (IoT) has been a focus of research. Security and privacy are the key issues for IoT applications, and still face some enormous challenges. In order to facilitate this emerging do...
详细信息
Wireless sensor networks (WSNs) are widely being used in many environments (e.g., disaster relief and target tracking), and WSNs localization is still a crucial research area because of the new localization requiremen...
详细信息
Cyber-Physical Systems (CPSs) integrate the virtual cyber world with the real physical world. In order to verify the new theories and methods, a novel CPSs simulation model for unmanned vehicle with Wireless Sensor Ne...
详细信息
Along with the development of network communication, digital media technology, multi-networks convergence and integrated business enlarge the ranges of mobile application, and data are produced, transported, stored an...
详细信息
Along with the development of network communication, digital media technology, multi-networks convergence and integrated business enlarge the ranges of mobile application, and data are produced, transported, stored and coped across heterogeneous networks, such as wired and wireless networks hybrid access. Data access is suffered from heterogeneous platforms and terminals with different ability, etc. It is a challenge how to ensure fair, efficient service. The article proposes hybrid access system model, analyzes data access performance, illustrates hybrid access optimization method to ensure fair and efficiency, and test the access delays and rates of the hybrid access system so as to provide reference to improve system design and optimization access performance.
Diffie-Hellman (DH) symmetric encryption plays an important role in the network security, and the security proof for many protocols relies on the DH assumption. The Logic of Local Sessions (LLS) is a practical securit...
详细信息
Implementation of controllers is always subject to timing uncertainties due to computing limitation, which will deteriorate control performance of a system. Integration from different perspectives is frequently adopte...
详细信息
Implementation of controllers is always subject to timing uncertainties due to computing limitation, which will deteriorate control performance of a system. Integration from different perspectives is frequently adopted to improve control performance in the literature. A streamlining codesign method is proposed which contains two critical parts: analysis model and the implementation architecture. Based on the former, the effect of timing attributes on control performance is estimated and key timing factors is chosen to manage, which makes the subsequent integration more efficient. In implementation architecture, the network bandwidth and CPU processing power is flexibly deployed. The method is validated with an application to CAN-based CNC systems. The sampling output jitter is the crucial timing factor concerned in the codesign. The scheduler designed regulates the sampling period to maximize the utilization of computing resources, and eventually improve the CNC control performance, which is shown by comparison of contour error in different sampling period.
In order to solve the computational complexity problem brought by high-order local features, the representation of high-order local features based on kernel function is proposed. At First, the feature space is mapped ...
详细信息
In order to use large numbers of unlabeled images effectively, this paper proposes an image classification method based on semi-supervised learning. The proposed method bridges a large amount of unlabeled images and l...
详细信息
In order to use large numbers of unlabeled images effectively, this paper proposes an image classification method based on semi-supervised learning. The proposed method bridges a large amount of unlabeled images and limited numbers of labeled images by exploiting the common topics. The classification accuracy is improved by using the must-link constraint and cannot-link constraint of labeled images. The experimental results on Caltech-101 demonstrate the feasibility and stability of proposed method. Furthermore, due to the present semi-supervised image classification methods lacking of incremental learning ability, we propose an incremental implementation of our method. Comparing with non-incremental learning model in literature, the incremental learning method can improve the computation efficiency of nearly 90%.
The frequent closed item set algorithm has shown its important advances by providing a minimal representation of frequent item sets without losing their support information. However, many applications showed that the ...
详细信息
暂无评论