Distributed AI embedded cluster for real-time video analysis systems with edge computing

Herein, on the basis of a distributed AI cluster, a real-time video analysis system is proposed for edge computing. With ARM cluster server as the hardware platform, a distributed software platform is constructed. The system is characterized by flexible expansion, flexible deployment, data security, and network bandwidth efficiency, which makes it suited to edge computing scenarios. According to the measurement data, the system is effective in increasing the speed of AI calculation by over 20 times in comparison with the embedded single board and achieving the calculation effect that matches GPU. Therefore, it is considered suited to the application in heavy computing power such as real-time AI computing.

Enhancement of observability using Kubernetes operator

Observability is the ability for us to monitor the state of the system, which involves monitoring standard metrics like central processing unit (CPU) utilization, memory usage, and network bandwidth. The more we can understand the state of the system, the better we can improve the performance by recognizing unwanted behavior, improving the stability and reliability of the system. To achieve this, it is essential to build an automated monitoring system that is easy to use and efficient in its working. To do so, we have built a Kubernetes operator that automates the deployment and monitoring of applications and notifies unwanted behavior in real time. It also enables the visualization of the metrics generated by the application and allows standardizing these visualization dashboards for each type of application. Thus, it improves the system's productivity and vastly saves time and resources in deploying monitored applications, upgrading Kubernetes resources for each application deployed, and migration of applications.

Dynamic event-triggered fault detection filter design for dynamical systems under fading channels

This article investigates the dynamic event-triggered fault detection filter (FDF) design problem for linear continuous-time networked systems, considering the fading channels phenomenon and randomly occurring faults. A dynamic event-triggered mechanism (ETM) is introduced to reduce the network bandwidth occupation more efficiently by utilizing an internal variable which can enlarge the event-triggered intervals. Besides, the Zeno phenomenon is eliminated fundamentally by ensuring that the event-triggered intervals are positive lower bounded. After that, sufficient conditions are derived to guarantee the stochastic stability of the residual system with a desired [Formula: see text] performance and the co-design criterion of the FDF and the dynamic ETM is developed. Finally, an unmanned surface vehicle (USV) system is used to illustrate the applicability of the presented approach.

Polynomial fault detection filter design under adaptive event-triggered scheme via line-integral Lyapunov functions

This paper investigates the problem of polynomial fault detection filter design under an adaptive event-triggered scheme for continuous-time networked polynomial fuzzy model–based (PFMB) systems considering network transmission delays. The proposed adaptive polynomial event-triggered scheme is checked only at the sampling instant to eliminate the Zeno behavior as well as save the network bandwidth. With the consideration of the mismatched membership functions (MFs), the asynchronous problem between the physical plant and the polynomial fault detection filter (PFDF) is examined. A Lyapunov–Krasovskii (L-K) function is introduced to deal with the time delays caused by the network transmission and the zero-order holder (ZOH), and a proper line-integral Lyapunov function is also introduced to reduce the conservation of the design constraints, whose analytical procedure is rule-dependent. The design constraints are given in the form of sum of squares (SOS) to keep the PFMB fault detection system asymptotically stable with [Formula: see text] performance [Formula: see text]. Finally, an inverted pendulum example together with a numerical example is given to verify the effectiveness and superiority of the proposed scheme in terms of transfer rate and conservatism.

Research on optimization of real-time efficient storage algorithm in data information serialization

Background Along with the vigorous development of Internet technology, increasing the functions of the various types of equipment, network communication easy and diversity, at the same time, the amount of data is very huge, under the network bandwidth limitations, through long lead to a data need to be cut into more, one by one, transfer times, information delay problems. Results Aiming at the problems of poor data integrity, low efficiency and poor serialization efficiency of traditional data storage information, this article introduces Protobuf technology to study the serialization of data storage information. The serpentine gap method is used to complete the allocation interval of the sequence nodes, so that the working state and the resting state always maintain a dynamic balance. According to the first-level rules, the storage data of the completed target node is obtained, and the grammatical structure and the semantics of the target data are analyzed, Meanwhile corresponding correspondences are established, and the data storage information is serialized. In order to verify the effectiveness of Protobuf’s data storage information serialization method, a comparative experiment is designed. By using three methods of HDVM, Redis and Protobuf to serialize JSON data, the comparative analysis shows that HDVM has the longest processing time and Protobuf has the shortest processing time, and the data integrity is not affected. The simulation data shows that the Protobuf serialization method has short conversion time, high space utilization, and the Obvious advantages in correctness and integrity. It is vary suitable for serialization of JSON data in the case of limited bandwidth.

Review of light field technologies

Light fields are vector functions that map the geometry of light rays to the corresponding plenoptic attributes. They describe the holographic information of scenes by representing the amount of light flowing in every direction through every point in space. The physical concept of light fields was first proposed in 1936, and light fields are becoming increasingly important in the field of computer graphics, especially with the fast growth of computing capacity as well as network bandwidth. In this article, light field imaging is reviewed from the following aspects with an emphasis on the achievements of the past five years: (1) depth estimation, (2) content editing, (3) image quality, (4) scene reconstruction and view synthesis, and (5) industrial products because the technologies of lights fields also intersect with industrial applications. State-of-the-art research has focused on light field acquisition, manipulation, and display. In addition, the research has extended from the laboratory to industry. According to these achievements and challenges, in the near future, the applications of light fields could offer more portability, accessibility, compatibility, and ability to visualize the world.

Simulation analysis of a method to improve data-transmission performance of Nanshan 26m Radio Telescope based on Software-Defined Networks

Data Center of Xinjiang Astronomical Observatory (XAO-DC) commenced operating in 2015, and provides services including archiving, releasing and retrieving precious astronomical data collected by the Nanshan 26m Radio Telescope (NSRT) over the years, and realises the open sharing of astronomical observation data. The observation data from NSRT are transmitted to XAO-DC 100 km away through dedicated fiber for long-term storage. With the continuous increase of data, the static architecture of the current network cannot meet NSRT data-transmission requirements due to limited network bandwidth. To get high-speed data-transmission using the existing static network architecture, a method for reconstruction data-transmission network using Software-Defined Networks (SDN) is proposed. Benefit from the SDNʼs data and control plane separation, and open programmable, combined with the Mininet simulation platform for experiments, the TCP throughput (of single thread) was improved by ∼24.7%, the TCP throughput (of multi threads) was improved by ∼9.8%, ∼40.9%, ∼35.5% and ∼11.7%. Compared with the current network architecture, the Latency was reduced by ∼63.2%.

Minimizing the In-Cloud Bandwidth for On-Demand Reactive and Proactive Streaming Applications

Video streaming services are one of the most resource-consuming applications on the Internet. Thus, minimizing the consumed resources at runtime in general and the server/network bandwidth in particular are still challenging for researchers. Currently, most streaming techniques used on the Internet open one stream per client request, which makes the consumed bandwidth increases linearly. Hence, many broadcasting/streaming protocols have been proposed in the literature to minimize the streaming bandwidth. These protocols can be divided into two main categories, namely, reactive and proactive broadcasting protocols. While the first category is recommended for streaming unpopular videos, the second category is recommended for streaming popular videos. In this context, in this paper we propose an enhanced version of the reactive protocol Slotted Stream Tapping (SST) called Share All SST (SASST), which we prove to further reduce the streaming bandwidth with regard to SST. We also propose a new proactive protocol named the New Optimal Proactive Protocol (NOPP) based on an optimal scheduling of video segments on streaming-channel. SASST and NOPP are to be used in cloud and CDN (content delivery network) networks where the IP multicast or multicast HTTP on QUIC could be enabled, as their key principle is to allow the sharing of ongoing streams among clients requesting the same video content. Thus, clients and servers are often services running on virtual machines or in containers belonging to the same cloud or CDN infrastructure.

Machine Learning Algorithm for Delay Prediction in IoT and Tactile Internet

The next-generation cellular systems, including fifth-generation cellular systems (5G), are empowered with the recent advances in artificial intelligence (AI) and other recent paradigms. The internet of things (IoT) and the tactile internet are paradigms that can be empowered with AI solutions and integrated with 5G systems to deliver novel services that impact the future. Machine learning technologies (ML) can understand examples of nonlinearity from the environment and are suitable for network traffic prediction. Network traffic prediction is one of the most active research areas that integrates AI with information networks. Traffic prediction is an integral approach to ensure security, reliability, and quality of service (QoS) requirements. Nowadays, it can be used in various applications, such as network monitoring, resource management, congestion control, network bandwidth allocation, network intrusion detection, etc. This paper performs time series prediction for IoT and tactile internet delays, using the k-step-ahead prediction approach with nonlinear autoregressive with external input (NARX)-enabled recurrent neural network (RNN). The ML was trained with four different training functions: Bayesian regularization backpropagation (Trainbr), Levenberg–Marquardt backpropagation (Trainlm), conjugate gradient backpropagation with Fletcher–Reeves updates (Traincgf), and the resilient backpropagation algorithm (Trainrp). The accuracy of the predicted delay was measured using three functions based on ML: mean square error (MSE), root mean square error (RMSE), and mean absolute percentage error (MAPE).

GAF: A General Auction Framework for Secure Combinatorial Auctions

<p>Auctions are an economic mechanism for allocating goods to interested parties. There are many methods, each of which is an Auction Protocol. Some protocols are relatively simple such as English and Dutch auctions, but there are also more complicated auctions, for example combinatorial auctions which sell multiple goods at a time, and secure auctions which incorporate security solutions. Corresponding to the large number of protocols, there is a variety of purposes for which protocols are used. Each protocol has different properties and they differ between how applicable they are to a particular domain. In this thesis, the protocols explored are privacy preserving secure combinatorial auctions which are particularly well suited to our target domain of computational grid system resource allocation. In grid resource allocation systems, goods are best sold in sets as bidders value different sets of goods differently. For example, when purchasing CPU cycles, memory is also required but a bidder may additionally require network bandwidth. In untrusted distributed systems such as a publicly accessible grid, security properties are paramount. The type of secure combinatorial auction protocols explored in this thesis are privacy preserving protocols which hide the bid values of losing bidder’s bids. These protocols allow bidders to place bids without fear of private information being leaked. With the large number of permutations of different protocols and configurations, it is difficult to manage the idiosyncrasies of many different protocol implementations within an individual application. This thesis proposes a specification, design, and implementation for a General Auction Framework (GAF). GAF provides a consistent method of implementing different types of auction protocols from the standard English auction through to the more complicated combinatorial and secure auctions. The benefit of using GAF is the ability to easily leverage multiple protocols within a single application due to the consistent specification of protocol construction. The framework has be tested with three different protocols: the Secure Polynomial auction protocol, the Secure Homomorphic auction protocol and the Secure Garbled Circuits auction protocol. These three protocols and a statistics collecting application is a proof of concept for the framework and provides the beginning of an analysis designed at determining suitable protocol candidates for grid systems.</p>