Fault tolerance and real-time performance are two essential goals for directed acyclic graph (DAG) scheduling. However, the redundant tasks to compensate for faults can significantly prolong the completion time of a DAG, i.e., the makespan. In addition, the unpredictable runtime failure status, i.e., a fault may or may not occur during task execution, results in a huge difference between the offline schedule and the actual execution. Existing list scheduling methods can support fault-tolerant execution with all redundant tasks taken into account before runtime. However, such methods cannot effectively reduce the actual makespan as conditional execution of redundant tasks is not considered during scheduling. To address the above issues, this paper proposes a fault-tolerant DAG scheduling method with a runtime reconfiguration facility to optimize the actual makespan. First, the fault-aware makespan is optimized by a fine-grained offline scheduling method considering the worst-case scenario and the runtime flexibility. Then, the runtime reconfiguration mechanism safely moves the influential nodes ahead using the additional time interval to minimize the actual makespan. The experimental results indicate that the proposed method outperforms the state-of-the-art (SOTA) methods in terms of schedulability and actual makespan.

Real time operating system (RTOS) is an operating system which is designed to execute tasks in a guaranteed time. In this paper, we introduce SylixOS - a powerful and open source RTOS. As an industrial RTOS, SylixOS has been used in aerospace, military and many more embedded devices. We introduce the development path and some advanced features of SylixOS. In addition, SylixOS also supports the multiprocess. We discuss how SMP scheduling algorithms can be applied in SylixOS. Performance evaluations of SylixOS are presented to assess the real-time efficiency of SylixOS.

It requires an enormous amount of media server and network resources for on-demand delivery of video objects to a large number of clients with high start-up latency and packet loss ratio. An effective solution is to deploy a streaming proxy server close to client to cache media streaming data. In this paper, we propose an integral and scalable data sharing scheme, Optimal Prefix Caching and Data Sharing (OPC-DS), which combines prefix caching and interval caching in the proxy cache. In OPC-DS, the sizes of the appropriate prefix cache and interval cache are calculated according to the current request distribution. It can reduce the average start-up delay experienced by users as well as assure that the major of requests can share the data in the proxy buffer. So, utility of the proxy resources can be improved, and the resources consumption can be reduced significantly for the media server and back-bone network bandwidth by using it. Experimentally comparing with several existing methods, OPC-DS achieves significant performance improvement.

It requires an enormous amount of server and network bandwidth resources for providing real-time video streams to a large number of clients. And when using variable-bit-rate (VBR) encoded streams, frequent traffic bursts of bit rate have high variability for resource requirements. In this paper, we propose a smoothing of prefix caching and data sharing (S-PCDS) scheme based on the proxy caching, which can assist in smoothing operation for VBR encoded streams and reduce the start-up delay of users, through caching video prefix in the proxy. It can share the data in the buffer of proxy through interval caching which size is determined dynamically according to the current request distribution. With S-PCDS, the peak requirements of the resources can be reduced for VBR stream, the major of requests can share the data in the buffer, utility of the resource can be improved for the proxy, and the resource consumption can be minimized for the media server and back-bone network bandwidth. These are proved by comparing with several existing methods experimentally.

In this paper we firstly improve a virtual network buffer model (VB), based on which we propose a joint sender/receiver rate control algorithm for scalable video streaming. A transmission rate is determined by a rate control algorithm at the sender which employs the program clock reference (PCR) embedded in the video streams to work in a refined way. An over-boundary playback rate adjustment mechanism based on proportional-integral (PI) controller is performed at the receiver to maximize the visual quality of the displayed video according to the receiver buffer occupancy. Simulation results show that our proposed algorithm can reduce the overflow and underflow of the sender and receiver buffer, and achieve better video quality and quality smoothness than traditional rate control algorithms.

In order to obtain better video quality, media files are required to use variable-bit-rate (VBR) encoding. However, it produces traffic burst and unbalanced resource utilizations to translate VBR-encoded video. In this paper, we propose a novel bandwidth smoothing algorithm, buffer sharing and bandwidth smoothing of VBR streams (BSBS-VBR), which combines with prefetching and interval caching, allows video server to transmit a VBR-encoded stream at a fixed rate and makes users to share a disk stream. BSBS-VBR can also allocate and adjust buffer size dynamically according to the current request distribution and available resources. It can reduce the peak requirements of disk bandwidth and network bandwidth, improve utility of the resources, and serve more users by using this algorithm. These conclusions are proved by comparing with several existing methods experimentally.

Most of the existing buffer allocation and sharing schemes can not make full use of the system resources while being implemented independently. Here, based on analysis of the existing schemes, we propose a new algorithm, Balanced Buffer Sharing of Limited Resource (BBSLR), which allocates resources for each request according to the available cache and disk bandwidth, and adjusts buffer size according to the current request distribution and available resources dynamically. It does a good job of managing resources to maximize the number of simultaneous clients and enhance start-up delay. With BBSLR, the resources consumption will be balanced by rational allocation of the available resources, the average start-up delay will be reduced by caching the data at start and more clients will be served consequently. These conclusions are proved by comparing with several existing methods experimentally.

In this paper we present a video communication scheme that integrates rate control and buffer management at the source. A transmission rate is obtained via a rate control algorithm, which employs the Program Clock References (PCR) embedded in the video streams to regulate the transmission rate in a refined way and thus reduce the client buffer requirement. The server side also maintains multiple buffers to trade off random loss for controlled loss of visually less important data. We test our system with Standard Definition Television (SDTV) and High Definition Television (HDTV) traces, and find that the proposed scheme serves best for the transmission by slowing down the transmission rate without higher buffer requirement in both cases.

In this paper, we present a practical end-to-end video transmission system with refined rate allocation at the server and adaptive playback adjustment at the client that enables High Definition (HD) video streaming via bandwidth-constraint IP network. A transmission rate is determined by a rate control algorithm which employs the Program Clock References (PCR) embedded in the video streams to regulate the transmission rate in a refined way and thus reduce the client buffer requirement. An over-boundary playback adjustment mechanism based on Proportional-Integra (PI) controller is performed at the receiver to maximize the visual quality of the displayed video according to the receiver buffer occupancy. We test our system with Standard Definition Television (SDTV) and High Definition Television (HDTV) traces, and find that our proposed algorithm can reduce overflow and underflow of sender and receiver buffer, and achieve better video quality and quality smoothness than traditional rate control algorithms.