Improved Peňa-Rodriguez portmanteau test

Several problems with the diagnostic check suggested by Peňa and Rodriguez [2002. A powerful portmanteau test of lack of fit for time series. J. Amer. Statist. Assoc. 97, 601-610.] are noted and an improved Monte-Carlo version of this test is suggested. It is shown that quite often the test statistic recommended by Peňa and Rodriguez [2002. A powerful portmanteau test of lack of fit for time series. J. Amer. Statist. Assoc. 97, 601-610.] may not exist and their asymptotic distribution of the test does not agree with the suggested gamma approximation very well if the number of lags used by the test is small. It is shown that the convergence of this test statistic to its asymptotic distribution may be quite slow when the series length is less than 1000 and so a Monte-Carlo test is recommended. Simulation experiments suggest the Monte-Carlo test is usually more powerful than the test given by Peňa and Rodriguez [2002. A powerful portmanteau test of lack of fit for time series. J. Amer. Statist. Assoc. 97, 601-610.] and often much more powerful than the Ljung-Box portmanteau test. Two illustrative examples of enhanced diagnostic checking with the Monte-Carlo test are given.     

CAR: a low latency video-on-demand broadcasting scheme for heterogeneous receivers

As different types of wireless networks are converging into an all-IP network, i.e., the Internet, it can be expected that in the near future video-on-demand (VoD) will be widely applied to many interesting services, and users can access these services using heterogeneous terminals via heterogeneous wired/wireless access networks. Many periodic broadcasting protocols have been proposed to reduce the implementation cost of VoD systems. However, most of the protocols assumed homogeneity for user terminals, while in practice, user terminals are usually quite different in their processing power, buffer space, and power. To address this problem, a few periodic broadcasting protocols providing the same video quality for all heterogeneous clients have been proposed recently. In this paper, we proposed a novel heterogeneous VoD broadcasting technique called Catch and Rest (CAR) to accommodate bandwidth heterogeneity without sacrificing user video quality. Then, we provide mathematic analysis to calculate the client bandwidth and buffer space requirements of CAR. Finally, we present our performance evaluation results for CAR. Our results show that under the same system resources (i.e., server and network bandwidth), CAR provides more uniform and acceptable service latency for all heterogeneous clients compared to previous works.     

A Unified Approach to Heterogeneous Video-on-Demand Broadcasting

Video-on-demand (VoD) is an essential technology for many multimedia applications. However, it remains challenging to efficiently deliver on-demand streams to a large number of receivers in a heterogeneous network environment. Although a number of VoD broadcasting protocols have been proposed for heterogeneous receivers, these protocols considers only single-bit-rate VoD services, which cannot satisfy the diverse quality-of-service requirements of heterogeneous receivers. In this paper, we propose a unified approach to heterogeneous VoD broadcasting, called unified heterogeneous broadcasting (UHB). As a unified approach, UHB is design to integrate different VoD broadcasting protocols and different scalable video coding schemes. The main merits of UHB are threefold. First, unlike conventional VoD broadcasting protocols, UHB supports multiple-bit-rate VoD services. Second, UHB is capable of adapting to network bandwidth fluctuations. Third, UHB enables each heterogeneous receiver to obtain the best visual quality by selecting the optimum system joining time. Our performance evaluation results indicate that UHB is superior to conventional heterogeneous VoD broadcasting protocols under various network conditions.     

Resource-Based Striping: An Efficient Striping Strategy for Video Servers Using Heterogeneous Disk-Subsystems

In building a large-scale video server, it is highly desirable to use heterogeneous disk-subsystems for the following reasons. First, existing disks may fail, especially in an environment with a large number of disks, enforcing the use of new disks. Second, for a scalable server, to cope with the increasing demand of customers, new disks may be needed to increase the server's storage capacity and throughput. With rapid advances in the performance of disks, the newly added disks generally have a higher data transfer rate and a larger storage capacity than the disks originally in the system. In this paper, we propose a novel striping scheme, termed as resource-based striping (RBS), for video servers built on heterogeneous disks. RBS combines the techniques of wide striping and narrow striping so that it can obtain the optimal stripe allocation and efficiently utilize both the I/O bandwidth and storage capacity of all disks. RBS is suitable for applications whose files are not updated frequently, such as course-on-demand and movie-on-demand. We examine the performance of RBS via simulation experiments. Our results show that RBS greatly outperforms the conventional striping schemes proposed for video servers with heterogeneous or homogeneous disks, in terms of the number of simultaneous streams supported and the number of files that can be stored.     

Video Super-resolution using Edge-based Optical Flow and Intensity Prediction

Full-image based motion prediction is widely used in video super-resolution (VSR) that results outstanding outputs with arbitrary scenes but costs huge time complexity. In this paper, we propose an edge-based motion and intensity prediction scheme to reduce the computation cost while maintain good enough quality simultaneously. The key point of reducing computation cost is to focus on extracted edges rather than the whole frame when finding motion vectors (optical flow) of the video sequence in accordance with human vision system (HVS). Bi-directional optical flow is usually adopted to increase the prediction accuracy but it also increase the computation time. Here we propose to obtain the backward flow from foregoing forward flow prediction which effectively save the heavy load. We perform a series of experiments and comparisons between existed VSR methods and our proposed edge-based method with different sequences and upscaling factors. The results reveal that our proposed scheme can successfully keep the super-resolved sequence quality and get about 4x speed up in computation time.     

Packet Permutation: A Robust Transmission Technique for Continuous Media Streaming Over the Internet

With the growing popularity of the Internet, there is an increasing demand to deliver continuous media (CM) streams over the Internet. However, packets may be damaged or lost during transmission over the current Internet. In particular, periodic network overloads often result in bursty packet losses, degrading the perceptual quality of CM streaming. In this paper, we focus on reducing the impact of this bursty loss behavior. We propose a novel robust end-to-end transmission scheme, referred to as packet permutation (PP), to deliver pre-compressed continuous media streams over the Internet. At the server side, PP permutes, prior to transmission, the normal packet delivery sequence of CM streams in a specific way. The packets are then re-permuted at the receiver side before they are presented to the application. In this way, the probability of losing a large number of packets within each CM frame can be significantly reduced. To validate the effectiveness of PP, a series of trace-driven simulations are conducted. Our results show that for a given quality of service (QoS) requirement of CM streaming, PP greatly reduces the overhead required by traditional error control schemes, such as forward error correction (FEC) and feedback/retransmission-based schemes.     

Constant time permutation: an efficient block allocation strategy for variable-bit-rate continuous media data

To provide high accessibility of continuous-media (CM) data, CM servers generally stripe data across multiple disks. Currently, the most widely used striping scheme for CM data is round-robin permutation (RRP). Unfortunately, when RRP is applied to variable-bit-rate (VBR) CM data, load imbalance across multiple disks occurs, thereby reducing overall system performance. In this paper, the performance of a VBR CM server with RRP is analyzed. In addition, we propose an efficient striping scheme called constant time permutation (CTP), which takes the VBR characteristic into account and obtains a more balanced load than RRP. Analytic models of both RRP and CTP are presented, and the models are verified via trace-driven simulations. Analysis and simulation results show that CTP can substantially increase the number of clients supported, though it might introduce a few seconds/minutes of initial delay. Received June 9, 1998 / Accepted January 21, 1999