文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > A New MBAC algorithm for Video Streaming based on AutoRegressive Adaptive Filtering
侵权投诉

A New MBAC algorithm for Video Streaming based on AutoRegressive Adaptive Filtering

Abstract: A novel strategy of Measurement Based Admission Control (MBAC) for video distribution system has been developed, aiming at obtaining the needed Quality of Service (QoS), expressed in terms of probability for the video aggregation to exceed the ch

A New MBAC algorithm for Video Streaming based on AutoRegressive

Adaptive Filtering

Pietro Camarda, Cataldo Guaragnella, Domenico Striccoli

Politecnico di Bari – Dipartimento di Elettrotecnica ed Elettronica,

Via E. Orabona, 4 – 70125 Bari (Italy)

e-mail: {camarda, guaragnella, d.striccoli} @poliba.it

Abstract: A novel strategy of Measurement Based Admission Control (MBAC) for video distribution system has been developed, aiming at obtaining the needed Quality of Service (QoS), expressed in terms of probability for the video aggregation to exceed the channel capacity. Such an algorithm is based on an AutoRegressive (AR) adaptive filter taking into account slow and quick time bit rate variations of video stream aggregation. A simple admission control strategy is presented, able to shift from the server to the client the decision about the admission of a new video stream. Simulation results, based on real MPEG encoded videos justify the application.

Keywords: admission control, linear prediction, congestion control, multimedia streaming system

1. INTRODUCTION

In the modern Telecommunication world, the rapid growth of multimedia transmission like Video on Demand, Distance Learning, video conferences, video telephony, etc., has renewed the research interest in this area. Quality of Service (QoS) support is necessary to guarantee multimedia delivery with low frame losses, low packet delays and jitters.

It is well known that video applications usually present Variable Bit Rate (VBR) bandwidth requirements, like the case of video sources compressed with standards (MPEG, H.263, etc). As stated in [1], VBR video traffic exhibits self-similarity and Long Range Dependence (LRD) properties that make difficult an accurate statistical analysis. On the other side, a correct video data analysis is fruitful to efficiently plan network infrastructures and Admission Control schemes.

Measurement Based Admission Control (MBAC) algorithms perform admission control by measuring actual traffic load in the network link and exploiting past measurements to estimate actual bandwidth utilization [1][2][3].

In this work, we propose a new MBAC algorithm suitable for compressed video delivery with specified QoS guarantees. It establishes if a new video flow can be admitted into the network link, while respecting the QoS guarantees of the flows already running together with the new admitted flow QoS specifications. The QoS parameter taken into account is the probability for the video aggregation to exceed the channel capacity. The MBAC algorithm is based on an AutoRegressive (AR) adaptive filter taking into account slow time and quick time bit rate variations of stream aggregation, to efficiently perform video stream admission control and respect the QoS guarantees. Statistics of the stream aggregation are taken

into account through past filtered measurements of bit rate aggregation.

The algorithm proposed in this work can be fruitfully applied in a context in which an aggregate behaviour analysis is crucial for admission control, like the DiffServ architecture scenario [4], in which video streaming applications often require different QoS levels according to the particular application running into the network link [5]. It has to be noted that the main disadvantage of the DiffServ architecture is that it does not provide QoS guarantees to every flow, since it establishes the behaviour of flow aggregations. This problem becomes particularly relevant for multimedia applications, that suffer from service degradations. A correct Admission Control policy is thus essential to respect the QoS guarantees for different types of multimedia applications [6].

A variety of approaches have been studied to guarantee the correct QoS level to video flows in a DiffServ context. In [5] a strategy for bandwidth allocation is proposed, for different streaming service classes. It exploits past information about streaming bit rates (the QoS metric taken into account) as feedback for bandwidth estimation, to improve QoS differentiation robustness. In [6] a novel scheduling algorithm is proposed in a DiffServ scenario for multimedia applications, to guarantee the fairness of the admitted flows at the same time obtaining an efficient performance in terms of QoS parameters (drop rate and delays among data packets). In [7] a QoS mapping is proposed for multimedia applications running into a DiffServ architecture. A QoS mapping is proposed, that is an algorithm to make a correspondence between the various kinds of data packets and the DiffServ QoS classes. In [8] the QoS issue on DiffServ networks has been applied to MPEG video traffic, treating the video flows as Assured Service (AF) flows and managing I-frames, P-frames and

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email