Download presentation

Presentation is loading. Please wait.

Published byCarley Swallows Modified over 9 years ago

1
LOGO Video Packet Selection and Scheduling for Multipath Streaming IEEE TRANSACTIONS ON MULTIMEDIA, VOL. 9, NO. 3, APRIL 2007 Dan Jurca, Student Member, IEEE, and Pascal Frossard, Senior Member, IEEE 指導老師：童曉儒 教授 學生：許益晨

2
Outline Introduction Multipath Video Streaming General Framework Streaming Model and Notations Packet Scheduling Analysis Unlimited Buffer Nodes Constrained Buffer Nodes Distortion Optimized Streaming Optimal Solution: Depth-First Branch & Bound(B&B) Heuristic Solution: Load-Balancing Algorithm(LBA) Simulation Results Conclusions

3
Introduction

4
This paper addresses the problem of choosing the best streaming policy for distortion optimal multipath video delivery, under network bandwidth and playback delay constraints.

5
Internet media streaming applications still suffer from limited and highly varying bandwidth, and from packet loss. Introduction Solution ： increase the streaming bandwidth by balancing the load over multiple (disjoint) network paths As solution 1 ： inter-dependent video packets to be transmitted (equivalently the adaptive coding of the video sequence) propose a detailed analysis of timing constraints imposed by delay sensitive streaming applications. propose a heuristic-based approach to the optimization problem, which leads to a polynomial time algorithm, based on load-balancing techniques.

6
Internet media streaming applications still suffer from limited and highly varying bandwidth, and from packet loss. Introduction Solution ： increase the streaming bandwidth by balancing the load over multiple (disjoint) network paths As solution 1 ： inter-dependent video packets to be transmitted (equivalently the adaptive coding of the video sequence) propose a detailed analysis of timing constraints imposed by delay sensitive streaming applications. propose a heuristic-based approach to the optimization problem, which leads to a polynomial time algorithm, based on load-balancing techniques.

7
Multipath Video Streaming

8
Multipath Video Streaming A. General Framework Multipath streaming scenario. The client accesses the streaming server simultaneously through two different paths, each one composed of two segments with intermediate buffers. The network channels between the server and the client are represented as variable bandwidth, lossless links. network segment i instantaneous rate r i ( t ) instantaneous latency d i ( t )

9
Multipath Video Streaming A. General Framework Multipath streaming scenario. The client accesses the streaming server simultaneously through two different paths, each one composed of two segments with intermediate buffers. The network channels between the server and the client are represented as variable bandwidth, lossless links. network segment i instantaneous rate r i ( t ) instantaneous latency d i ( t )

10
Represented in the last page: The rate r i ( t ) is the total bandwidth allocated to the streaming application on segment i at time instant t. denote the cumulative rate on segment i, up to time instant t, by Multipath Video Streaming B. Streaming Model and Notations We first assume that all segment rates and latencies along with intermediate buffer capacities are accurately predicted by the server at all time instants, possibly with feedback of the overlay nodes.

11
Multipath Video Streaming B. Streaming Model and Notations The video sequence is encoded into a bitstream using a scalable (layered) video encoder. The general rule stating ： Each network packet contains data relative to at most one video frame Encoded video frame can be fragmented into several network packets

12
Multipath Video Streaming B. Streaming Model and Notations Directed acyclic dependency graph representation for a typical MPEG layered-encoded video sequence (one network packet per layer, with IPBPB format). Each packet p n Packet size s n Decoding timestamp t d n Weight w n The successful decoding of one packet is contingent on the successful decoding of some other packets, called ancestors of p n.

13
By transmission policy π =(π 1, π 2,…, π N ) The policy π N used for packet p n consists in a couple a variables [q n, t s n ] that respectively represent the action q n chosen for packet p n, and its sending time t s n. Multipath Video Streaming B. Streaming Model and Notationst

14
Represented is equal to 1 if the packet arrives on time at the decoder, and if all its ancestors have been successfully decoded. That arrival time,t c n

15
Packet Scheduling Analysis

16
We consider buffering space in the network nodes and the client is not constrained. Packet Scheduling Analysis A. Unlimited Buffer Nodes the sending time t s n of each packet sent on path a

17
Packet Scheduling Analysis A. Unlimited Buffer Nodes p n enters the node buffer The arrival time of packet p n

18
Packet Scheduling Analysis A. Unlimited Buffer Nodes The queuing time b n = transmit bit present in the buffer The minimal playback delay D( π )

19
Lemma 1: Given that the streaming server sends the network packets in parallel on two paths, and that on each path the packets are sent sequentially, the playback delay D( π ) under the given policy vector π is a nondecreasing function of n. Proof: Packet Scheduling Analysis A. Unlimited Buffer Nodes

20
Lemma 2: Ω n is a nonincreasing function of the packet number. Packet Scheduling Analysis A. Unlimited Buffer Nodes

21
the buffering space in the intermediate nodes on each path is limited to and respectively. It tries to avoid buffer overflows by adapting the sending time of each packet to the buffer fullness. Note that it may no longer use the full available bandwidth, without risking loss of packets. Packet Scheduling Analysis B. Constrained Buffer Nodes

22
it verifies the inequality ： also define the maximum buffer occupancy during the whole streaming process

23
Distortion Optimized Streaming

24
Earliest Delivery Path First(EDPF) DISTORTION OPTIMIZED STREAMING

25
Earliest Delivery Path First(EDPF) DISTORTION OPTIMIZED STREAMING

26
Earliest Delivery Path First(EDPF) DISTORTION OPTIMIZED STREAMING

27
DISTORTION OPTIMIZED STREAMING A. Optimal Solution: Depth-First Branch & Bound (B&B) At each stage in the tree we can compute D( π ), the minimum playback delay and Ω n, the cumulative video quality measure, for a partial scheduling up to packet p n.

28
DISTORTION OPTIMIZED STREAMING B. Heuristic Solution: Load-Balancing Algorithm 1.If a packet can be scheduled on both network paths without interfering with the packets already scheduled, the algorithm will choose the path that offers the shortest arrival time for packet. 2.If packet can only be scheduled on one path, the algorithm will insert the packet on that path. 3.Otherwise packet cannot be scheduled on any of the two paths, without interfering with the already scheduled packets, and the algorithm will drop packet without transmitting it.

29
Simulation Results

30
Simulation Results A. Simulation Setup Video sequences are compressed with an MPEG4-FGS encoder, at 30 fps with various GOP structures. simulate network scenarios containing two and three disjoint paths between the server and the client.

31
Simulation Results B. Stored Streaming Scenarios

32
Simulation Results C. Streaming With Limited Look-Ahead Encoded video frame rate (cumulative) and decoded video frame rates (cumulative) in the case of infinite and constrained intermediate buffers

33
Simulation Results D. Streaming With Link Rate Estimation and Channel Losses

34
Simulation Results E. Complexity Considerations LBA performance versus complexity (100 frames, average aggregated bandwidth of 450 kbps).

35
Conclusions

36
polynomial time algorithms that still offer close to optimal solutions, in the case of stored videos, and real-time streaming. Simulation results in both scenarios prove that our proposed heuristic-based solution performs well in terms of final video quality, and is moreover suitable for the case of real-time streaming under strict delay constraints.

Similar presentations

© 2024 SlidePlayer.com Inc.

All rights reserved.

To make this website work, we log user data and share it with processors. To use this website, you must agree to our Privacy Policy, including cookie policy.

Ads by Google