在异构无线网络中将能量最小化的多路径视频传输到移动设备资源-CSDN文库

研究论文

36 浏览量 2021-03-11 17:09:09 上传评论收藏 3.04MB PDF 举报

资源推荐

资源详情

资源评论

1160 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 34, NO. 5, MAY 2016

Energy-Minimized Multipath Video Transport to

Mobile Devices in Heterogeneous Wireless Networks

Jiyan Wu, Member, IEEE, Chau Yuen, Senior Member, IEEE, Bo Cheng, Member, IEEE,

Ming Wang, and Junliang Chen

Abstract—The technological evolutions in wireless communi-

cation systems prompt the bandwidth aggregation (e.g., Wi-Fi

and LTE radio interfaces) for concurrent video transmission to

hand-held devices. However, multipath video transport to the

battery-limited mobile terminals is confronted with challenging

technical problems: 1) high-quality real-time video streaming

is throughput-demanding and delay-sensitive; 2) mobile device

energy and video quality are not adequately considered in con-

ventional multipath protocols; and 3) wireless networks are error-

prone and bandwidth-limited. To enable the energy-efﬁcient and

quality-guaranteed live video streaming over heterogeneous wire-

less access networks, this paper proposes an energy-video aware

multipath transport protocol (EVIS). First, we present a math-

ematical framework to analyze the frame-level energy-quality

tradeoff for delay-constrained multihomed video communication

over multiple communication paths. Second, we develop schedul-

ing algorithms for prioritized frame scheduling and unequal loss

protection to achieve target video quality with minimum device

energy consumption. EVIS is able to effectively leverage video

frame priority and rateless Raptor coding to jointly optimize

energy efﬁciency and perceived quality. We conduct performance

evaluation through extensive emulations in Exata involving real-

time H.264 video streaming. Emulation results demonstrate that

EVIS advances the state-of-the-art with remarkable improve-

ments in energy conservation, video peak signal-to-noise ratio

(PSNR), end-to-end delay, and goodput.

Index Terms—Multipath transport protocol, energy efﬁciency,

real-time video, Raptor codes, heterogeneous wireless networks.

Manuscript received August 1, 2015; revised January 4, 2016; accepted

March 10, 2016. Date of publication April 6, 2016; date of current version

May 19, 2016. This work was supported by the National Natural Science

Foundation of China (NSFC) under Grant 61132001 and Grant 61550110244,

in part by National High-tech R&D Program of China (863 Program) under

Grant 2013AA102301, in part by the Program for New Century Excellent

Talents in University under Grant NCET-11-0592, in part by Project of New

Generation Broadband Wireless Network under Grant 2014ZX03006003, and

in part by National Research Foundation, Prime Minister’s Ofﬁce, Singapore

under its IDM Futures Funding Initiative. (Corresponding author: Bo Cheng.)

J. Wu is with the State Key Laboratory of Networking and Switching

Technology, Beijing University of Posts and Telecommunications, Beijing

100876, China, and also with the Department of Research and Development,

OmniVision Technologies Singapore Pte. Ltd., Singapore 068898 (e-mail:

wujiyan@126.com; jiyan.wu@ovt.com).

C. Yuen is with the Engineering Product Development Pillar, Singapore

University of Technology and Design, Singapore 487372 (e-mail: yuen-

chau@sutd.edu.sg).

B. Cheng, M. Wang, and J. Chen are with the State Key Laboratory

of Networking and Switching Technology, Beijing University of Posts and

Telecommunications, Beijing 100876, China (e-mail: chengbo@bupt.edu.cn;

wangming_bupt@bupt.edu.cn; chjl@bupt.edu.cn).

Color versions of one or more of the ﬁgures in this paper are available online

at http://ieeexplore.ieee.org.

Digital Object Identiﬁer 10.1109/JSAC.2016.2551483

I. INTRODUCTION

ECENT advancements in wireless infrastructures and

hand-held devices enable mobile users to receive rich

multimedia contents with ubiquitous access options, e.g., cel-

lular networks (UMTS, HSDPA, LTE), wireless local area

networks (802.11 family), and broadband wireless networks

(LTE, WiMAX). Supported by the unprecedented achievements

in communication t echnologies, mobile video streaming ser-

vices (e.g., Youtube, online gaming, live sports program, video

call, etc.) have undergone explosive growth during the past few

years. According to the latest market research of the Cisco com-

pany [1], video streaming accounts for 55% of the mobile data

trafﬁc over the Internet in 2014 and will exceed 72% by 2019. In

parallel, the global mobile data is predicted to increase 10-fold

in the next ﬁve years.

The exponential growth of mobile video trafﬁc signiﬁcantly

outpaces the network capacity of wireless platforms. Resource

restrictions of single wireless networks prompt the link (band-

width) integration of heterogeneous access medium for con-

current video transmission as shown in Figure 1 [2]–[7]. The

state-of-the-art mobile terminals (e.g., the Samsung S5 smart

phones [8] and Mushroom products [9]) are equipped with

different radio interfaces to concurrently receive data through

multiple wireless access networks. With the popularity of such

multihomed mobile terminals, the future wireless networking

is expected to incorporate heterogeneous access options for

providing high-quality mobile services.

A plethora of application and transport layer solutions have

been proposed to deliver multipath video streaming over het-

erogeneous wireless networks [2]–[7], [10]–[12], [56]. In par-

ticular, an effective transport-layer protocol is able to guarantee

the application-layer QoS (Quality-of-Service) and enhance the

network-level utilization. Furthermore, transport protocols only

require modiﬁcations at communication terminals (e.g., Linux

kernel implementation) and ease the burden of application

programs to enable multipath data transfer.

A. Problem Statement

Concurrent multipath video transport also requires higher

energy consumption of wireless devices. Contemporary mobile

terminals are usually powered by batteries with limited capac-

ities and the radio communication modules (e.g., using Wi-Fi,

3G, and LTE network interfaces) constitute the main portion

of energy consumption [13], [14]. As mobile video stream-

ing is energy-hungry and QoS-demanding application, it is

See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

WU et al.: ENERGY-MINIMIZED MULTIPATH VIDEO TRANSPORT TO MOBILE DEVICES 1161

Fig. 1. Concurrent video transmission with EVIS (Energy-Video aware

multI

path tranSport) protocol to multihomed mobile devices in heterogeneous

wireless n etworks (e.g.,Wi-Fi,4G,5G).

necessary to develop an effective transport protocol to provide

user-satisﬁed mobile streaming services [15], [16]. However,

the existing multipath protocols are inefﬁcient for delivering

energy-efﬁcient and quality-guaranteed video streaming over

heterogeneous wireless networks within stringent delay con-

straint. The challenging problems are summarized as follows.

(P1) QoS requirements. High-quality real-time video trans-

port is bandwidth-intensive and delay-sensitive. The

throughput demand for high-deﬁnition video distribution

mainly ranges from 2–6 Mbps. While the end-to-end

delay is constrained to be less than 150 ms to achieve

excellent real-time video quality [17].

(P2) Network limitation. Wireless networks are characterized

by scarce radio resources and time-varying channel status.

Recent studies [4], [18] reveal that the available band-

width for individual mobile users in 4G LTE networks

generally ranges from 1.5 to 2.5 Mbps.

(P3) Energy consumption. Mobile device energy is not con-

sidered in the multipath transport protocols recommended

by IETF (Internet Engineering Task Force), i.e., Multipath

TCP (MPTCP) [19] and Stream Control Transmission

Protocol (SCTP) [20]. Furthermore, MPTCP and SCTP

are ineffective to deliver energy-efﬁcient real-time video

due to the retransmission mechanism.

(P4) Video contents. Video trafﬁc is featured by complex

content parameters, e.g., frame priority and decoding

dependency. However, the existing multipath protocols

schedule the video data in a content-agnostic fashion,

which signiﬁcantly degrades the video quality and net-

work utilization.

B. Contributions

To jointly optimize the energy efﬁciency and perceived

quality of multihomed video streaming, this research presents

an energy-video aware multipath transport (EVIS) protocol.

Compared with the existing multipath protocols, we advance

the state-of-the-art by modeling and leveraging the frame-

level energy-quality tradeoff for concurrent video transport over

heterogeneous wireless networks. EVIS minimizes the device

energy consumption by taking advantage of priority-aware

frame scheduling and Raptor coding adaptation. The proposed

multipath protocol is able to effectively mitigate the packet

reordering and link heterogeneity to integrate heterogeneous

network resources. The detailed solution procedure will be pre-

sented in Section IV. Speciﬁcally, the contributions of this

paper can be summarized as follows.

(C1) Develop an analytical framework to model the frame-

level tradeoff between energy consumption and video dis-

tortion for real-time t rafﬁc over multiple wireless access

networks.

(C2) Propose a multipath transport protocol that effectively

integrates the following components:

• a prioritized frame scheduling algorithm leveraging

the frame ﬁltering and path selection to achieve target

quality with minimal energy consumption.

• an unequal loss protection scheme taking advantage of

the code rate and symbol size adaptation to minimize

sum of total video distortion.

(C3) Conduct extensive emulations in the Exata platform

involving real-time H.264 video streaming. Evaluation

results demonstrate that EVIS achieves remarkable

improvements than reference schemes in energy conser-

vation, video PSNR, end-to-end delay and goodput.

C. Paper Organization

The remainder of this paper is organized as follows.

Section II brieﬂy reviews the related work to this research. In

Section III, we present the protocol design and system model.

Section IV introduces the scheduling algorithms for prioritized

frame scheduling and unequal loss protection. The performance

evaluation is provided in Section V and Section VI gives the

concluding remarks.

II. R

ELATED WORK

The related work to this research can be classiﬁed into

three categories: 1) multipath transport protocols; 2) bandwidth

aggregation for multihomed video communication; 3) energy-

efﬁcient multimedia transmission to mobile devices.

A. Multipath Transport Protocols

1) SCTP Solutions: In reference [21], Wallace et al. review

the recent progresses and research issues in SCTP. Iyengar

et al. [22] study three negative effects in CMT, i.e., unnecessary

fast retransmissions, overly conservative congestion window

growth, and increased acknowledgment trafﬁc. In literature [4],

the authors propose a Distortion-Aware Concurrent Multipath

Transfer scheme (CMT-DA) to minimize the end-to-end dis-

tortion in mobile video delivery over heterogeneous wireless

networks. However, CMT-DA provides data protection with

retransmissions and is not appropriate for the real-time video

applications with stringent delay constraint. Xu et al. [5] pro-

pose a Quality-Aware Adaptive Concurrent Multipath Transfer

(CMT-QA) scheme that includes the components of data dis-

tribution, path quality estimation and optimal retransmission.

Speciﬁcally, the path quality is estimated with the sending

buffer size and transmission delay.

1162 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 34, NO. 5, MAY 2016

TABLE I

IFFERENCES OF THE PROPOSED EVIS WITH OUR EARLIER WORKS [4], [12], [26], [27], [28]

2) MPTCP Solutions: The Fountain Code-based Multipath

TCP (FMTCP) proposed by Cui et al. [23] leverages the rate-

less fountain coding to overcome channel erasures and path

heterogeneity in multipath data transfer. However, the video

quality is not considered in [23] and this metric is signiﬁcantly

different from the network-level parameters (e.g., throughput,

delay, packet loss rate, etc.). Peng et al. [11] propose an energy-

efﬁcient MPTCP scheme that leverages the throughput-energy

tradeoff for path selection and congestion control. Chen et al.

[24] conduct a measurement study of MPTCP performance

over cellular and Wi-Fi networks to investigate the impact of

path diversity on application-level metrics.

Singh et al. [10] propose a multipath real-time transport

protocol (MPRTP) that extends RTP to multipath communi-

cation scenario. However, MPRTP does not provide reliable

data transfer against channels losses. In literature [25], the

authors propose a Multipath Loss Tolerant (MPLOT) protocol

that exploits multipath diversity in wireless networks based on

block erasure code.

B. Bandwidth Aggregation

Han et al. [2] design and implement a multipath transmission

system to enable stable live video streaming over heteroge-

neous wireless networks based on fountain code. This system

is designed to maximize the video encoding rate on the basis

of aggregate bandwidth, as well as overcoming the wireless

channel loss. In literature [12], the authors propose a sub-frame

level (SFL) scheduling approach, which splits large-size video

frames to optimize the delay performance of HD video stream-

ing over heterogeneous wireless networks. In [26], the authors

introduce a dynamic rate allocation algorithm into joint source-

channel coding (JSCC) to optimize the mobile video quality

over heterogeneous networks. Xing et al. [3] propose a real-

time adaptive algorithm for HTTP-based video streaming over

multiple access networks. In [27], the authors develop a Loss

Tolerant Bandwidth Aggregation (LTBA) scheme to spread the

FEC packets over multiple wireless networks to minimize the

channel distortion. In reference [28], a goodput-aware load dis-

tribution model is developed to optimize the real-time trafﬁc

performance over multipath networks. Table I summarizes the

main differences of the proposed EVIS with our earlier works

[4], [12], [26]–[28]. EVIS is different from the solutions [4],

[12], [26]–[28] in the protocol type, data allocation, energy and

priority-awareness.

C. Energy-Efﬁcient Multimedia Transmission to Mobiles

The general review of recent studies in this ﬁeld can be

referred to [14]. Hoque et al. [13] propose a novel download

scheduling algorithm based on crowd-sourced video viewing

statistics to reduce the aggressive prefetching and tail energy

consumption. Hu et al. [29] propose an ofﬂine method to esti-

mate minimum power consumption, and an online solution for

energy saving in user skip/quit scenarios. Bui et al. [6] pro-

pose the GreenBag that includes the load balancing, segment

management and energy-aware mode control to deliver mobile

video over heterogeneous wireless networks. Ismail et al. [15]

propose a content and energy aware packet scheduling approach

for video upload service in heterogeneous networks. In ref-

erence [16], an energy management sub-system is developed

for mobile terminals to sustain multihomed video streaming.

Pluntke et al. [30] design a multipath scheduler that uses

MPTCP to maximize energy savings based on energy mod-

els and accumulated communication history for different radio

interfaces.

In summary, the existing multipath transport protocols are

ineffective to deliver real-time video in an energy-efﬁcient and

reliable manner. To the best of our knowledge, the proposed

EVIS is the ﬁrst multipath transport protocol that exploits the

frame-level energy-quality tradeoff for real-time video trans-

mission to multihomed mobile terminals in heterogeneous

wireless networks.

III. P

ROTOC OL DESIGN AND SYSTEM MODEL

A. System Overview

Figure 2 presents the system diagram of the proposed EVIS

framework, which is a completely transport-layer protocol.The

goal of this protocol is to deliver quality-guaranteed real-time

video over heterogeneous wireless networks with minimum

energy consumption of multihomed mobile devices. The work-

ing components involved in the communication session of

EVIS are implemented at both sender and receiver sides. To

ensure the compatibility and portability of the proposed sys-

tem, EVIS employs UDP (User Datagram Protocol) to transmit

video data and TCP for control information exchange (e.g., con-

nection establishment and feedback path status information).

We assume the video to be transmitted is encoded with H.264

codec [31], which is commonly adopted for video compression

in industrial communities. The quality requirement d

min

and

delay constraint T are imposed by the real-time video appli-

cations, e.g., the delay constraint for a GoP (Group of Pictures)

with 15 frames should be 500 ms if encoding frame rate is 30

frames per second. This is because the playout duration of the

15 frames in a GoP is 500 ms. The delay constraint should be

less than 500 ms in order to prevent playback buffer starvation.

剩余18页未读，继续阅读

评论收藏

内容反馈

weixin_38620741

粉丝: 1
资源: 909

在异构无线网络中将能量最小化的多路径视频传输到移动设备

异构无线网络中用于移动视频流的失真感知并发多路径传输

在异构无线网络中通过多路径TCP流传输高质量的移动视频

异构无线网络上保证质量的实时视频的带宽有效多路径传输协议

网络游戏-异构无线网络中基于MPTCP的视频传输拥塞控制方法.zip

融合多路径传输技术的移动车联网传输机制研究.pdf

网络游戏-无线异构网络的数据传输方法、分流设备、基站设备.zip

面向智慧农业的异构无线网络研究.doc

基于自适应网络编码的异构无线链路并发传输控制方法研究.docx

无线视频传输相关论文

论文研究-异构无线网络中的带宽分配策略 .pdf

异构网络_ns2异构网络_异构网络基于NS2_网络选择_异构网络_异构

基于GAN的异构无线网络故障检测与诊断算法.docx

基于DA优化模糊神经网络的异构无线网络接入选择算法

网络游戏-异构无线网络的分流系统.zip

面向智慧农业的异构无线网络研究_.doc

异构无线网络融合的关键问题及发展研究.pdf

SCI 异构无线网络

网络收益最大化的异构无线网络呼叫接纳控制.pdf

基于FMIPv6的异构无线网络垂直切换算法.pdf

异构无线网络资源管理技术研究.pdf

网络游戏-基于异构无线网络的多路存取多单元分布式资源分配方法.zip

网络游戏-多出口异构无线网络分布式资源共享方法.zip

最新资源