基于智慧协同的异构多路流媒体传输控制机制研究

With the increasing availability of wireless communication technologies, and growing computing power and storage capacity of mobile devices, the content-rich multimedia streaming will be one of the most attractive services in the future mobile Internet. Concurrent multipath multimedia transfer has been recognized as a promising technology to increase the throughput performance and improve users’ quality of experience for multimedia streaming service in a heterogeneous wireless network environment. However, existing concurrent multipath transfer technologies lack smart cooperation and service-differentiating abilities, such limitations seriously degrade the application-level performance of multimedia delivery. By jointly considering the characteristics of multimedia streaming service-the most attractive service in the future mobile Internet, and MPTCP (Multipath TCP)-a promising transport protocol in the future mobile Internet, this project is proposed to launch a frontier research on smart and cooperative data transmission. The goals of this project are: (i) to design a smart and cooperative MPTCP-based transmission pattern in order to maximize the resource utilization and achieving load balancing, (ii) to design a quality demand-aware partial reliability-based MPTCP extension in order to meet the real-time demand of multimedia streaming, (iii) to design a energy-aware mechanism for MPTCP in order to reduce the energy consumption and maximize the battery lifetime of wireless terminals, (iv) to design a multi-QoS-aware bandwidth aggregation scheme for MPTCP in order to reduce the out-of-order multimedia data arrival and receive buffer blocking problem.

随着无线通信技术的快速发展，以及无线用户终端的计算和存储能力的增加，内容丰富的流媒体服务将成为未来移动互联网的核心应用。流媒体并行多路传输被认为是能够在异构无线网络环境下提高流媒体吞吐量和用户体验质量的最佳方案。然而，现有并行多路传输机制缺乏智慧协同以及区分业务等能力，成为制约实时流媒体传输的性能瓶颈。本项目以未来互联网核心应用流媒体服务为应用驱动，以未来互联网传输层核心协议MPTCP为协议基础，拟在智慧协同传输领域开展前沿研究。提出智慧协同并行多路传输模型，实现网络资源适配和节点负载均衡，并为后续研究工作奠定基础；提出基于质量需求感知的部分可靠传输协议，保证流媒体传输和播放的实时性；提出基于能量感知的绿色传输控制算法，降低并行多路传输总能耗及最大化无线终端续航时间；提出基于多域感知的多路带宽聚合策略，减少数据失序和避免接收缓冲区阻塞，提高流媒体吞吐量性能。

本项目以未来互联网核心应用流媒体服务为应用驱动，以未来互联网传输层核心协议MPTCP为协议基础，拟在智慧协同传输领域开展前沿研究。提出智慧协同并行多路传输模型，实现网络资源适配和节点负载均衡，并为后续研究工作奠定基础；提出基于质量需求感知的部分可靠传输协议，保证流媒体传输和播放的实时性；提出基于能量感知的绿色传输控制算法，降低并行多路传输总能耗及最大化无线终端续航时间；提出基于多域感知的多路带宽聚合策略，减少数据失序和避免接收缓冲 区阻塞，提高流媒体吞吐量性能。经过四年的研究，工作课题组圆满完成了预期研究任务，取得了一系列具有创新性以及前瞻性的研究成果。课题组共计发表（录用）高质量论文14篇，其中，在国际学术期刊《IEEE Access》、《EURASIP Journal on Wireless Communications and Networking》、《Visual Computer》、《Concurrency and Computation-Practice & Experience》、《KSII Transactions on Internet and Information Systems》、《China Communications》上发表SCI期刊论文8篇；在国际高水平学术会议IEEE ICPADS 2019、IEEE VCIP 2016等发表EI收录论文5篇；以及在CCF推荐中文科技期刊《小型微型计算机系统》1篇；申请发明专利3项、应用新型专利5项；培养硕士研究生2名。