基于信息中心网络的无线局域网组播机制研究

随着智能终端的普及，移动互联网流量飞速发展，特别是视频流量爆炸式增长。组播是一种高效的单点对多点的传输技术，却很少在无线局域网中被使用。我们认为问题的根源在于现有的无线局域网组播的设计基于TCP/IP这种端到端的体系结构。为了解决组播动态性和组播易用性之间的矛盾，组播可靠性和组播数据传输效率之间的矛盾，组播可用性和终端能耗之间的矛盾这三个科学问题，本课题提出基于信息中心网络来设计无线局域网组播机制，拟在基于命名的动态组播管理机制、基于强化学习的组播速率控制机制、基于PIT状态的组播节能机制来对无线局域网组播关键技术进行创新和突破。我们将无线局域网组播应用于移动互联网流量占比最高的视频流传输，并设计能够充分发挥无线局域网组播优势的自适应视频流传输机制：基于无线访问点的视频流请求代理和基于SVC的分层组播，来验证基于信息中心网络的无线局域网组播关键技术的有效性。

With the popularity of smart mobile devices, mobile data traffic has grown explosively. Multicast is an efficient point-to-multipoint transmission technology, but it is rarely used in WLANs. We find that the root cause of this problem lies in that it is based on TCP/IP, an end-to-end networking architecture. To achieve dynamic, reliability and energy efficiency, we propose a WLAN multicast mechanism based on information-centric networking. we propose a naming based dynamic multicast management mechanism , a reinforcement learning based multicast rate adaptive algorithm, and a Pending Interest Table (PIT) state-based multicast power saving mechanism. Furthermore, We propose an adaptive video streaming mechanism using AP based video streaming proxy and SVC based layered multicast to evaluate the effectiveness of WLAN multicast design.