低功耗移动自组网的邻居发现机制研究

Neighbor discovery is not only the foundation of the networking and routing for self-organizing networks, but also the routine functions of mobile self-organizing networks. In order to save energy effectively, the Low Power Consumption Mobile Self-Organizing Network normally works in the low duty cycle mode, where the scheduled sleep/wakeup cycles are exploited. However, the low duty cycle mode and the potential node mobility make the neighbor discovery a challenging issue. This project mainly focuses on the algorithms of probabilistic neighbor discovery, the selective reference neighbor discovery and the intra-group collaborative neighbor discovery. A set of efficient neighbor discovery mechanisms will be proposed, as well as a mobile campus information sharing system developed to demonstrate these mechanisms. Key innovations of this project include the mechanism of probabilistic neighbor discovery based on prime-set with low average discovery delay, the mechanism of selective reference neighbor discovery based on a combination of multiple historical information, the mechanism of intra-two-hop-group collaborative neighbor discovery for heterogeneous networks, and so on. Our team members, having published several papers in some top-level journals and conferences, are very experienced and knowledgeable in the neighbor discovery. Through this project, we expect our research results to be generally recognized by the international community in this area. The detailed result expectations include more than 12 high-level papers, 3-5 published in the prestigious international conferences such as SenSys, MobiHoc, and INFOCOM, 5 patents, and training of more than 12 graduate students.

邻居发现是自组网组网和路由的基础，也是移动自组网的常规工作。为节省能量，低功耗移动自组网常常采用"苏醒-睡眠"的低占空比工作模式，该模式及节点移动性使得邻居发现具有很大的挑战性。本项目主要研究概率性邻居发现、选择推荐邻居发现和组内协同邻居发现等多种算法，提出一套高效的邻居发现机制，以及研发基于智能手机的移动校园信息共享演示系统进行验证。主要创新点包括基于素数集合的低平均发现延时的概率性邻居发现机制、综合多种历史信息的高效的选择推荐邻居发现机制，以及针对异构网络的二跳组组内协同邻居发现机制等。团队成员在邻居发现研究方面具有良好的积累，多篇论文发表在国内外重要会议或期刊上，期望通过本项目的执行在国际上产生一定的影响力，预期成果包括发表高水平论文12篇以上，争取其中3-5篇发表在SenSys、MobiHoc、INFOCOM等重要国际会议上，申请国家发明专利5项，培养研究生12名以上。

邻居发现是自组网组网和路由的基础，也是移动自组网的常规工作。为节省能量，低功耗移动自组网常常采用“苏醒-睡眠”的低占空比工作模式，该模式及节点移动性使得邻居发现具有很大的挑战性。本项目主要研究概率性邻居发现、选择推荐邻居发现和组内协同邻居发现等多种算法，提出了一套高效的邻居发现机制，以及研发了基于智能手机的移动校园信息共享演示系统进行验证。主要创新点包括基于素数集合的低平均发现延时的概率性邻居发现机制、确定性邻居发现机制、综合多种历史信息的高效的选择推荐邻居发现机制。本项目取得了丰硕的研究成果：共发表学术论文25篇，其中SCI/EI 检索论文17篇，1篇会议论文被顶级学术会议INFOCOM 2018 接收，1篇TMC期刊论文正在投稿中；申请国家专利12项，包含已授权专利5项；人才培养方面，培养出已经毕业硕士生14人，参与本项目的在读硕士14人；邀请相关领域国际资深专家来校进行合作交流7人次；2项软件著作权正在申报中。