物联网异构数据通信机制研究

Internet of Things (IoT) has huge application prospects and economic benefits, which has attracted much attention throughout the world. To provide data communication among large-scale heterogeneous users, it is urgently to manage the communication resources such as time/space/frequency efficiently and dynamically, and to incorporate time-varying traffic-loads and node’s battery charging process into the network design. In this project, heterogeneous data communication mechanism is studied to achieve a green network and meet the diversity application needs of IoT, jointly considering the large-scale and heterogeneous data transmission requirements, short packet communication characteristics, and heterogeneous energy harvesting process. Firstly, the heterogeneous data transmission scheduling mechanism is designed, which alleviates the energy bottleneck of nodes in the large-scale IoT network effectively, satisfies the large-scale and heterogeneous data transmission requirements, and alleviates the dependence of network performance on network state information. Secondly, the resource allocation is optimized, which coordinates the energy supply and data transmission, thereby reducing the constraints on network performance of heterogeneous data transmission requirements in the network, and further improves the communication performance. Thirdly, the impacts of system parameters such as energy harvesting capacity and user density on network communication performance are theoretically derived, which helps us to further understand IoT. This project will lay a scientific foundation for basic research in the field of IoT and promote the application of the IoT in various fields.

物联网具有巨大的应用前景和经济效益，是近年来国内外研究的热点。如何高效、动态地管理时/空/频等通信资源，并将时变的网络负载和节点的电池充电过程纳入网络设计中，以提供大规模异构用户之间的数据通信是目前亟待解决的问题。本项目拟结合物联网中大规模且异构的数据传输需求、短数据包通信和异构能量收集过程等特征，研究异构数据通信机制，以实现能量可持续物联网，并满足物联网的多样化应用需求：首先，设计异构数据传输调度机制，有效缓解物联网中节点的能量瓶颈，满足物联网中的数据通信需求，并减少网络性能对状态信息的依赖；其次，提出资源分配策略，协调能量供应和数据传输，降低异构数据传输需求对网络性能的制约，进一步提升通信性能；从理论上推导出能量收集能力和用户密度等系统参数对网络通信性能的影响以加深对物联网的认识。本项目的研究将为物联网领域的基础研究奠定科学基础，并推动物联网在各领域的应用。

物联网中具有大规模且异构的数据传输需求、短数据包通信和异构能量收集过程等特点，需要高效、动态地管理时/空/频等通信资源，并将时变的网络负载和节点电池充电过程的时间变化纳入网络设计中。本项目综合考虑物联网中的实际数据传输需求，主要研究：（1）提出了一种基于分层链的传输模型能缓解物联网节点的能量瓶颈，从而有效提高网络的吞吐量性能；并基于所提出的传输模型提出了三种基于链的传输协议，能有效满足物联网中大规模的数据传输需求，并减少网络性能对状态信息的依赖；（2）提出了一种基于联邦强化学习的数据传输方案，能有效提高传输延迟和减少丢包率，并确保安全和用户隐私；（3）提出了一种基于Lyapunov随机网络优化方法的自适应资源分配策略，该策略不需要任何先验的随机过程信息，可以实现渐进最优的网络性能，同时确保网络的稳定性；（4）揭示了能量收集能力等系统参数对网络通信性能的影响。