Optimized Orchestration of Massive Next-Generation IoT Applications

5G通信最近制定的许多物联网(IoT)应用预计将遵循的新标准，即海量吞吐量、海量低延迟、海量传感、海量能效、海量安全保障和海量分形异构性，推动了对集成平台的新研究。在这个多租户、多服务和多速度的生态系统中创建、部署和协调应用程序非常具有挑战性，因为有各种不同的技术、不兼容的协议、不同的用户配置文件，特别是5G应用程序的毫秒延迟要求。此外，还需要新的绿色技术来抵消巨大需求带来的能源消耗爆炸性增长。 云计算、5G、软件定义网络(SDN)、网络功能虚拟化(NFV)、机器对机器(M2M)通信和大数据/机器学习这五项使能技术的融合，将为创建具有前所未有的技术能力、可扩展性、能效、安全性、灵活性和经济性的物联网应用平台提供机会。本研究对这些技术进行了全面研究，这些技术的结合使下一代应用能够同时感知环境和环境。这项研究的目标是发现和突破实时优化、动态资源分配和数据传输范式的潜力和边界，以满足大规模下一代物联网应用在质量和环境排放方面的新需求。 我们的研究计划将创建新的虚拟化、路由和调度算法、优化模型、认知决策和计算博弈理论，以协调大规模物联网应用的平台。特别是，这项研究将支持该领域最具挑战性的两个领域：i)通过远距离实时无缝地将服务迁移到边缘，并几乎打破光速以实现毫秒延迟；以及ii)以各种关系和连接的形式实时协调前所未有的数量(数十亿)的智能对象，以及大量的时间、空间、经济和环境要求。 这项研究的影响将是重大的，无论是在基本方法和新技术的进步方面。它通过新兴的理论模型增强知识，提供一致的实践经验，并通过可持续的用例实现。这项研究产生的平台、模型和范式将成为下一代物联网应用开发的主导力量，以释放我们网络社会的巨大创新，并保持经济增长和高质量的生活。

New standards recently set by 5G communications that many Internet of Things (IoT) applications are expected to follow, namely, massive throughput, massive low latency, massive sensing, massive energy efficiency, massive safety and security, and massive fractal heterogeneity, have driven new research on integrated platforms. Creating, deploying, and orchestrating applications across this multi-tenant, multi-service, and multi-speed ecosystem is very challenging because of the variety of diverse technologies, incompatible protocols, heterogeneous user profiles, and particularly the millisecond latency requirement of 5G applications. Also, new green technologies are required to offset the explosive increase of energy consumption of massive demands. The convergence of five enabling technologies - cloud computing, 5G, software-defined networking (SDN), network function virtualization (NFV), machine-to-machine (M2M) communication, and big data/machine learning will provide an opportunity to create IoT application platforms that offer unprecedented technical capabilities, scalability, energy efficiency, security, flexibility, and economics. The combination of these technologies, holistically investigated in this research, enables next-generation applications that are both context and environment aware. The goal of this research is to discover the potentials and push the boundaries of real-time optimization, dynamic resource allocation, and data transmission paradigms to meet new requirements of massive next-generation IoT applications, both in terms of quality and environmental emissions. Our research program will create new virtualization, routing and scheduling algorithms, optimization models, cognitive decisions, and computational game theory to orchestrate a platform for massive IoT applications at scale. In particular, this research will underpin two of the most challenging realms in the field: i) seamlessly migrating services to the edge in real-time through long distances, and virtually breaking the speed of light to achieve millisecond latency, and ii) real-time orchestration of unprecedented number of smart objects (order of billions) in vast forms of relationships and connections, as well as a multitude of temporal, spatial, economic, and environmental requirements. The impact of this research will be significant, both in terms of fundamental methodologies and advancement of novel technologies. It empowers knowledge enhancement through emerging theoretical models delivered with consistent practical experiences and enabled by sustainable use-cases. The platform, models, and paradigms resulting from this research will serve as the leading force in the development of next-generation IoT applications to unlock vast innovations in our networked-society and sustain economic growth and a high quality of life.