Machine-to-Machine Communications and Internet-of-Things for Future Generations of Wireless Networks

近年来，互联网技术经历了巨大的变化，并已成为一个重要的通信基础设施，目标是随时随地连接。鉴于人与人之间的通信主要是基于语音通信，传统的无线网络是针对以人为本的交通特性而构建和优化的。最近，一个完全不同的通信模式已经出现，包括“机器”在通信景观。任何机器生成的流量的交换被称为机器对机器（M2M）通信。因此，关于预计在不久的将来访问互联网的设备数量的猜测每天都在增加。这确实得到了物联网（IoT）框架的支持，该框架允许大量的“事物”在不需要人类交互的情况下相互生成和传递信息。通信机器的预计大规模引入必须针对需要广泛的要求和特性（诸如移动性支持、可靠性、所需的数据速率、功耗、硬件复杂性和设备成本）的应用进行规划和适应。M2M通信的其他规划和设计问题包括未来的网络架构、设备的大规模增长以及在未来几代无线网络中实现物联网概念的各种设备要求。** 在本提案中，我们调查了与M2M通信发展相关的重要问题，M2M通信是物联网和未来5G系统的关键推动因素。为了解决这些问题，拟议的计划分为三个互补的主题，所有目标都是实现未来的M2M通信。其中一个主题是关于机器学习理论在连接机器中的发展，以提高M2M网络的整体性能。第二个主题代表了向5G系统迈进了一步，我们考虑了超密集网络（UDN）的概念。在这方面，我们专注于与UDN及其对M2M应用程序的影响相关的问题。利用随机几何的理论，我们提供了这些问题的解决方案，如干扰管理和协调的人与人和M2M通信。鉴于M2M的主要挑战是由于设计方面，如动态频谱管理，异构性，功率效率的设计，和自适应配置能力，在这个建议的第三个主题涉及认知无线电的M2M的应用，其中提供的挑战和解决方案的认知M2M网络。 通过工业合作，该计划中开发的技术将被纳入构建原型以进行验证和进一步开发，为工业界提供广泛应用的高效快速实施。

Recently, the Internet technology has undergone enormous changes and has become an important communication infrastructure targeting anywhere, anytime connectivity. Given that human-to-human communication was mainly based on voice communication, traditional wireless networks are built and optimized for human-oriented traffic characteristics. Recently, an entirely different paradigm of communication has emerged with the inclusion of "machines" in the communications landscape. The exchange of any machine-generated traffic is known as machine-to-machine (M2M) communication. As a result, the speculations about the number of devices expected to access the Internet in the near future is increasing every day. This is indeed supported by the Internet-of-things (IoT) framework which allows a tremendous number of "Things" to generate and communicate information among each other without the need of human interaction. The projected massive introduction of communicating machines has to be planned for and accommodated with applications requiring a wide range of requirements and characteristics such as mobility support, reliability, required data rate, power consumption, hardware complexity, and device cost. Other planning and design issues for M2M communications include the future network architecture, the massive growth in devices, and the various device requirements that enable the concept of IoT in future generations of wireless networks. ******In this proposal, we investigate important problems related to the development of M2M communications as being a key enabler for the IoT and the future 5G systems. To address these problems, the proposed program is divided into three complementing themes, all targeting the realization of future M2M communications. One of these themes is concerned with the development of the machine learning theory in connected machines to improve the overall performance of M2M networks. The second theme represents one step closer to 5G systems where we consider the concept ultra-dense networks (UDNs). In this, we focus on problems associated with UDNs and their effect on M2M applications. Using the theory of stochastic geometry, we offer solutions to these problems such as interference management and coordination of human-to-human and M2M communications. Given that the major challenges of M2M are due to design aspects such as dynamic spectrum management, heterogeneity, power-efficient designs, and adaptive configuration capabilities, the third theme in this proposal deals with the application of cognitive radio to M2M where challenges and solutions of cognitive M2M network are provided. Through industrial collaborations, the techniques developed in this program will be incorporated to build prototypes for verification and further development to provide industry with efficient and fast implementations over the wide range of applications.