Cloud Assisted Two-Tier Wireless Networks

云辅助两层无线网络

• 批准号: RGPIN-2021-04298
• 负责人: Sousa, Elvino
• 金额: $ 2.4万
• 依托单位: University of Toronto
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2021
• 资助国家: 加拿大
• 起止时间: 2021-01-01 至 2022-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=741929
• 关键词: Cloud; Assisted; Two; Tier; Wireless

This project will address a new concept for public wireless cellular networks based on the Two-Tier concept that we have been proposing to the industry. This concept will target beyond 5G networks and we utilize data driven neural network and machine learning methodologies for network configuration, interference management, and resource allocation. The Two-Tier concept involves an architecture with the regular base stations which we refer to as primary nodes, secondary nodes that are installed close to the user terminals which act in a sense like relays but are quite different from the relays studied in the current literature, and the user terminals. The link from the primary nodes to the secondary nodes is referred to as the primary link and it is the focus of this project. The secondary nodes will have the capability for a larger number of antenna elements than is possible in a regular user terminal, and the link from the primary nodes to the secondary nodes will be well-behaved in terms of channel propagation and prediction. The focus of the proposal is to devise data driven techniques to optimize this link. There are two main focus cases: 1) The secondary nodes are stationary, and 2) The secondary nodes are placed in a moving platform such as a car, truck, bus, streetcar, or train. The project will focus on novel techniques to associate a channel with a time and location so that we can do much more efficient physical layer adaptation than in normal cellular schemes. The techniques will involve the use of cloud databases to contain information about the propagation environment in relation to the location of the secondary node, either in a fixed installation such as a home or office, or in a moving platform along a street. The techniques utilized in the project will involve neural networks and machine learning in order to do power level adaptation, antenna beam configurations, and co-operation between primary nodes. The ultimate goal in this research is to reduce the radiation footprint of transmitters significantly so as to enhance network capacity, increase physical layer security, decrease energy consumption, and minimize radiation.  In our research we will develop an architecture that complements the traditional approaches of small-cells and massive MIMO on traditional base stations. The one-tier small cell approach has the drawback of requiring a back-haul network to interconnect the small cells, and the massive MIMO approach used on conventional base stations has the drawback that when we have a large number of terminals such as in IoT applications even the massive MIMO approach will have limitations due to excessive overhead. Our approach will complement these two classical approaches and allow for custom network configuration implementations that are specific to a City and a network operator.

该项目将基于我们一直向业界提出的双层概念，为公共无线蜂窝网络提出一个新概念。这一概念将超越5G网络，我们利用数据驱动的神经网络和机器学习方法进行网络配置、干扰管理和资源分配。两层概念涉及一个架构，其中包括我们称之为主节点的常规基站，安装在用户终端附近的辅助节点，其作用在某种意义上类似于中继，但与当前文献中研究的中继有很大不同，以及用户终端。从主要节点到次要节点的链接称为主要链接，它是本项目的重点。辅助节点将比普通用户终端具有容纳更多天线单元的能力，并且从主节点到辅助节点的链路将在信道传播和预测方面表现良好。该建议的重点是设计数据驱动技术来优化这一链接。主要有两种情况：1)辅助节点是静止的，2)辅助节点放置在移动的平台上，如汽车、卡车、公共汽车、有轨电车或火车。该项目将专注于将信道与时间和地点联系起来的新技术，这样我们就可以比正常的蜂窝方案更有效地进行物理层适应。这些技术将涉及使用云数据库来包含与次要节点位置相关的传播环境信息，无论是在家庭或办公室等固定装置中，还是在沿街的移动平台中。项目中使用的技术将涉及神经网络和机器学习，以进行功率水平适应、天线波束配置和主节点之间的合作。本研究的最终目标是大幅减少发射机的辐射足迹，从而提高网络容量，提高物理层安全性，降低能耗，最大限度地减少辐射。在我们的研究中，我们将开发一种架构，以补充传统基站的小蜂窝和大规模MIMO的传统方法。单层小蜂窝方法的缺点是需要回程网络来互连小蜂窝，而在传统基站上使用的大规模MIMO方法的缺点是，当我们有大量终端时，例如在物联网应用中，即使是大规模MIMO方法也会由于过多的开销而受到限制。我们的方法将补充这两种经典方法，并允许特定于城市和网络运营商的自定义网络配置实现。