Building Robust and Optimized Cloud-Native Mobile Networks

在过去几年中，对高带宽网络连接的需求——由于对设备/机器/实体的实时或近实时远程控制、增强现实的实现以及提供高分辨率视觉/空间、触觉/触觉和其他感官数据以提供沉浸式体验的能力的综合需求而产生的极其严格的性能要求——已经显著增长。尽管这种快速增长为移动运营商带来了商机，但它也带来了迫切的网络容量/需求挑战，需要及时响应。投资额外的基础设施总是一个解决方案。然而，在这种情况下，这远不是一根魔法棒，因为资本投资的回报可能是最小的。不可否认，网络容量的提升在很大程度上取决于网络所依赖的基础设施。然而，实现特定硬件设置的真正潜力是一个具有更现实的改进空间和可见影响的目标。运营商一直在寻找一种创新的解决方案，以提高网络容量，超出预期的网络性能水平。此外，网络运营商正在应对其他多层次的挑战，这些挑战与日益增长的需求一样重要。不断增加的能源、财务和社会成本，加上缺乏专业人员来设计、实施和协调日益复杂的基于硬件的基础设施，是运营商的主要关注点。这些问题的范围不仅限于收入损失；连锁反应表现为上市时间滞后，以及供应链挑战和信息通信技术（ICT）行业创新的普遍障碍。网络功能虚拟化（NFV）、移动边缘计算（MEC）、人工智能/机器学习（AI/ML）和软件定义网络（SDN）技术获得了更大的动力，为降低资本和运营（CAPEX和OPEX）支出预算、提高灵活性、可扩展性和可移植性、易于开发和缩短应用程序上市时间提供了途径。NFV、AI/ML、SDN和MEC技术具有促进灵活的硬件和软件解耦的能力，以及实现更有效的资本投资和更高的资本回报的手段，它们有可能彻底改变电信行业。在本研究项目中，申请人的目标是研究和开发一个人工智能驱动的端到端连续编排架构框架，该框架扩展到极端边缘，并在考虑网络资源利用率和快速变化的计算资源市场的同时有效地支持未来应用的严格要求。

The demand for high-bandwidth network connectivity - with extremely stringent performance requirements stemming from the combined need for remote control of devices/machines/entities in real or near real-time, realization of augmented reality, and the ability to provide high-resolution visual/spatial, tactile/haptic, and other sensory data to offer an immersive experience - has been growing significantly over the past few years. As much as this rapid increase represents a business opportunity for mobile operators, it also creates pressing network capacity/requirements challenges in need of a timely response. Investing in additional infrastructure is always there as a solution. However, this is far from a magic wand in this situation since the return-on-capital investments can be minimal. It is undeniable that upgrading network capacity depends heavily on the infrastructure upon which the network relies. However, realizing the true potential of a specific hardware setup is an objective that has more realistic room for improvement and visible impact. Operators are in insistent search for an innovative game-changing solution to elevate the network capacity beyond the expected network performance levels. Furthermore, network operators are juggling other multi-layered challenges as critical as increasing demand. Increasing energy, financial and social costs coupled with the lack of personnel with expertise to design, implement, and orchestrate a progressively complex hardware-based infrastructure are the primary concerns of operators. The scope of these issues is not limited merely to revenue loss; ripple effects manifest through lags in time-to-market, as well as supply chain challenges and general hindrances to innovation within the Information and Communications Technologies (ICT) industries. Network Function Virtualization (NFV), Mobile Edge Computing (MEC), Artificial Intelligence/Machine Learning (AI/ML), and Software-Defined Networking (SDN) technologies gained more momentum offering a gateway to reduced capital and operating (CAPEX and OPEX) expenditure budgets, improved flexibility, scalability and portability, ease of development, and time-to-market applications releases. Empowered with the ability to facilitate flexible hardware and software decoupling and the means to achieve more effective capital investments with higher returns on capital, NFV, AI/ML, SDN and MEC technologies are/have the potential to revolutionize the telecommunications industry. In this research program, the applicant aims to investigate and develop a AI-driven end-to-end continuous orchestration architectural framework that extends to the extreme edge and efficiently support the stringent requirements of future applications while considering network resource utilization, and the rapidly changing computing resource market.