SWIFT: SMALL: xNGRAN Navigating Spectral Utilization, LTE/WiFi Coexistence, and Cost Tradeoffs in Next Gen Radio Access Networks through Cross-Layer Design

移动的多媒体应用和物联网（IoT）的快速增长对无线网络基础设施提出了苛刻的要求，例如超低延迟、用户体验连续性和高可靠性。移动的设备由于其计算和通信能力的增加而成为当今访问互联网服务的主要媒介。然而，使用移动的设备来实现需要实时、现场数据收集和处理的应用仍然具有挑战性，这是由于（1）计算能力不足以及在各个移动的设备上聚合/全局数据的不可用性，以及（2）将数据卸载到远程计算资源以进行集中式计算所涉及的通信成本和响应时间。鉴于这些限制，移动的边缘计算（MEC）概念已经出现，其目的是将电信、IT和云计算联合起来，直接从网络边缘提供云服务。边缘计算允许在终端用户附近执行应用程序，这减少了端到端延迟和返回到集中式云计算环境的昂贵回程带宽消耗。下一代（NG）无线电接入网络（RAN）的灵活性允许在MEC中从“完全集中化”移动到“分布式方法”。所提出的用于5G（虚拟化）蜂窝网络的xl_NGRAN框架针对按需资源分配和网络内内容缓存做出优化的跨层决策，并在无线电资源、系统成本、LTE/WiFi技术共存和缓存服务之间进行权衡。通过基于云的框架，特别是通过NG-RAN虚拟化，包括物理基础设施和频谱在内的网络资源被抽象化，以提供支持各种服务的开发平台，从而最大限度地提高资源利用率。框架性能将通过三个研究任务和一个验证/评估计划进行评估，考虑智能设备环境中基于增强现实（AR）的应用程序。在任务1中，在考虑LTE/WiFi共存要求的同时，将设计资源分配解决方案，以通过基于业务波动动态地适配分布式单元（DU）密度和托管DU池的虚拟机（VM）的大小来最小化小区站点和中央单元（CU）池处的功耗。在任务2中，将通过跨层设计实现功能划分;将引入新的动态无线电资源分配和灵活的功能划分，以优化NG-RAN中的累积数据速率和网络功耗。在任务3中，将在密集网络中研究服务缓存和任务卸载分配的联合问题，在密集网络中，每个用户都可以利用自由度将其计算任务的不同部分卸载到附近的DU。该奖项反映了NSF的法定使命，并被认为值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估来支持。

The rapid growth of mobile multimedia applications and the Internet of Things (IoTs) have placed severe demands on wireless network infrastructures such as ultra-low latency, user experience continuity, and high reliability. Mobile devices are nowadays the predominant medium of access to Internet services due to an increase in their computation and communication capabilities. However, enabling applications that require real-time, in-the-field data collection and processing using mobile devices is still challenging due to (1) the insufficient computing capabilities and unavailability of aggregated/global data on individual mobile devices and (2) the communication cost and response time involved in offloading data to remote computing resources for centralized computation. In light of these limitations, the Mobile Edge Computing (MEC) concept has emerged, which aims at uniting telco, IT, and cloud computing to deliver cloud services directly from the network edge.. Edge Computing allows for the execution of applications in close proximity to the end users, which reduces the end-to-end delay and the costly backhaul bandwidth consumption back to a centralized cloud computing environment. The flexibility of Next Gen (NG)-Radio Access Networks (RAN) allows a move from a "full centralization" to a "distributed approach" in MEC.The proposed xl_NGRAN framework for 5G (virtualized) cellular networks makes optimized cross-layer decisions for on-demand resource allocation and in-network content caching, and navigates the tradeoffs among radio resources, system cost, LTE/WiFi technology coexistence, and caching service. With a cloud-based framework, and specifically via NG-RAN virtualization, network resources including physical infrastructure and spectrum are abstracted in such a way as to provide a developing platform to support various services, thus maximizing resource utilization. The framework performance will be assessed via three research tasks and one validation/assessment plan considering Augmented Reality (AR)-based applications in a smart-device context. In Task 1, resource-allocation solutions will be designed, while considering LTE/WiFi coexistence requirements, to minimize the power consumption at both the cell sites and the Central Unit (CU) pool by dynamically adapting the Distributed Unit (DU) density and size of the Virtual Machines (VMs) hosting the DU pool based on traffic fluctuations. In Task 2, functional splitting will be enabled thought cross-layer design; a novel dynamic radio-resource allocation and the flexible functional split will be introduced to optimize the accumulated data rate and network power consumption in NG-RANs. In Task 3, the joint problem of service caching and task-offloading assignment will be studied in a dense network where each user can exploit the degrees of freedom in offloading different portions of its computation task to nearby DUs.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.