DyMoNet: Dynamic Mobile Networks

当前无线接入网络向5G及以后发展的特点之一是提供高带宽业务，并能够为来自大量异构设备的流量提供服务。在此类网络中提供高容量的关键方法中，网络致密化起着突出的作用。然而，许多小型蜂窝基站的密集部署意味着巨大的投资，增加了移动网络的CAPEX和OPEX。此外，由于流量需求随时间和空间的变化，密集化增加了过度供应的网络资源的数量。根据最近对联网汽车的研究，解决这些问题最有希望的方法之一是开发车载小型移动基站。这允许利用用户和车辆的时空模式之间的相关性，以便创建一个灵活和自然的网络，随时随地需要许多用户，从而减少了密集部署静态基站的需要。提出的概念构成了网络设计和运营的新范式，将静态基础设施与车载基站无缝集成，具有前所未有的灵活性和适应性。然而，为了使这种移动网络范式有效地实现动态和灵活的致密蜂窝网络基础设施，需要解决一些关键的开放式研究问题。DyMoNet项目的主要目标是通过解决一些阻碍其实际可行性的关键开放研究挑战来实现这种移动网络范式。具体而言，该项目的目标是：(1)确定移动网络范式相对于传统静态部署而言，在提高网络性能和资源效率方面具有最大潜力的场景和用例；(2)开发可靠的无线移动回程机制，实现移动基站向网络核心的互联；(3)详细阐述动态干扰管理的QoS感知机制，在严格的QoS约束下实现有效的连接供应；(4)研究动态资源分配和切片，以满足特定的QoS约束，同时考虑到服务需求和移动网络基础设施的动态变化。DyMoNet的成果将通过为未来无线接入网络的设计和运营提供新的解决方案，推动当前网络规划和管理的艺术状态。结果将有助于评估移动网络模式的潜力，使提供具有成本效益和资源效率的通信服务成为可能，为经济评价提供技术投入，同时消除其执行的一些主要技术障碍。

The evolution of current wireless access networks towards 5G and beyond is characterized, among others, by the provisioning of high-bandwidth services and by the capability of serving traffic from a large number of heterogeneous devices. Among the key approaches for provisioning high capacity in such networks, a prominent role is played by network densification. However, dense deployments of many small cell base stations imply huge investments, increasing both CAPEX and OPEX for the mobile network. Additionally, densification increases the amount of over-provisioned network resources, due to variability in traffic demand over space and time.Following up recent research on connected cars, one of the most promising approaches to address these issues consists in exploiting vehicle-mounted small cell moving base stations. This allows taking advantage of the correlation between spatio-temporal patterns of users and of vehicles, in order to create a network that flexibly and naturally densifies whenever and wherever needed by many users, hence reducing the need for dense deployments of static base stations.The proposed concept constitutes a new paradigm for network design and operation, which seamlessly integrates static infrastructure with base stations operating on vehicles, with an unprecedented level of flexibility and adaptability. However, for such a moving network paradigm to effectively enable a dynamic and flexibly densified cellular network infrastructure, a number of key open research issues need to be addressed.The main goal of the DyMoNet project is to enable such moving network paradigm, by addressing some of the key open research challenges which stand in the way of its practical feasibility. Specifically, the project will aim at:(1) identifying those scenarios and use cases in which the moving network paradigm holds the highest potential to increase networking performance and resource efficiency with respect to traditional static deployments;(2) developing mechanisms for reliable wireless mobile backhaul, for the interconnection of moving base stations to the core of the network;(3) elaborating QoS-aware mechanisms for dynamic interference management, enabling an efficient provisioning of connectivity under tight QoS constraints; and (4) investigating dynamically resource allocation and slicing in order to satisfy specific QoS constraints, while accounting for the dynamics of service demand as well as of the moving network infrastructure.The outcomes of DyMoNet will advance the current state of the art in network planning and management, by providing new solutions for the design and the operation of future wireless access networks. Results will help assessing the potential of the moving network paradigm for enabling cost-effective and resource-efficient provisioning of communication services, offering technical inputs for economic evaluations, while removing some of the main technical obstacles to its implementation.