Vehicular Computation Offloading for Real-time Applications

实时应用的车辆计算卸载

• 批准号: RGPIN-2018-06499
• 负责人: Zhao, Dongmei
• 金额: $ 4.81万
• 依托单位: McMaster University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=752324
• 关键词: Vehicular; Computation; Offloading; Real; time

With the ever increasing expectation for intelligent transportation, very powerful computing capabilities are required to improve road safety, enhance transport efficiency, and enrich the travel experience. For example, video clips captured by many cameras in and around vehicles are processed for avoiding collisions, interactive maps and augmented reality are embedded in many transportation-related applications, and games with high resolution virtual reality are common in the mobile devices of passengers. These computation-intensive applications are often time critical, due to road safety or their interactive nature. As a result, supporting these services and applications may greatly exceed the processing capability of single vehicles or mobile devices.The concepts of mobile cloud computation and computation offloading enable vehicles and mobile devices to offload part or all the computation load to other devices that help speed up the executions. As a result, the traditional vehicular communication systems are being transformed from mainly providing communication services for drivers and passengers to an integrated system of communications, computing, and control. Powerful cloud servers may be accessed by vehicular users through wireless networks and the Internet, but they are more suitable for delay-tolerable services due to the relatively long delay to access the servers. For time-sensitive applications, servers should be deployed close to the vehicular users to reduce the communication delay. If road side units and cellular base stations are available along a road, dedicated edge servers can be installed and co-located with these infrastructure devices. However, it is unlikely that this fixed infrastructure can cover all roads in the near future due to possible low capital returns, complicated design issues, unpredictable communication and computation demands, and other reasons. On the other hand, future smart vehicles are likely to have, or can be equipped with, relatively powerful computation units, and can either help other devices or collaborate together in processing complicated tasks. Given the large number of smart devices on road and their wide geographical distribution, their aggregate computing capability can be enormous. The objective of this research is to study how to effectively integrate the available computing resources at pre-installed computing devices and moving vehicles along roads, and efficiently utilize these resources with the support of the existing vehicular communication networks to provide safe and convenient transportation and enriched travel experiences. The new ideas and knowledge generated from this research will promote a great market opportunity for various road and travel related products and applications, which will be of strong interest to automobile industry, wireless network providers and users, and software industry.

随着人们对智能交通的期望不断提高，需要非常强大的计算能力来改善道路安全、提高交通效率、丰富出行体验。例如，车辆内部和周围的许多摄像机捕获的视频片段经过处理以避免碰撞，交互式地图和增强现实被嵌入到许多交通相关应用中，高分辨率虚拟现实的游戏在乘客的移动设备中很常见。由于道路安全或其交互性质，这些计算密集型应用程序通常对时间至关重要。因此，支持这些服务和应用程序可能会大大超出单个车辆或移动设备的处理能力。移动云计算和计算卸载的概念使车辆和移动设备能够将部分或全部计算负载卸载到其他设备上，从而有助于加速执行死刑。传统的车载通信系统正在从主要为驾驶员和乘客提供通信服务转变为通信、计算和控制的综合系统。功能强大的云服务器可以被车辆用户通过无线网络和互联网访问，但由于访问服务器的延迟相对较长，因此更适合延迟容忍的服务。对于时间敏感的应用，服务器应部署在靠近车辆用户的地方，以减少通信延迟。如果道路沿线有路边单元和蜂窝基站，则可以安装专用边缘服务器并与这些基础设施设备共置。然而，由于可能的低资本回报、复杂的设计问题、不可预测的通信和计算需求以及其他原因，这种固定基础设施不太可能在不久的将来覆盖所有道路。另一方面，未来的智能汽车很可能拥有或可以配备相对强大的计算单元，并且可以帮助其他设备或协同处理复杂的任务。鉴于道路上的智能设备数量众多且地理分布广泛，它们的聚合计算能力可能非常巨大。本研究的目的是研究如何有效整合预装计算设备和沿道路移动车辆的可用计算资源，并在现有车载通信网络的支持下高效利用这些资源，提供安全、便捷的交通和丰富的生活。旅行经历。这项研究产生的新想法和知识将为各种道路和旅行相关产品和应用带来巨大的市场机会，这将引起汽车行业、无线网络提供商和用户以及软件行业的浓厚兴趣。