Collaborative Research: CNS Core: Medium: Foundations and Scalable Algorithms for Personalized and Collaborative Virtual Reality Over Wireless Networks

协作研究：CNS 核心：中：无线网络上个性化和协作虚拟现实的基础和可扩展算法

• 批准号: 2106679
• 负责人: Atilla Eryilmaz
• 金额: $ 26.6万
• 依托单位: Ohio State University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-10-01 至 2025-09-30
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2106679&HistoricalAwards=false
• 关键词: Collaborative; Research; CNS; Core; Medium

Virtual reality (VR) over wireless networks can provide an interactive and immersive experience for multiple users simultaneously and thus has many applications, especially in VR-based education/training. However, satisfactory personalized user experience in such wireless immersive services demands stringent performance requirements, including: (1) high-speed and high-resolution panoramic image rendering; (2) extremely low delay guarantees; and (3) seamless user experience. Besides the aforementioned requirements, collaborative user experience requires both scalability and fairness of VR service. Existing VR systems heavily rely on various heuristic designs and do not efficiently exploit VR content commonality and its predictability, which impede their large-scale deployment. This project aims to develop the theoretical foundations and complete implementation of a system for providing both personalized and scalable collaborative VR experience over wireless networks. This project will integrate machine learning, wireless networking, and mobile computing to enable high-quality and scalable wireless immersive applications on commodity mobile devices. The theory and practical implementations to be developed in this project will be integrated into both undergraduate and graduate curriculum, as well as exposing K-12 students to state-of-the-art wireless and VR technologies.The proposed designs are motivated by a number of insights that we have developed from our preliminary work, including (1) viewport-adaptive rendering; (2) commonality among VR content for multiple users to enable multicasting; and (3) predictability of VR content to enable prefetching. The proposed research will contribute to and advance both theoretical and system-oriented research in the fields of wireless networks and virtual reality. The project explicitly exploits the unique characteristics of both immersive VR applications and wireless networks, and propose the following four interdependent research thrusts: (I) Dealing with network and prediction uncertainties: This thrust will investigate algorithm designs to optimize personalized user experience given both network and viewport prediction uncertainties. (II) Meeting stringent immersive service requirements: This thrust will develop wireless scheduling algorithms that provide stringent immersive, personalized service guarantees for multiple VR users. (III) Supporting smooth collaborative interaction: This thrust will focus on the algorithm design that leverages the VR content similarities and predictabilities that naturally emerge during collaborative interactions. (IV) Scalable system integration, implementation, evaluation, and deployment: This thrust will integrate Research Thrusts I through III into a holistic system, perform system-level optimizations, and evaluate it through lab experiments and real-world classroom deployment.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.

通过无线网络的虚拟现实（VR）可以同时为多个用户提供交互式和沉浸式体验，因此具有许多应用，特别是在基于VR的教育/培训中。然而，在这种无线沉浸式服务中，令人满意的个性化用户体验需要严格的性能要求，包括：（1）高速、高分辨率的全景图像渲染； (2) 极低的延迟保证； (3) 无缝的用户体验。除了上述要求外，协作用户体验还需要VR服务的可扩展性和公平性。现有的 VR 系统严重依赖各种启发式设计，无法有效利用 VR 内容的共性及其可预测性，这阻碍了其大规模部署。该项目旨在开发系统的理论基础和完整实现，以通过无线网络提供个性化和可扩展的协作 VR 体验。该项目将集成机器学习、无线网络和移动计算，以在商用移动设备上实现高质量和可扩展的无线沉浸式应用程序。该项目中要开发的理论和实际实现将整合到本科生和研究生课程中，并使 K-12 学生接触最先进的无线和 VR 技术。所提出的设计受到我们从前期工作中得出的许多见解的启发，包括 (1) 视口自适应渲染； (2) VR内容之间的共性，供多个用户启用组播； (3) VR 内容的可预测性以实现预取。拟议的研究将有助于并推进无线网络和虚拟现实领域的理论和系统研究。该项目明确利用沉浸式 VR 应用和无线网络的独特特征，并提出以下四个相互依赖的研究主旨：（I）处理网络和预测不确定性：该主旨将研究算法设计，以在网络和视口预测不确定性的情况下优化个性化用户体验。 （二）满足严格的沉浸式服务要求：该重点将开发无线调度算法，为多个VR用户提供严格的沉浸式、个性化服务保证。 (III) 支持流畅的协作交互：该重点将侧重于利用协作交互过程中自然出现的 VR 内容相似性和可预测性的算法设计。 (IV) 可扩展的系统集成、实施、评估和部署：该主旨将把研究主旨 I 至 III 整合到一个整体系统中，执行系统级优化，并通过实验室实验和现实世界的课堂部署对其进行评估。该奖项反映了 NSF 的法定使命，并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，被认为值得支持。

项目成果

Efficient Distributed MAC for Dynamic Demands: Congestion and Age Based Designs

• DOI: 10.1109/tnet.2022.3191607
• 发表时间: 2023-02
• 期刊: IEEE/ACM Transactions on Networking
• 作者: Xujin Zhou;Irem Koprulu;A. Eryilmaz;M. Neely

A Lyapunov-Based Methodology for Constrained Optimization with Bandit Feedback

• DOI: 10.1609/aaai.v36i4.20285
• 发表时间: 2021-06
• 作者: Semih Cayci;Yilin Zheng;A. Eryilmaz

Fresh Caching of Dynamic Content Over the Wireless Edge

• DOI: 10.1109/tnet.2022.3170245
• 发表时间: 2022-10
• 期刊: IEEE/ACM Transactions on Networking
• 作者: B. Abolhassani;John Tadrous;A. Eryilmaz;E. Yeh

Communication-efficient Subspace Methods for High-dimensional Federated Learning

• DOI: 10.1109/msn53354.2021.00085
• 发表时间: 2021-12
• 期刊: 2021 17th International Conference on Mobility, Sensing and Networking (MSN)
• 作者: Zai Shi;A. Eryilmaz

Regret-Optimal Learning for Minimizing Edge Caching Service Costs

• DOI: 10.23919/wiopt56218.2022.9930560
• 发表时间: 2022-09
• 期刊: 2022 20th International Symposium on Modeling and Optimization in Mobile, Ad hoc, and Wireless Networks (WiOpt)
• 作者: Guocong Quan;A. Eryilmaz;N. Shroff