CAREER: Leveraging Context-Aware Sensing to Enhance QoE for Mobile Users: A Practical Framework Design

由于功能强大的移动的终端的激增，诸如视频流、视频通话和在线游戏的资源需求多媒体应用在过去十年中已经激增。如何提高终端用户主观感知的这些应用的满意度，广泛地解释为体验质量（QoE），对于无线网络的运营商来说至关重要。流行的QoE增强方法大多是面向运营商的，而用户扮演的角色有些有限。另一方面，移动的设备配备有多个嵌入式传感器，其使得能够收集和推断丰富的上下文信息。受这些观察的启发，该项目旨在利用以用户为中心的上下文感知来增强无线网络中的QoE。所提出的用户辅助QoE增强如果成功，则有可能作为对现有集中式QoE管理框架的有希望的补充。此外，它将开辟新的机会，将移动的用户纳入网络测量和管理，这主要是由网络运营商单独进行的。该项目计划让女性和代表性不足的少数民族学生参与研究活动，并积极向初中和高中学生推广。该学术发现将广泛传播给社区。该项目旨在开发一个实用的框架，以利用上下文感知来增强QoE，并解决其中涉及的关键挑战。更具体地说，拟议的研究将通过三个密切相关的研究重点进行。第一个推力研究细粒度的用户上下文预测的基础上收集的历史记录从移动的设备。作为另一个关键因素，它进一步研究如何利用预测结果来帮助无线网络中的资源配置和QoE改善。意识到上下文感测可能会产生额外的通信和能量开销，第二个推力旨在通过开发具有成本效益的上下文采样和适应上下文动态的准确重建来降低成本。第三个推力解决了上下文感知过程中的隐私问题。将开发灵活的上下文交易框架，以允许个人用户决定在QoE改善交易中披露多少隐私。拟议的研究是跨学科的，涵盖了资源管理，学习技术，压缩感知，差分隐私和经济理论。建议的机制和设计将通过测量活动，模拟和实验研究的组合进行彻底评估。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Due to the proliferation of powerful mobile terminals, resource-demanding multimedia applications, such as video streaming, video calls, and online gaming, have been surging during the past decade. How to improve the satisfaction of these applications as perceived subjectively by end-users, widely interpreted as the quality of experience (QoE), is critical to the operators of wireless networks. Prevailing QoE enhancement methods are mostly operator-oriented, while users play somewhat limited roles. On the other hand, mobile devices are equipped with multiple embedded sensors that enable the collection and inference of rich contextual information. Motivated by these observations, this project aims to leverage user-centric context sensing to enhance QoE in wireless networks. The proposed user-assisted QoE enhancement, if successful, has the potential to serve as a promising complement to the existing centralized QoE management framework. Besides, it will open up new opportunities that incorporate mobile users for network measurement and management that has been mostly conducted by network operators alone. The project plans to engage female and under-represented minority students in the research activities and actively outreach to middle and high school students. The scholarly discovery will be disseminated broadly to the community.This project intends to develop a practical framework to leverage context sensing for QoE enhancement and address critical challenges involved therein. More specifically, the proposed research will be carried out through three closely related research thrusts. The first thrust studies fine-granular user context prediction based on historical records harvested from mobile devices. As another key ingredient, it further investigates how to utilize the prediction results to assist with resource provisioning and QoE improvement in wireless networks. Realizing that context sensing may incur extra communication and energy overhead, the second thrust intends to reduce the cost by developing cost-efficient context sampling and accurate reconstruction that adapt to context dynamics. The third thrust addresses the privacy concern during context sensing. A flexible context trading framework will be developed to allow individual users to decide how much privacy to disclose in the trade of QoE improvement. The proposed research is inter-disciplinary, which spans resource management, learning techniques, compressive sensing, differential privacy, and economic theories. The proposed mechanisms and designs will be thoroughly evaluated via a combination of measurement campaigns, simulations, and experimental studies. A medium-scale prototype of off-the-shelf mobile devices, routers, servers, and cloud will be implemented.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.