CSR: Small: Collaborative Research: CAM: A Cloud-Assisted mHealth Monitoring System

移动医疗（mHealth），特别是移动医疗监测，被认为是最具活力的移动应用程序，在医疗保健行业的革命和不断提高个人生活质量方面发挥着至关重要的作用。不幸的是，由于移动健康监测服务处理的健康和健身相关数据的敏感性和私密性，隐私问题成为广泛部署的绊脚石，必须加以解决。由于资本投资有限，中小型移动医疗公司可能不得不寻求云计算设施，以降低IT支持成本。然而，外包给云将加剧隐私问题，因为公司的监控程序也是专有信息。这个项目的重点是设计一个架构框架，称为CAM：云辅助移动健康监测系统，将其开发为中间件，并将昂贵的计算外包给云。在高层次上，拟议的研究是开发一种潜在的广泛采用移动健康监测服务的使能技术。特别是安全框架的设计，既能保护用户健康和健身数据的隐私，也能保护企业的监控程序，同时又能让云正确执行程序，并向用户返回正确的建议。该设计考虑了外包范例，将大多数计算密集型任务转移到云端，同时仍保留隐私，这是生产实际可部署系统的关键。然后，通过解决诸如合适的编程模型、应用程序开发人员在安全性保证和灵活性之间的平衡等实际问题，该框架被开发成中间件。通过模拟独特的攻击进行全面的渗透测试，以评估所提出的框架在实际系统设置中的安全性。尽管受到移动健康监测应用的推动，但所提出的安全框架可以推广到具有许多其他重要应用（如财务分析和软件故障诊断）的诊断程序的隐私保护外包。因此，拟议的研究将通过对多学科的贡献而产生更广泛的影响，并为研究生和本科生提供大量的多学科研究机会。

Mobile Health (mHealth), particularly mobile healthcare monitoring, has been perceived to be the most dynamic mobile apps which play a crucial role in revolutionizing healthcare industries and steadily improving the quality of individuals' lives. Unfortunately, due to the sensitive and private nature of the health and fitness related data handled by mHealth monitoring services, privacy issues become the stumbling blocks to wide deployment and must be addressed. With limited capital investments, small to medium sized mHealth companies may have to seek cloud computing facilities to reduce the cost on IT support. However, outsourcing to the cloud will aggravate the privacy issues since companies' monitoring programs are also proprietary information. This project focuses on designing an architectural framework, called CAM: a cloud-assisted mHealth monitoring system, developing it into a middleware, and outsourcing expensive computations to the cloud. At a high level, the proposed research is to develop an enabling technology for the potentially wide adoption of mHealth monitoring services. In particular, a security framework is designed to preserve the privacy of users' health and fitness data and companies' monitoring programs while still allowing the cloud to correctly execute the programs and return proper advices to users. The design takes the outsourcing paradigm into account by shifting most computationally intensive tasks to the cloud while still preserving privacy, which is the key to producing a practically deployable system. The framework is then developed into a middleware by tackling practical issues such as a suitable programming model, balancing between security guarantees and flexibility for app developers, etc. Comprehensive penetration testing is conducted by simulating unique attacks to evaluate the security of the proposed framework in practical system settings. Although motivated by mHealth monitoring applications, the proposed security framework can be generalized for privacy-preserving outsourcing of diagnostic programs which have many other important applications such as financial analysis and software fault diagnosis. The proposed research will thus have broader impact by contributing to multiple disciplines and offering both graduate and undergraduate students plentiful opportunities for multidisciplinary research.