IRES Track 1: Secure Crowdsensing for Improving Smart City Applications

该项目将每年派遣3个小组，每组5名学生，在德国的班贝格大学进行为期10周的研究。 班贝格大学在整个大学和德国的班贝格市运营着一个智能城市生活实验室，该实验室能够从整个城市的传感器收集特定区域的噪音水平，二氧化碳水平和人数等信息。 智慧城市收集信息并进行处理，以帮助决策者更全面地了解城市中人与服务的复杂互动，从而做出更好的决策。 一些信息来自被认为值得信赖的来源，如城市工作人员部署的闭路电视摄像机;然而，这些值得信赖的来源的设置和维护成本往往很高。 众包信息提供了一种更便宜、更有效的获取信息的方式;然而，由于人们可能会撒谎，或者恶意软件可能会代表他们撒谎，因此这些信息的可信度要低得多。 我们的研究考察了在这样一个智能城市环境中所有实体之间的相互作用，并开发了利用众包信息的方法，同时保护城市免受恶意用户的攻击，并保护用户免受可能通过其数据监视他们的人的攻击。 研究人员将在德国的生活实验室班贝格实施和测试他们的工作，以验证我们的安全系统的有效性。 预计这项工作将带来更安全，更智能的智慧城市，并有可能提高能源效率，通信，交通和公共卫生。该项目将解决智慧城市安全和隐私方面的几个科学挑战。选择班贝格大学的地点是因为该大学在校园和城市都运营着智能城市生活实验室。 研究人员将解决以下三个科学挑战：i）利用来自不可信来源的数据至关重要，因为智能城市中的许多数据源都是用户的智能手机，这些用户可以故意操纵数据，或者可能在他们的手机上安装恶意软件来策略性地操纵数据。 学生将使用博弈论和机器学习的组合来构建防御机制。用户的隐私对于在智慧城市中采用人群感知以及防止用户和管理员滥用系统至关重要。 学生将建立隐私增强技术，使用真实生成的覆盖流量来混淆用户数据和网络物理相关性，以便对城市收集的数据进行隐私保护审计。（三） 安全激励有助于保护任务的部署者免受恶意用户的攻击，这些恶意用户试图优化他们的收益而不对系统做出贡献。 为了确保恶意用户不利用奖励系统，学生将创建上下文信息的时空模型，以检测用户在贡献信息时是否确实在该位置。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。

This project will send 3 cohorts of 5 students per year for 10 weeks each to conduct research at the University of Bamberg, Germany. The University of Bamberg operates a Smart City Living Laboratory throughout the university and the city of Bamberg, Germany that is able to collect information such as noise levels, CO2 levels, and the number of people in a particular area from sensors throughout the city. Smart cities collect information and process it to help decision makers make better decisions with a fuller understanding of the complex interactions of people and services in a city. Some of the information comes from sources that are considered trustworthy such as CCTV cameras deployed by city workers; however, these trustworthy sources are often expensive to set up and maintain. Crowd sourcing the information provides a cheaper and often more efficient way of obtaining information; however, since people can lie or malicious software could lie on their behalf, this information is significantly less trustworthy. Our research examines the interactions between all of the entities in such a smart city environment and develops methods for utilizing crowd sourced information while protecting the city from malicious users and protects users from people who might spy on them through their data. Researchers will be implementing and testing their work in the Living Lab Bamberg in Germany to validate the effectiveness of our security systems. It is expected that this work will result in safer, more intelligent smart cities with the potential to improve energy efficiency, communication, transportation, and public health.This project will address several scientific challenges in the security and privacy of smart cities. The site at the University of Bamberg was chosen due to the smart city living laboratory that the university operates on both the campus and city. The researchers will address the following three scientific challenges: i) Utilization of data from untrusted sources is critical because many of the data sources in a smart city are the smart phones of users who can intentionally manipulate data or who might have malware on their phone that is strategically manipulating data. Students will build defense mechanisms using a combination of game theory and machine learning. ii) Privacy of users is critical to the adoption of crowd sensing for use in a smart city and to prevent the abuse of the system by users and administrators alike. Students will build privacy-enhancing technologies that use realistically generated cover traffic to obfuscate user data and cyber-physical correlation to enable privacy-preserving audits of the data collected by the city. iii) Secure incentives help to protect the deployers of tasks from malicious users that are seeking to optimize their gain without contributing to the system. To ensure that malicious users do not exploit the incentive system, students will create spatio-temporal models of context information that detect if a user was actually at the location at the time they reported they were when contributing information.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.