Source-Channel Coding, Data Privacy and Contagion Mitigation in Communication and Information Networks

这一建议涉及到我们在网络信息理论和通信方面的研究计划的加强和延续。具体而言，我们将为通信和信息网络开发先进的源信道编码、数据隐私和传染缓解解决方案。这一三管齐下的目标解决了未来一代大数据网络面临的关键和互补挑战，其中信息需要在多个设备之间可靠和迅速地交换，在不损害用户隐私的情况下有效利用，并安全地存储在数据库中，免受恶意攻击的污染。在取得成果的基础上，我们将重点抓好以下几个主题：******(i)双向通信：建立每个终端同时接收和发送信号的双向通信网络的核心香农理论极限。根据我们的理论发现，我们将为全双工通信系统设计最先进的源信道代码。******(ii)数据隐私：使用信息和估计理论方法研究分散网络的数据隐私机制。关键目标是使数据收集代理能够在不泄露敏感用户信息的情况下优化系统效用，并为大数据存储库构建有效的隐私保护算法。******（三）减轻传染：为网络流行病建立随机传染模型，特别强调准确表示通信网络中的恶意软件传播和社会网络中的意见动态。我们还将设计并彻底验证基于最大熵和博弈论原理的稳健治疗技术，用于大规模网络中的感染控制和缓解。******建议的研究将以每年严格培训七名高素质人员（3名博士，2名硕士和2名本科生）为特色，使他们沉浸在信息论，无线通信，数据安全和网络流行病等具有挑战性的主题中，并产生创新的高质量出版物。本研究结果可望为双向网路的源信道编码理论与实务提供更深入的了解，并有助于无线通讯与感测网路技术的永续发展。预计它还将有助于部署弹性用户隐私技术，以及明智地控制和遏制大数据网络中的传染性现象

This proposal concerns the enhancement and continuation of our research program in network information theory and communications. Specifically, we will develop advanced source-channel coding, data privacy and contagion mitigation solutions for communication and information networks. This three-pronged objective addresses critical and complementary challenges facing future generation big data networks in which information needs to be reliably and promptly exchanged between multiple devices, efficiently harnessed without compromising user privacy and securely stored in databases that are immune to contamination from malicious attacks. Building on our achievements, we will focus on the following main themes.******(i) Bi-Directional Communication: Establish core Shannon-theoretic limits for two-way communication networks where each terminal receives and transmits signals simultaneously. Guided by our theoretical findings, we will then design state-of-the-art source-channel codes for full duplex communication systems.******(ii) Data Privacy: Investigate data privacy mechanisms for decentralized networks using information and estimation theoretic methods. Key aims are to enable data collection agents to optimize system utility without leaking sensitive user information and construct effective privacy-preserving algorithms for big data repositories.******(iii) Contagion Mitigation: Develop stochastic contagion models for network epidemics with special emphasis on the accurate representation of both malware spread in communication networks and opinion dynamics in social networks. We will also design and thoroughly validate robust curing techniques based on maximum-entropy and game-theoretic principles for infection control and mitigation in large-scale networks.******The proposed research will feature the rigorous training of seven highly qualified personnel per year (3 Ph.D., 2 M.Sc. and 2 undergraduate students), immersing them in challenging topics in information theory, wireless communications, data security and network epidemics and resulting in innovative high-quality publications. The results of this research are anticipated to provide a deeper understanding of the theory and practice of source-channel coding for two-way networks and contribute to the sustainable evolution of wireless communications and sensor networks technologies. It is also expected to contribute to the deployment of resilient user privacy techniques and the judicious control and containment of contagious phenomena in big data networks.**