Smart city development is important to achieve sustainable cities and societies which help enhance urban services, reduce resource consumption and decrease overall cost. The incorporation of smart cities with the Internet has given us the Next Generation of Internet (NGI) where every smart device exploits the interconnected services and infrastructure of the Internet. The underlying structure of NGI is composed of large scale heterogeneous multilevel systems-of-systems (SoSs) where each system represents a sensor, mobile phone, computer or smart device.

Security and privacy is a fundamental requirement of NGI which is heavily dependent on the composition of services and connectivity of the underlying systems. Meaning any unsecure system can affect the security of the entire networked infrastructure/SoSs. Therefore, it is important to analyse and understand the composition of different systems at different levels in NGI in order to identify and mitigate vulnerabilities. This paper proposes a solution to identify and mitigate vulnerabilities within multilevel SoSs, to enhance security without deploying additional security at endpoints, and quantify security levels of individual systems and the entire composed system. The solution was tested and evaluated using simulation and a network testbed. Results show that NGI security can be enhanced with better composition of systems.

智慧城市发展对于实现可持续的城市和社会至关重要，这有助于增强城市服务，减少资源消耗并降低总体成本。智能城市与Internet的融合为我们提供了下一代Internet（NGI），其中每个智能设备都可以利用Internet的互连服务和基础设施。NGI的基础结构由大规模异构多级系统（SoS）组成，其中每个系统代表传感器，移动电话，计算机或智能设备。

安全和隐私是NGI的基本要求，它在很大程度上取决于服务的组成和底层系统的连接性。意味着任何不安全的系统都可能影响整个网络基础架构/ SoS的安全性。因此，重要的是分析和理解NGI中不同级别的不同系统的组成，以便识别和缓解漏洞。本文提出了一种解决方案，用于识别和缓解多级SoS中的漏洞，以增强安全性而无需在端点上部署额外的安全性，并量化单个系统和整个组合系统的安全级别。使用模拟和网络测试平台对解决方案进行了测试和评估。结果表明，更好的系统组成可以增强NGI的安全性。