Smart city (SC) technologies can provide appropriate services according to citizens’ demands. One of the key enablers in a SC is the Internet of Things (IoT) technology, which enables a massive number of devices to connect with each other. However, these devices usually come from different manufacturers with different product standards, which confront interactive control problems. Moreover, these devices will produce large amounts of data, and efficiently analyzing these data for intelligent services. In this paper, we propose a novel artificial intelligence-based semantic IoT (AI-SIoT) hybrid service architecture to integrate heterogeneous IoT devices to support intelligent services. In particular, the proposed architecture is empowered by semantic and AI technologies, which enable flexible connections among heterogeneous devices. The AI technology can support very implement efficient data analysis and make accurate decisions on service provisions in various kinds. Furthermore, we also present several practical use cases of the proposed AI-SIoT architecture and the opportunities and challenges to implement the proposed AI-SIoT for future SCs are also discussed.

Integrated Space-Terrestrial Network (ISTN) can effectively improve the global communication coverage and SDN framework in ISTN has emerged as a promising solution to guarantee the Quality of Service (QoS). At present, the researches on routing algorithm are mainly focused on the QoS parameters about resource availability and do not consider the impact of the network attacks comprehensively, such as DDoS attack. Plenty of available resources will be occupied by network attacks and normal traffic may be blocked incorrectly by attack defense solutions. In this paper, we propose a Trusted Routing (TR) model based on Trusted Resource Matrix (TRM) to protect the normal traffic from network attacks in ISTN. In the process, the SDN controller constructs the characteristic matrices by collecting the state information of each routing node and the entropy estimation method which describes the changes of traffic characteristics is used to determine whether a routing node is suffering from network attacks. Corresponding trust values are allocated to the routing nodes for constructing TRM and trusted transmission paths are discovered by the TR model. Based on the comprehensive consideration of available and trusted resources, a Hybrid Routing (HR) model is proposed, taking into account the traditional QoS parameters and TRM. In the simulation experiments, the TR and HR models-based algorithms are evaluated. The experimental results show that TR and HR algorithms can effectively avoid the routing nodes suffering from network attacks and improve the communication QoS. When considering a scenario of a single attacked node under resource constraints, the average packet pass rate of HR algorithm is about 7.0% higher than that of TR algorithm and about 52.1% higher than that of traditional QoS algorithm.

Wireless smart home system is to facilitate people's lives and it trends to adopt a more intelligent way to provide services. It is beneficial to design an intelligent smart home (SH) to precisely recognize users' behaviors and automatically responsd with the corresponding activities to satisfy users' actual demands. However, activity models in the existing approaches are usually constructed separately through statistic probability. These models cannot recognize the user's dynamical intentions accurately. To address the problem, we propose a new SH architecture with smart device enabled sensor networks and develop the prototype system. Moreover, we propose the hybrid semantic model based on the statistic probability model and the semantic association model, and an assistance algorithm is presented. In our prototype system, the smart devices are described by semantic models. When the user needs assistance, smart gateway can provide appropriate services according to the inference results of the algorithm. The algorithm has been implemented and the results show that the accuracy of the algorithm based on the hybrid model is higher than the statistic probability model.