The number of smart cities is increasing rapidly around the world with the continuous increase of governments’ interest in exploiting Information and Communication Technologies (ICT) to solve issues arising from rapid urbanization. Most smart city services rely fundamentally on ubiquitous sensing, enabled by Wireless Sensor Network (WSN) technologies. However, WSNs in smart cities are naturally vulnerable to unavoidable external challenges like storms, fires, and other natural disasters. Such challenges pose a great threat to smart city infrastructure, including WSNs, as they might affect network connectivity or result in complete blockages of network services. However, some particular smart city services are critical, to the point where they must remain available in all situations, especially during disasters; to monitor the disaster and obtain sensory information needed for controlling it, limiting its danger, or for decision-making during rescue operations. Thus, it is crucial to design a smart-city network to maintain connectivity against such challenges. In this paper, we introduce MPResiSDN, a MultiPath Resilient routing system based on Software Defined Networking (SDN). The system introduced exploits SDN’s capabilities and aided-multipath routing to reactively provide connectivity in smart city networks in the presence of challenges. We evaluated our proposed system under simulations of different natural disasters. The results demonstrate that the system improved data delivery under the challenges by as much as 100% compared to the Spanning Tree Protocol when a suitable value for k diverse paths was selected.

Smart cities are considered to be one of the most important applications of the IoT notion. Most smart city applications rely fundamentally on ubiquitous sensing, enabled by Wireless Sensor Network (WSN) technologies. These sensor networks are vulnerable to different challenges that cause failures in some parts of the network, which in turn interfere with the availability of network services and weaken the user experience. In this paper, we introduce a graph-theoretic model of wireless sensor networks used in smart cities. Moreover, we present several challenges, such as natural disasters and random failures and evaluate the system's performance in terms of data delivery, end to end delay, and energy consumption. The evaluation results show that fire is the challenge that causes the most damage among the three challenges examined, while random failure has the least effect on network performance. The results also show that the modeled WSN's can cope well with the challenge of random failures.