Introduction

Increasing individual use of connected smart devices, rapid growth of worldwide urban population, gradual aging of society in many countries as well as the rising demand for sustainable energy resources have given momentum to the emergence of smart cities and smart spaces [1]. Smart city services span a wide spectrum of applications ranging from smart utilities, to smart health, smart transportation, smart governance, and smart environment [2], which Utilize real-time sens- ing, knowledge engineering, and presentation of the analyzed data in an interpretable format. To fulfill the requirements of diverse applications and services, a smart city architecture consists of five planes as illustrated in Fig. 1 in a minimalist manner. The application plane provides services to the end users for any relevant application such as smart utilities, energy, transportation, health, environment and safety. Serving mobile, home, and corporate sectors, these applications rely on an underlying substrate, which encompasses sensing, communication, data, and security planes as the core of a smart city architecture [3]. The data acquisition (sensing) networks—implemented by utilizing either hard sensors or soft sensors—form the sensing and actuation plane; whereas processing, analysis, and storage of data shape the main functionality of the data plane. These two planes are bridged by the communication and aggregation plane. Addressing users’ growing cognizance about smart city cybersecurity, each component shown in Fig. 1 calls for security and privacy assurance mechanisms. The challenging task of satisfying these requirements is relegated to the security and privacy plane. As shown in Fig. 1, a comprehensive and effective security and privacy plane must be implemented adjacent to every individual building block of this architecture. The Internet of Things (IoT) is a major enabler for diverse smart applications that involve massive data acquisition and intelligent decision making [4]. Therefore, in the realization of smart cities, the IoT is a bridging component between the sensing devices and the data plane [5, 6]. Indeed, in order for the IoT concept to operate thoroughly, its interface with the local wireless networks, backhaul networks, as well as local wired networks needs to be addressed properly. Furthermore, protocols for reliable and efficient data acquisition methodologies for fusing data are of paramount importance particularly to ensure a robust communication back-end in a smart city infrastructure. In a smart city architecture, communication back-end is one of the most crucial components, which is responsible for pre-processing and aggregation of the sensory data. Reliability, usefulness and trustworthiness of the data acquired by the communication back-end depends on the effectiveness of the sensing plane. As studied in [2], sensors in a smart city setting can be deployed in either a dedicated or non-dedicated manner; each deployment strategy has its own pros and cons. Based on the observations above, in this paper, we present a smart city architecture by briefly introducing its building blocks (termed planes), namely the application, communication, sensing, data, and security planes. Upon brief introduction of the architectural building blocks, we move to our main foci, which are the sensing, communication and security planes. In the study of the sensing plane, we present the dedicated and nondedicated sensing paradigms from the standpoint of various smart city applications. We thoroughly investigate the sensor types, problems experienced by the corresponding smart city applications, existing solutions and the communication technologies used by the sensors. Sensing plane is followed by the communication plane, which is studied in terms of the requirements, implications, and common solutions. We partition the communication plane into several sub-planes to study the communication infrastructure and protocol design for a smart city with fine granularity.