These days, the demand for smart city infrastructures and systems that can support real-time functions are rising, fog computing technology is increasingly adopted to meet the demand for real-time processing and latency-sensitive and scalable applications, this is due to fog capabilities of performing processing closer to the edge nodes. However, the adoption of fog computing is still behind the cloud which takes up a huge portion of the implementation approaches and causes time-critical services to suffer. In this paper, we reviewed relevant literature on fog computing and mobile development approaches to highlight the development approaches taken in developing smart cities with this technology, a classification table was made to classify the papers based on three classes such as service-based, application-based, and resource-based which was then broken into subclasses based on each paper's contribution. This study also extracts a component table highlighting the components, tools, metrics, etc. used by the authors. This paper contributes to highlighting new approaches taken in the development of smart cities.