Designing machine learning based security system for smart cities

Immense urbanization is putting tremendous pressure on policy makers, city planners and managers to render sustainable plans and designs which ensures good quality of life for the residents of the smart cities. In order to achieve this goal, a city can be made smart with the ubiquitous development and usage in hardware and software technologies, vast advancement in information and communication technologies and speedy furtherance of connected devices. The newest style of smart city initiative, in turn makes the implementation arduous and challenging. Furthermore, the standards for the design and implementation of the smart city is yet to be released. Consequently, security in this ad-hoc technology is premature and overlooked. In this paper we propose security system that offers active and planned approach to protect the smart city in four spots. Firstly, the machine learning based security system should be placed at the data collection points that has IoT enabled devices. Secondly, the data must be protected by implementing complex cryptographic technique while travelling from the collection points to the intended storage location. Thirdly, the data at rest in the storage area for analysis should be conserved by applying some intelligent intrusion detection technique. Finally, the specific machine generated action should be ensured at the target.