Reliable FPGA-Based Network Architecture for Smart Cities

Nowadays, smart cities are progressively established for increasing governments operational efficiency and for citizens satisfaction. This research proposes a novel network architecture for smart cities integrating several functionalities such as traffic jam monitoring, water quality control and security. The architecture is based on three types of nodes: central node (sink), sensing node (bouquet of sensors) and actuating node. Both the sink and the sensing node are based on FPGAs for their high computation capability and for protection of investment. The Sink is responsible for the collection of data from the sensing nodes and performing Fog computing. Each bouquet of sensors also performs Dew computing. Video analytics are used for several purposes in the system. The architecture is verified by Riverbed Network Modeler and it is shown that there is no packet loss. To increase the reliability of the system, fault-tolerant sensor nodes are implemented using a proposed analog voting technique.