Monitoring of Drainage System in Urban Using Device Free Localization Neural Networks and Cloud computing

Abstract

Wireless Sensor Network is a Wi-Fi community consisting of spatially propagated and self-sufficient devices using sensors to detect physical or environmental conditions. During heavy rainfall, the urban drainage system cannot drain the water. A wireless sensor with many interconnected wireless sensor nodes captures real-time data from the network environment and transmits this data to a base station for analysis and operation. With wireless sensor nodes, it is possible to capture and monitor the amount of water in drainages and the difference in water flow between the two points in the drainage system. Nevertheless, the majority localization techniques aims on device based localization, which can find target with festinated devices. It is not suitable for applications such as terrain, drainage flow and flooding. Here device free wireless localization system using artificial neural networks and a cluster based wireless sensor network system to monitor urban drainage is proposed. There are two stages in the system. During the off-line preparation stage, Acceptable Signal Strength (RSS) differential metrics are calculated between the RSS metrics together while the monitor area is empty and calculated by a specialized in the region. Some RSS dissimilarity values ​​are selected in the RSS Difference Matrix. The RSS dissimilarity standards ​​and associated matrix indices are taken as the inputs of the ANN representation in addition to the identified position coordinate are in its outputs. The real-time data collected from the wireless sensor network is used to detect overflow and provide alarms before disturbances arise.