Study On Building Security Monitoring Technology Based On Wireless Sensor Networks

In recent years, with the rapid development of wireless sensor networks (WSNs), it has become a bridge which connects the physical world with the computer world. Wireless sensor networks not only have the advantages like low wiring cost, high monitoring precision, good fault-tolerance, etc., it also can be monitered remotely. It has outstanding merits in fault diagnosis, security monitoring, and so on. Since the building security accidents happened frequently, how to take real-time and effective monitoring for building’s health status, avoid the building security accidents caused by significant structure or equipment damage, fire or building indoor environment change, and provide timely and scientific guidance to health and maintenance of a building have get more and more people’s attention. Hence, it is of great theoretical significance and social benefits to set a building security monitoring system based on wireless sensor networks.This dissertation analyzes the current situations, progress and challenges faced by the building security monitoring research, summarizes the system structure, characteristics and technical requirements of wireless sensor networks, researchs the design and implementation of wireless sensor networks in the fields of building structure health monitoring, building fire security monitoring and building indoor environmental monitoring. The main contributions of the dissertation are summarized as follows.1. Due to the large number of time and resource consumption required by the traditional building structural health monitoring system, this dissertation tries to develop a lightweight monitoring system based on WSNs. A genetic ant colony algorithm is proposed to realize multicast routing and multihop routing, so as to improve the quality of service and stability of the networks. An improved ant colony algorithm is utilized to save the energy consumption of the networks, so as to enlarge the network lifetime.2. To deal with the problem of limited function, complex architecture, and high bandwidth requirement in current building fire security monitoring system, this dissertation proposes a novel fire and flammable gas detection WSNs system based on multiple agents theory. A dynamic scheduling scheme is proposed to solve the large scale computing problem involved in the multi-tasks scheduling, so as to reduce the running time and improve the efficiency of task scheduling. A novel D-S data fusing algorithm is utilized to deal with the realtime data set, so as to improve the accuracy and stability of the monitoring alarm system, and decrease the false positive and non-response rate as well.3. The traditional building indoor environmental monitoring system generates lots of data to transmit. However, there is only limited spectrum resource and power energy. To deal with this problem, this dissertation builds a spectrum sensing mode and proposes a spectrum sensing algorithm based on supporting vector machine. The proposed scheme can sense the available spectrum resource accurately, realtimely, and effectively. With the sensed spectrum resource, the throughput and transmission efficiency improve remarkably. Meanwhile, this dissertation also proposes an improved DV-hop algorithm based on compressive sensing theory. The proposed algorithm could reduce the number of samples, save power energy consumption, and improve network lifetime, while without any performance degradation of the networks.