A wireless sensor network system for border security and crossing detection

The protection of long stretches of countries’ borders has posed a number of challenges. Effective and continuous monitoring of a border requires the implementation of multi-surveillance technologies, such as Wireless Sensor Networks (WSN), that work as an integrated unit to meet the desired goals. The research presented in this thesis investigates the application of topologically Linear WSN (LWSNs) to international border monitoring and surveillance. The main research questions studied here are: What is the best form of node deployment and hierarchy? What is the minimum number of sensor nodes to achieve k− barrier coverage in a given belt region? iven an appropriate network density, how do we determine if a region is indeed k−barrier covered? What are the factors that affect barrier coverage? How to organise nodes into logical segments to perform in-network processing of data? How to transfer information from the networks to the end users while maintaining critical QoS measures such as timeliness and accuracy. To address these questions, we propose an architecture that specifies a mechanism to assign nodes to various network levels depending on their location. These levels are used by a cross-layer communication protocol to achieve data delivery at the lowest possible cost and minimal delivery delay. Building on this levelled architecture, we study the formation of weak and strong barriers and how they determine border crossing detection probability. We propose new method to calculate the required node density to provide higher intruder detection rate. Then, we study the effect of people movement models on the border crossing detection probability. At the data link layer, new energy balancing along with shifted MAC protocol are introduced to further increase the network lifetime and delivery speed. In addition, at network layer, a routing protocol called Level Division raph (LD ) is developed. LD utilises a complex link cost measurement to insure best QoS data delivery to the sink node at the lowest possible cost. The proposed system has the ability to work independently or cooperatively with other monitoring technologies, such as drowns and mobile monitoring stations. The performance of the proposed work is extensively evaluated analytically and in simulation using real-life conditions and parameters. The simulation results show significant performance gains when comparing LD to its best rivals in the literature Dynamic Source Routing. Compared to DSR, LD achieves higher performance in terms of average end-to-end delays by up to 95%, packet delivery ratio by up to 20%, and throughput by up to 60%, while maintaining similar performance in terms of normalised routing load and energy consumption.