"Development of a Reconfigurable Sensor Network for Intrusion Detection"

Andrzej Sluzek, Palaniappan Annamalai
Nanyang Technological University


Wireless sensor networks are expected to become a major tool for various security and/or surveillance applications. Although it is expected that eventually they may be used as fully autonomous systems, in a more realistic scenario, sensors network are devices supporting human decisions with data that have been preliminarily analyzed and interpreted.

The paper reports development of a wireless sensor network with a two-level structure of intrusion detection and classification. At the first level, relatively simple nodes with basic sensing devices (proximity, vibration, acoustic, magnetic, etc.) and wireless transmission capabilities are used. These sensor nodes, with very low energy consumption, are built around a micro-controller of low computational power. In the future, a low-cost FPGA will be added to perform selected data processing algorithms.

The second-level sensor node is built around a high-performance FPGA controlling an array of cameras. Each camera is activated after the corresponding first-level nodes acquire data that may indicate a presence of an intruder, and wirelessly transmit the warning message to the FPGA node. The FPGA can be dynamically configured to perform various types of visual data processing and analysis. Currently, three types of intrusion detection algorithms (sharing a library of image pre-processing operations) are available. These are:

ü       Static intrusion detection – the shape of the intruder(s) is extracted from a captured image and will be further used to classify the intruder by its shape;

ü       Dynamic intrusion detection – the variations of the intruder’s shape (which indicates how the intruder moves) are extracted from a sequence of images; the results can be subsequently used to classify the intruder by the type of its mobility;

ü       Intrusion visualization – the full image of the intruder is extracted from the scene; after adding an image compression algorithm and already implemented data encryption algorithms, this will be used to transmit the compressed and encrypted intruder’s image in the wireless network to a distant human operator for a visual verification.

The implemented algorithms use several image pre-processing operations that are necessary to compensate for the fluctuations of visibility conditions, minor motions of the environment components, vibrations of the sensor platform, etc. Mathematical morphology is the most important methodology used for that purpose.

The presented results are an example of a distributed sensor network combining reasonable energy requirements with a relatively high level of intelligence. In particular, the network can assist a human operator so that he/she is engaged only to confirm the system’s decision whether a detected intruder is a potential danger. The system discussed in the paper is still in the development phase and further improvements are being added to it.


2005 MAPLD International Conference Home Page