Information Technology seminars
A Scalable, Distributed and Real Time Pattern Recognition
- Date and time:
- 01/07/2009, 10:30
- Location:
- Building: 26, Room: 135, Clayton Campus
- Presenters:
- Amir H. Basirat
- Abstract:
- Wireless Sensor Networks (WSNs) are an emerging domain of deeply networked systems of low-power wireless devices characterized by tiny CPU-memory and large federated networks for intricate sensing of the environment. The capability to support a plethora of new and diverse applications has placed Wireless Sensor Network technology at the threshold of significant development. In this regards, pattern recognition especially in real-time applications will play a paramount role in event detection and securing the network against malicious activity of an adversary. This presentation will discuss attack models in WSNs. In addition, a novel method using a highly distributed associative memory technique called Hierarchical Graph Neuron (HGN) will be explained as an effective method to detect anomalous activity within the network. The proposed approach not only promises to deliver accurate results but it can also be deployed for diverse types of applications in a multidimensional domain.
A professionally designed Java program demonstrating the effectiveness of the proposed pattern recognition approach for a WSN will be presented at the end of this talk.
- Enquiries:
- Asad Khan
