Publications

Wu, Benjamin C.; Ko, Teresa H.

The Seldon terrorist model represents a multi-disciplinary approach to developing organization software for the study of terrorist recruitment and group formation. The need to incorporate aspects of social science added a significant contribution to the vision of the resulting Seldon toolkit. The unique addition of and abstract agent category provided a means for capturing social concepts like cliques, mosque, etc. in a manner that represents their social conceptualization and not simply as a physical or economical institution. This paper provides an overview of the Seldon terrorist model developed to study the formation of cliques, which are used as the major recruitment entity for terrorist organizations.

Ko, Teresa H.; Berry, Nina M.

Wireless sensor networks allow detailed sensing of otherwise unknown and inaccessible environments. While it would be beneficial to include cameras in a wireless sensor network because images are so rich in information, the power cost of transmitting an image across the wireless network can dramatically shorten the lifespan of the sensor nodes. This paper describe a new paradigm for the incorporation of imaging into wireless networks. Rather than focusing on transmitting images across the network, we show how an image can be processed locally for key features using simple detectors. Contrasted with traditional event detection systems that trigger an image capture, this enables a new class of sensors which uses a low power imaging sensor to detect a variety of visual cues. Sharing these features among relevant nodes cues specific actions to better provide information about the environment. We report on various existing techniques developed for traditional computer vision research which can aid in this work.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Ko, Teresa H.; Berry, Nina M.

An important component of ubiquitous computing is the ability to quickly sense the dynamic environment to learn context awareness in real-time. To pervasively capture detailed information of movements, we present a decentralized algorithm for feature extraction within a wireless sensor network. By approaching this problem in a distributed manner, we are able to work within the real constraint of wireless battery power and its effects on processing and network communications. We describe a hardware platform developed for low-power ubiquitous wireless sensing and a distributed feature extraction methodology which is capable of providing more information to the user of events while reducing power consumption. We demonstrate how the collaboration between sensor nodes can provide a means of organizing large networks into information-based clusters. © Springer-Verlag 2004.