Application of Self Organizing Neural Net to Animal Communications.
Advisor: Dr. Zohdy
In this project, Self Organizing Feature Maps are trained to categorize animal communication sounds into danger
hunger and mating calls for Humpback, Whales, Bottlenose, Dolphins and Coyotes. Features are extracted in time domain
frequency domain and joint time frequency domain from audio files of animal communication. Unknown calls are then fed into
the map to identify the sound. Several contributions have been made to the maps to improve the quality speed and accuracy.
Correlation has been implemented as the activation function to improve the accuracy of identifying sounds. A history function to
update previous winning nodes has been implemented to decrease the time necessary to complete the training phase. Sammon mapping
has also been proposed to provide better initialization of vectors.
Intra-Vehicle UWB Channel Measurements
Advisor: Dr. Li
Ultra Wide Bandwidth (UWB) acts as a potential way to construct the intra- vehicle wireless sensor network. We measured and modeled by UWB propagation channel for the commercial vehicle environment. We have two scenarios of channel sounding and their measurements. In the first scenario we have both antennas beneath the chassis, the other scenario both antennas are inside the engine compartment. The results that are given are to be group into clusters. A cluster is commonly known to be a group of multipath components with similar parameters. Clusters are usually identified manually. However we have a very large amount of data, this would be impractical to do. Therefore we designed an algorithm to cluster the data automatically. Our algorithm clusters our data with little or no error.
Issues and Challenges in Route Guidance
Advisor: Dr. Mili
Route guidance refers to all decision aids that help drivers select a
route from their source to destination location. In its simplest form,
route guidance is the search for a shortest path in a directed graph.
In practice, the problem is complicated by the facts that: 1.Many time
variable factors affect the quality of a route, 2. optimal decisions
made in light of a current configuration (state of the roads, traffic
congestion) change the situation by potentially causing congestions
elsewhere and making the selected route sub-optimal. In this project,
we survey existing algorithms, analyze their properties, and perform a
comparative simulation using the TRANSIMS environment. TRANSIMS is an
integrated system of travel forecasting open source traffic simulation
environment developed by the Los Alamos National Laboatory for the
Department of Transportation.
Privacy Issues with VANET
Advisor: Dr. Fu
Vehicular Ad-Hoc Networks (VANETs) are networks of communication between vehicles and roadside units. These networks have the potential to increase safety and provide many services to drivers, but they also present risks to privacy. Researching mechanisms to protect privacy requires two key ingredients: 1. a precise definition of privacy that reflects citizens concerns and perceptions, and 2. an understanding of the type of attacks that VANETS are vulnerable to. In this research, we formulate a workable definition of privacy, and focus on tracking attacks, which we found to be lacking. Although considerable research has been performed in tracking none of the published solutions ensures full protection. We propose to combine a set of published solutions, namely: Mix Zones, Silent Periods, and Group Signatures in order to improve the protection of privacy of drivers. Vehicles enter a region where, vehicles change their pseudonyms (Mix Zone) as well as network addresses, and then enter the silent period, and use one group key for communication. The solution prevents attackers from linking transmission to a particular vehicle after an intersection. It could help make tracking more difficult and increase the safety and confidence of drivers using VANET.