CS Poster Presentations at GSA Research Event

Three CS students, Tom Palmer, Shohrab Hossain, and Qing Zhao presented posters at the GSA 2011 Spring Research conference held at the National Weather Center in Norman, OK.

Shohrab Hossain

Abstract of Shohrab Hossain's Poster:

Mobile networks can be formed in bus, train, aircrafts, satellites with a wide variety of on-board IP-enable devices and Network Mobility (NEMO) protocols are required to support uninterrupted services to ongoing sessions. Node mobility has a direct impact on the performance evaluation of various NEMO protocols. However, most of the analysis on mobility protocols used random waypoint mobility model which does not always represent real-world movement patterns in city streets. In this paper, we have developed an analytical model to analyze the stochastic properties of city section mobility model, a realistic street mobility model. We have used the model to analyze the performance of the basic network-mobility protocol. We have used ns-2 simulation to compare the performance of NEMO using city section and random waypoint models. Results have been obtained for average throughput, packet drop probability, end-to-end delay, handoff frequency, and signaling overhead and show significant deviation between the mobility models. Our analysis thus can help in estimating the various performance metrics of mobile network deployed in city streets.

Tom Palmer

Abstract of Tom Palmer's poster:
Many logical planning systems depend on user-provided predicates and functions for representing the environment. In Blocks World tasks, for example, the basic predicate on(X, Y) represents that block X is on block Y. From such a basis, sophisticated algorithms can solve tasks of interesting complexity. However, automatically inventing such predicates is often considered a difficult problem. Our goal here is to work from physical attributes about items and thereby infer logical relations. Currently, we focus on human-labeled examples of world situations with different consequences. To learn relations, we employ the Spatiotemporal Multidimensional Relational Framework of Bodenhamer at al. (2009). Going forward, we seek to discover distinct situations in a more automated fashion as well as to improve the speed and reliability of learning.

Qing Zhao

Abstract of Qing Zhao's poster:

A high efficiency class-E power amplifier (PA) based on Gallium Nitride (GaN) high electron mobility transistor (HEMT) is designed and tested. Compact impedance transformation network is implemented using microstrip transmission lines, so that it simultaneously achieves fundamental load transformation and harmonic impedance control. At 2.705GHz, the PA provides 42dBm output power, power added efficiency (PAE) of 58% and gain of 12.5dB.