Poster Abstracts

Jins Alexander
Auburn University
832.588.8480
Masters
Electrical Engineering

An Efficient Power Estimation Tool for CMOS Circuits

We have developed a gate-level (total) power estimation tool forCMOS circuits. Average, maximum and minimum values of each component of power, namely, dynamic (separate logic and glitch), short circuit, and leakage are determined. Such a tool has useful applications because, in circuit design, the objective is total power optimization and methods to reduce any one component affecting others.

Lei Chen
Auburn University
334.844.3659
Doctorate
Computer Science

Secure Media Streaming with Path Diversity in Mobile Ad Hoc Networks (MANETs)

Media streaming in Mobile Ad Hoc Networks is quite challenging because of the limited bandwidth, signal loss and signal interference. Securing the huge amount of data in media communication makes this task even more difficult to complete. In our research, we try to increase the bandwidth and reduce delay by utilizing Path Diversity, or making use of multiple paths between the source and destination nodes. For security purpose, we propose a smart data distribution mechanism; to balance the traffic in MANETs without draining any node, a selective encryption/decryption mechanism is also applied.

E. Vincent Cross II
Auburn University
334.844.6353
Doctorate
Department of Computer Science and Software Engineering

Prime III

Prime III is a voting system that delivers the requisite system security, integrity, and user satisfaction safeguards in a user friendly, secure, electronic voting system.  Moreover, Prime III incorporates the current voting process that the voter is accustomed to using and significantly improves upon it.  This approach allows the voter to remain comfortable and confident while using an enhanced voting system. The design goal was to develop a robust multimodal application through user centered design principles that combines security with user interaction to create a usable security system that facilitates user interaction with the application, through multiple means (i.e. text-to-speech, speech-to-text, touch).  This multimodal approach permits voters to hear and/or see the candidate names while they cast their votes (via voice and/or touch). 

Shuangchi He
Auburn University
334.332.6656
Doctorate
Electrical Engineering

Direct FIR Linear Equalization of Doubly Selective Channels Based on Superimposed Training

Design of doubly-selective linear equalizers for single user frequency selective time-varying channels is considered using superimposed training and without first estimating the underlying channel response. Both the time-varying channel as well as the linear equalizers are assumed to be described by a complex exponential basis expansion model (CE-BEM). A periodic (non-random) training sequence is arithmetically added (superimposed) to the information sequence at the transmit ter before modulation and transmission. There is no loss in information rate. Knowledge of the superimposed training is exploited to design the FIR linear equalizer. An illustrative simulation example is presented.

Jorge S. Hurtarte
Auburn University
334.821.1569
Doctorate
Electrical Engineerinig

A Single-Chip Programmable RF MEMS Filter for SDR Cellular Pico BTS

The RF front end circuitry makes up more than 60% of the cost of a base transceiver station (BTS, also referred as Node B in 3G networks). Included in this cost is that of the Power Amplifier unit, RF filters, and the GPS unit for clock recovery and synchronization. Depending on the architecture complexity, number of supported sectors, power range specification, and wireless network standards supported, a base transceiver station can cost between $10,000 to $40,000. RF MEMS (microelectromechanical systems) is a technology that has the potential to not only solve the BTS cost and logistic issues, but to also complement SDR architectures by providing a true low cost, high performance, programmable RF Front End filter. The authors are researching the possibility of implementing a Single-Chip Programmable RF MEMS Filter for SDR Cellular Pico BTS for commercialization by 2010 or earlier, with particular attention to possible solutions to the issue of maximum switching speed limitation of RF MEMS.

Benjamin C. Kelley
Auburn University
859.948.4916
Undergraduate
Wireless Engineering

Mote Channel Selection in Noisy 2.4 GHz Enviroments

In environments where motes are deployed, several 802.11b/g access points frequently overlap a single area; more than three access points in an area limit the available frequencies because of the overlapping umbrellas associated with each 802.11 channel, limiting motes' operating quality. Motes attempting to operate in these noisy channels will experience high packet loss due to corruption from interference. The motes cannot compete with access points due to power requirements for transmission power levels. Motes allow setting the channel to a value within the 2.4Ghz spectrum, but only a small fraction of the available spectrum exists outside of the predominant 802.11 b/g channels (beyond channel 11). Using this small area of spectrum exclusively for motes will eventually result in increased mote-mote interference. Experimenting with channel selections and power levels for the motes will provide data indicating the sensitivity to surrounding 2.4Ghz noises.

Shuang Li
Auburn University
334.524.6647
Doctorate
Computer Science

A High-Throughput And Low-Delay Protocol for Sensor Networks

This paper presents a compound metric, Cost, which combines both throughput and delay in wireless sensor networks. We employ ETX (expected transmission count) metric for each link and we propose a new way to compute ETX for a path, which doesn't exaggerate the effect of interference among the successive links of a path. We put forward a self-defined equation to compute the Cost for a path, based on which a proper path is reinforced locally.  This paper describes the design and simulation of Cost for directed diffusion. We get the results from emulation on 20 computers connected by wired network that the metric Cost improves throughput with a factor of two or more and keeps the roughly satisfactory delay value.

Cong Liu
Auburn University
215.609.0179
Doctorate
Computer Science

P2PGS: A Distributed Grid Scheduler Adopting P2P Resource Discovery Strategy Y

In a grid computing environment, dynamicity and geographically distributed sites, make task scheduling problems challenging to solve. It is hard for a local site to obtain precise real-time information about other sites given that specific information on a site such as load and computing resources may change rapidly. Moreover, in data grid environment, large scale data intensive applications make scheduling problems even more challenging since both computational and data storage resources must be taken into consideration. In this paper we propose an innovative peer-to-peer scheduler to solve these problems. This scheduler is distributed and scalable. We used simulation to evaluate the performance of the scheduler under different circumstances, such as different number of hops to search suitable sites and different number of incoming tasks. Results show that our scheduler can successfully schedule around 75% of incoming tasks within their deadlines in average. For computation-intensive tasks, it can successfully schedule more than 90% of incoming tasks.

Yuanlin Lu
Auburn University
334.703.6891
Doctorate
Electrical Engineering

Statistical Leakage Optimization for Submicron Process Variation

A mixed integer linear programming method is proposed for dual-threshold design that minimizes the leakage power and circuit delay in a statistical sense such that the impact of process variation on the leakage and timing yields are minimized. Results show up to 30% greater leakage reduction than the deterministic approach.

Jeff E. Metzler
Auburn University
973.332.3706
Undergraduate
Software Engineering

The Top Ten Smart Phones

Smart phones have transformed cell phones into mobile, networked computers.  This technology unites an assortment of electronic devices — cell phones, handheld computers, MP3 players, cameras — into a single platform that has more capabilities than computers of a half-decade ago.  As versatile as these devices are, they have not caught the imagination of the consumer market as well as manufacturers predicted.  There are a myriad of reasons for this, but complexity is the predominate factor.  Smart phone software development requires the skill of a desktop computer software engineer as well as the attention to detail of an embedded systems programmer.  An examination of the top ten smart phones on the market today reveals the intricacies of cell phone software engineering:  operating system alternatives, service-level APIs, battery power management, user interface models, carrier relations, and business model support.

Raghu Kisore Neelisetti
Auburn University
334.444.1166
Doctorate
Computer Science and Engineering

Wireless Acoustic Sensor Networks for Target Tracking

Target tracking involves estimating various trajectory parameters, such as position, slope of the trajectory and velocity. We are designing and implementing a method for estimating these parameters using CPA (Closest Point of Approach) measurements from a group of four (minimum) wireless acoustic sensors. The tracking algorithm based on CPA measurements requires an uneven distribution of sensors deployed on either side of the target trajectory, with a reference node on one side of the target trajectory and the other three nodes on the other side. However, given a random deployment it is not easy to choose the right set of four nodes that can successfully track a target trajectory. But if we can group together five nodes instead of four, then irrespective of the target's trajectory we will always have three nodes on one side of the target trajectory and two nodes on the other side. We have implemented a simple algorithm (in Ns-2) that will dynamically group the deployed nodes into groups of five. The algorithm further provides desired level of efficiency and redundancy.

Ryan W. Northington
Auburn University
205.296.0256
Undergraduate
Wireless Software Engineering

Wireless Scoring System for Fencing Tournaments

The project involves the design of an automated, wireless scoring system for fencing tournaments. Sensors in the weapons detect scoring events during a match. This information is to be transmitted wirelessly to a "box" on the scorer's table that will signal events, and tabulate and display scores in real time for both of the fencing weapons (foil and epee). We will be using a microcontroller developed and built by Cypress Semiconductor Corp. in conjunction with their WirelessUSBTM solution. WirelessUSB is designed for short range multipoint to point wireless connectivity, while offering the benefits of low cost and low power.

Santosh Pandey
Auburn University
334.728.0716
Doctorate
Electrical Engineering

Sophisticated Node Placement Algorithms in Hierarchical Networks

In this work, we address the problem of node placement in a hierarchical heterogeneous wireless sensor network. We consider a two-tiered wireless sensor network where the resource constrained Lite Nodes (LNs) are used for sensing the environment and high-end Sophisticated Nodes (SNs) are added to aggregate and forward data. We intend to place minimum number of SNs to handle the traffic generated by LNs and ensure that the SNs form a connected network. We formulate the node placement problem as an optimization problem and use three different algorithms to solve it; namely, Binary Integer Linear Programming (BILP), Greedy algorithm (GREEDY) and Genetic Algorithm (GA). We also propose a hybrid approach (HYBRID) combining BILP, GREEDY and GA to improve results. It was found through simulations that GA performed better for random LN deployment. However, using HYBRID, results comparable to original GA could be obtained in only 11.46% of the time required for the original GA. We support the results with statistical tests.

Andrew R. Putnam
Auburn University
601.248.8987
Undergraduate
Wireless Software Engineering

JME - The Mobile Platform

The average cell phone user may not realize it, but they are using a small computer.  In fact, they may be carrying one that they can program.  Over 80% of the cell phone handset models on the US market today are programmable in the Java programming language, making Java one of the most widely accessible languages of all time.  The version of Java that runs on cell phones — Java Platform, Micro Edition (JME) — is the same language as used on desktop computers, but that is where the similarity stops.  JME software development requires radically more attention to detail.  An examination of the specifics of engineering software in JME shows that the developer must be adept at implementing applications that involve event loops, moderate computing capability, constrained user interfaces, limited power consumption, intermittent network connectivity, and increased robustness.  Moreover, JME programmers must contend with extra-software issues such as over-the-air provisioning, cell phone carrier constraints on content, and cost models.

Jie Qin
Auburn University
334.844.1800
Doctorate
Electrical Engineering

FPGA-Based Analog Functional Measurements for Adaptive Control in Mixed-Signal Systems

A Field Programmable Gate Array (FPGA)-based Built-In Self-Test (BIST) approach used for adaptive control in mixed-signal systems is presented.  It provides the capability to perform accurate analog functional measurements of critical parameters such as the 3rd order intercept point (IP3), frequency amplitude and phase responses and noise figure (NF).  The results of these measurements can then be used to adaptively control the analog circuitry for calibration and compensation.  The BIST circuitry consists of a direct digital synthesizer (DDS) based test pattern generator (TPG) and a multiplier/accumulator based output response analyzer (ORA).  The BIST approach has been implemented in an FPGA-based mixed-signal system and used for actual analog functional measurements.  The BIST measurements agree quite well with the results obtained with the traditional analog test equipment.  The proposed BIST circuitry provides a unique means for high-performance adaptive control in mixed-signal systems.

Christopher Rose
Auburn University
256.520.0319
Undergraduate
Electrical Engineering

Mobile Cell Phone Booth

The principle of noise cancellation is applied to speaker's speech. With the goal of providing the privacy of a phone booth to a cell phone user on the move, a digital signal processor feeds the speech signal in opposite phase to a loudspeaker with similar directional characteristics as the human speaker.

Nathan C. Shafer
Auburn University
706.315.1719
Undergraduate
Wireless Software Engineering

Emerging Trends in Wireless Technology

Software development has come what seems like a full circle.  Programming in the early days of computers required a careful marriage of hardware and software.  Programmers had to have an intimate knowledge of the hardware with which they were working in order to make the software fit, both functionally and physically. 
As programming languages became more high-level, developers were able to ignore to a large extent the underlying hardware when implementing software applications.  Mobile devices have returned software engineers to their original roots — designing and writing software for these platforms require careful attention to a host of low-level details, such as program size, system speed, power consumption, network bandwidth, and user interaction all of which are typically tightly constrained.  We examine emerging trends in software engineering technology based on cutting-edge features available on mobile devices today, and forecast what is required of software developers to keep pace with those trends. 

Chengjun Wang
Auburn University
334.524.6641
Doctorate
Computer Science

Cache Interconnection Architecture for Chip Multiprocessors

There is a growing trend to put multiple cores on the die. The interconnects among the cores on the same chip are quite different from those well understood for connecting chips, multi-chip modules, and board-level nodes because they share common limited resources, such as power, area. Our goal is to design the best cache to memory and cache to cache (in multi-core) interconnection architecture to maximize on-chip service and minimize off-chip misses for best performance and fault tolerance. Possible interconnection architectures are shared bus, wireless, point-to-point, or crossbar. Power, area, latency, and bandwidth are all design constraints. Using the design constraints, we assess different interconnects. Our cache to memory design uses a secondary channel to commit write backs while the main bus is busy with I/O operations. The parallelization of I/O with memory traffic improves performance significantly.

Xiaoyun Wei
Auburn University
334.844.1873
Doctorate
Electrical Engineering

An Improved On-chip 4-port Parasitics De-embedding Method with Application to RF CMOS

For on-wafer 2-port measurement, the measured test structure includes not only the desired internal device, but also probing pads and interconnects. These parasitics can be significant at high frequencies, and must be deembedded. The on-chip parasitics between the two measurement ports and the two device ports can be viewed as a 4-port. We presents an improved algorithm to solve the general 4-port parasitics de-embedding problem. Experimental results on 0.13 µm RF CMOS device are presented. The reciprocity and symmetry of the 4-port parasitics are also examined.

Christopher G. Wilson
Auburn University
334.462.8682
Masters
Electrical Engineering

Wireless Communication for Collaborative Search and Rescue

Collaborative map building for search and rescue missions requires wireless communication. This communication must be optimized with respect to a number of parameters including the amount of data to be exchanged, security requirements, number of entities involved, available power, and the communication interval. Previous work has investigated these tradeoffs in simulation. Presently, a laboratory-scale hardware demonstration is under construction which uses collaborating mobile robots equipped with WiFi transceivers configured as a mobile ad-hoc network. Theoretical and practical aspects of the communication system's design will be presented.

Shaoen Wu
Auburn University
334.844.6918
Doctorate
Computer Science and Software Engineering

Multi-Interface Multi-Channel Mobile Terminals

We propose a multiple interface and multiple channel architecture for wireless terminals. With such architecture, each wireless terminal is configured with multiple interfaces, each of which operates on a different wireless channel to associate with a different neighboring access point. Such wireless terminal architecture can work either as active-standby or load-balance mode. It can significantly improve throughput, enhance user response and enable seamless handoff. Moreover, this architecture explores the potential capacity of the existing wireless network infrastructure deployed ubiquitously today. Our measurement from real experiments confirms these benefits.

Qing Yang
Auburn University
334.844.6324
Doctorate
Computer Science

Connection-based Routing Protocol for Vehicular Networks

A connection-based routing protocol is proposed for the vehicular network. Due to the high mobility feature, the end-to-end connection in such networks will become more fragile.  Different from other works, we focus on establishing route paths which have the highest probability of full connection. Connectivity model for vehicles on street and intersection is invented from the information of the road density and average velocity of traffic flow.