|
Martin Heinz Roth
Goal - Objective: My goal is to develop new technologies in the very broad field of emergent systems. I am interested in studying and harnessing the interaction of many artificial individuals on a large scale. There are many potential applications of such technology of interest to both industry and academia, especially as real world systems continue to grow and expand to global scales featuring billions of elements. Examples include (but are definitely not limited to), cooperative/distributed/evolutionary robotics, artificial agents (for any purpose; my PhD thesis focused on their use to solve the routing problem in ad-hoc networks), autonomous computing, or artificial intelligence. Ideally, I would like to continue as a researcher at an institution which will allow me the flexibility and freedom to fully exploit the potential of my ideas together with those of my coworkers. For the time being I would like to have the opportunity to spend time developing and implementing new ideas. Contact:
Education: Cornell University Ithaca, New York, USA August 2000 - January 2005 Doctor of Philosophy, Electrical and Computer Engineering, under Dr. Stephen B. Wicker Dissertation: "Termite: A Swarm Intelligent Routing Algorithm for Mobile Wireless Ad-Hoc Networks" Committee: Dr. Stephen Wicker, Dr. Gun Sirer, Dr. Tom Seeley, Dr. Bruce Land Cornell University Ithaca, New York, USA August 1996 - May 2000 Bachelor of Science in Electrical Engineering, May 2000; with Honors; Cum Laude Concentration in Telecommunications, Digital Signal Processing, and Radars and Antennas Charlotte Country Day School Charlotte, North Carolina, USA August 1992 - May 1996 High School Diploma Skills:
Course Work:
Experience: Research Assistant Cornell University, Ithaca, New Yok, USA August 2000 - December 2004 Researching distributed, swarm intelligent, and emergent algorithms applied to the routing problem in mobile wirelss ad-hoc networks for doctoral degree. Duties included research, analytial analysis, program design and development in C, C++, Java, Matlab, and the Opnet network simulator on Unix, Linux, and Windows computer systems, and technical writing, review, and publication. Studied under Dr. Stephen Wicker in the Wireless Intelligent Systems Laboratory. Consultant Dr. Stephen Wicker, Ithaca, New York, USA 2003 - 2004 Researched technological issues with regards to patent coverage in such diverse topics as wireless system, media content control, wireless data aquisition, and downloadable media. Results used in litigation and settlement disputes. Undergraduate Research Project Mentor Cornell University, Ithaca, New York, USA May - December 2004 Designed, reviewed, and directed an undergraduate student in a summer research project in the area of distributed self-configuration for maximum-lifetime detection coverage in a sensor network. Work based on the use of genetic algorithms to determine a state machine for cellular automata based sensors giving optimal coverage and lifetime. Teaching Assistant: ECE 561 - Error Control Codes Cornell University, Ithaca, New York, USA August - December 2003 Organized, administered, lectured, and graded the final project for students enrolled in this class. Designed and programmed a common Java framework enabling communications over a noisy IEEE 802.11b wireless channel between two laptop computers. Students were required to implement a number of error control codes which could be swapped into the framework as modules. Consultant, Bluetronix Inc. Cleveland, Ohio, USA June - August 2003 Applied thesis research to design and simulate a swarm intelligent ad-hoc routing algorithm. Wrote government grant proposals to secure additional funding. Reviewed business strategy and plans. Lecturer: ECE 296 - HP Nomadic Computing Project Cornell University, Ithaca, New York, USA January - May 2003 Organized and administrated a course of fourteen students, each involved in independant research spanning many engineering diciplines. Coordinated student efforts, made research resources available, and provided guidance over the course of the semester. Served as liason to Hewlett-Packard sponsoring team. Project Management: AKIRA Project Cornell University, Ithaca, New York, USA August 2001 - May 2003 Created and lead the development of an ad-hoc networking testbed including seventy PDAs and ten laptops; the Ad-hoK Information Relay Application (AKIRA) project. Managed a team of four undergraduate and graduate students to write system software using C++ and Java. Used the testbed to test and verify three different MANET routing algorithms. Results used to verify simulations for thesis research. Guest Lecturer: ENGRI 126 - Introduction to Telecommunications Systems Cornell University, Ithaca, New York, USA August - December 2002 Taught select lectures to first year students on topics ranging from local area network technologies (ethernet) to medium access (MAC) technologies (Aloha, CSMA-CD/CA, TDMA, FDMA, CDMA) to satellite communications (link budgets). Research Internship EPFL, Lausanne, Vaud, Switzerland May - August 2001 Learned the funamentals of cryptography and the GSM cellular system. Applied this knowledge to design and develop a secure protocol for financial transactions using GSM handsets. The work was based on the Secure Electronic Transaction (SET) protocol for financial transactions on an open network. The project was implemented and simulated in Java, using the Java Cryptography Extension. The work was accomplished under Dr. Jean-Yves Le Boudec at the Laboratory for computer Communications and Applications. Student Researcher Cornell University, Ithaca, New York, USA August 1999 - May 2000 Revised and updated user movement prediction project from August 1998. This work included additional implementation, analysis, and presentation. A final report was given in March 2000, including a tutorial on bayesian networks and simulation results. Research Internship Zürich Research Laboratory, IBM, Rüschlikon, Zürich, Switzerland May - August 1999 Studied and implemented a turbo decoder using a bayesian network. The similarities between the BCJR Turbo Decoding algorithm and Pearl’s algorithm were considered, followed by an analysis between the latter and Jensen's algorithm for belief propagation in bayesian networks. A simulator was developed in Java in order to test the performance of the bayesian network based decoder. Student Researcher Cornell University, Ithaca, New York, USA August 1998 - May 1999 Self taught Jensen's algorithm for information propagation in Bayesian Networks. Implemented algorithm in Java for use in user movement prediction in mobile networks. These tools were integrated in a larger overall system with several other students in order to test resource allocation schemes in cellular networks. Research Internship Cornell University, Ithaca, New York, USA May - August 1998 Learned function and details of AMPS cellular system. Developed a model of a mobile node using the OPNET network simulator in C. Worked in conjunction with a graduate student to generate thesis results. Results later expanded upon in the following semester; node position data was collected for use with user movement prediction algorithms. Research Internship Cornell University, Ithaca, New York, USA May - August 1997 Developed web based tutorial of various modulation (PSK, FSK, ASK) and medium sharing (TDMA, FDMA, CDMA) schemes for wireless communications systems. Designed a web site featuring Dr. Stephen Wicker's Wireless Multimedia Laboratory at Cornell University. Publications: M. Roth, S. Wicker, Comparison of Swarm Intelligent MANET Routing, (in preparation) M. Roth, S. Wicker, Performance Evaluation of Pheromone Update in Swarm Intellignet MANETs, (under review) > M. Roth, S. Wicker, Asymptotic Pheromone Behavior in Swarm Intelligent MANETs: An Analytical Analysis of Routing Behavior, Sixth IFIP IEEE International Conference on Mobile and Wireless Communications Networks (MWCN), October 2004. An analytical justification is proposed for the design and global routing performance of three pheromone update methods proposed for use in Termite, a swarm intelligent routing algorithm for mobile wireless ad-hoc networks. A simple model is used in order to determine the average amount of pheromone present on a link, as well as some basic aspects of the pheromone dynamics. This includes a tendency towards a one-zero pheromone distribution favoring the better link. The pheromone update methods are investigated with the perspective that link pheromone is more an estimate of link utility than simply a routing heuristic. This allows the routing solution to be rephrased from a biological analogy to a more traditional best-metric routing terminology. A signal estimation perspective is suggested. > M. Roth, S. Wicker, Termite: Ad-Hoc Networking with Stigmergy, Globecom 2003, December 2003. A biologically inspired algorithm is presented to route messages in mobile wireless ad-hoc networks. The principles of swarm intelligence are used to define a probabilistic algorithm for which routing through paths of maximum throughput is an emergent property. This adaptive algorithm, dubbed \emph{Termite}, uses stigmergy to reduce the amount of control traffic needed to maintain a high data goodput. Stigmergy is a process by which information is indirectly communicated between individuals through their environment. The Termite environment is the contents of all routing tables. The movement of packets is influenced at each node, and communicating nodes observe this influence to update their own tables. Strong routing robustness is achieved through the use of multiple paths; each packet is routed randomly and independently. > M. Roth, S. Wicker, Termite: Emergent Ad-Hoc Networking, The 2nd Mediterranean Workshop on Ad-Hoc Networking, June 2003. A biologically inspired algorithm is presented to route messages in mobile wireless ad-hoc networks. The principles of swarm intelligence are used to define a probabilistic algorithm for which routing through paths of maximum throughput is an emergent property. This adaptive algorithm, dubbed \emph{Termite}, uses stigmergy to reduce the amount of control traffic needed to maintain a high data goodput. Stigmergy is a process by which information is indirectly communicated between individuals through their environment. The Termite environment is the contents of all routing tables. The movement of packets is influenced at each node, and communicating nodes observe this influence to update their own tables. Strong routing robustness is achieved through the use of multiple paths; each packet is routed randomly and independently. Presentations: - October 29, 2004, Termite: A Swarm Intelligent Routing Algorithm for Mobile Wireless Ad-Hoc Networks, Swarm Intelligent Systems Research Group, Prof. Alcherio Martinoli, EPFL Citizenship and Residency: Citizen of Switzerland Green Card holder of the United States of America Hobbies - DJing (Drum & Bass and Hardcore Techno) - Electronic music production |