Telecom Research at Rutgers University Fred S Roberts

Telecom Research at Rutgers University Fred S. Roberts Director, DIMACS froberts@dimacs. rutgers. edu July 11, 2003 Commission on Jobs Growth and Economic Development 1

Outline • What is Telecom? • The research players and their background • A selection of research topics suggested by Rutgers faculty – Short synopsis – Food for thought; aimed at stimulating discussion July 11, 2003 Commission on Jobs Growth and Economic Development 2

What is Telecom? The set of technologies and sciences at the intersection of communications, information, and computing Communications Information Telecom Computing July 11, 2003 Commission on Jobs Growth and Economic Development 3

What is Telecom? • 20 th century: transport, switching, and storage of narrowband voice and data • 21 st century? Reasonable goal: fully integrated and networked broadband multimedia including: – data of all types – text, images, audio, video – virtual reality – searchable, browseable multimedia documents – shared reality tele-collaboration July 11, 2003 Commission on Jobs Growth and Economic Development 4

What is Telecom? So: telecom is all about networks: • interconnections of networks (e. g. , the Internet) • operation and maintenance of networks • things that make up networks (routers, hubs, switches) • things that get moved around networks (data, text, voice, images, video, …) • things that attach to networks (devices, sensors, monitors) • services that run on networks July 11, 2003 Commission on Jobs Growth and Economic Development 5

Outline • What is Telecom? • The research players and their background • A selection of research topics suggested by Rutgers faculty – Short synopsis – Food for thought; aimed at stimulating discussion July 11, 2003 Commission on Jobs Growth and Economic Development 6

Major Hubs of Telecom Research at Rutgers CAIP (Center for Advanced Information Processing) DIMACS (Center for Discrete Mathematics and Theoretical Computer Science) WINLAB (Wireless Information Laboratory) Computer Science Dept. Statistics Dept. July 11, 2003 Commission on Jobs Growth and Economic Development 7

CAIP • Established by NJCST in 1985 • Mission: industrial applications of advanced computing technologies • Industry partners: AT&T, Avaya, Cisco, Datacube, Fujitsu, General Motors, IBM, Info. Value, Intel, Iscan, Kodak, Lucent, NEC, NIST, Oracle, OSS Nokalva, Panasonic, Sarnoff, Siemens, Speech. Works, SUN, Telcordia, Texas Inst. , CECOM, Picatinny Arsenal, Verizon, Xybernaut • University partners: UMDNJ, NJIT, Princeton, Stanford, MIT, Johns Hopkins, CMU, Colorado, Cal Tech, Columbia, New Mexico State • 86 faculty, visiting scientists, staff, and students July 11, 2003 Commission on Jobs Growth and Economic Development 8

• • • Telecom-related Research at CAIP Multimodal interfaces (NSF) Image and speech pattern recognition (DOD) VLSI design (NJCST) Bio/nano mechatronics (NSF) Applications to homeland security (DOD, CECOM) Si. C semiconductors (DARPA, Union Carbide) Collaborative networking (DOD, NSF) Distributed grid computing (NSF) Data visualization (NSF) Telemedicine/rehabilitation (NSF, Novartis) Virtual environments (NSF) Speech production (NIH) July 11, 2003 Commission on Jobs Growth and Economic Development 9

Telecom-related Research at CAIP: Future Opportunities • • Natural communication with information systems. Virtual environments for collaboration Internet delivery of rehabilitative therapies Autonomic grid computing Systems and sensors on a chip Detection of radioactive materials Human imaging for dosimetry analysis Low bit-rate communication for security July 11, 2003 Commission on Jobs Growth and Economic Development 10

CAIP Cumulative Impacts: • External Contract Funding: ($17 M in current contracts) • Ph. D. ’s and MS’s graduated: • Patents filed: • Startup companies assisted: • CAIP spinoff companies created: • Small business outreach, new jobs: July 11, 2003 Commission on Jobs Growth and Economic Development $55 M 213 80 20+ 3 100+ 11

DIMACS • The importance of discrete math and theoretical CS (algorithm development) led Rutgers, Princeton, AT&T Bell Labs, and Bellcore to develop strong research groups • In 1988, they joined to form DIMACS • Telecommunications: AT&T Labs, Bell Labs, Telcordia, Avaya • Computing: NEC Research, IBM Research, Microsoft Research, HP Labs (Princeton) • 1989: prestigious NSF “science and technology center” award. $10 M grant largest at Rutgers. NJCST played important role. July 11, 2003 Commission on Jobs Growth and Economic Development 12

Telecom-related Research at DIMACS • • next generation networks technologies computational information theory and coding communication security simulations of communication architectures computer-aided verification of software massively parallel computing massive data sets applications of large scale discrete optimization to communication networks • cryptography • complexity of interactive computing July 11, 2003 Commission on Jobs Growth and Economic Development 13

Telecom-related Research at DIMACS - II telecom researchers find new applications of their methods: • homeland security research • epidemiology/public health • computational biology • DNA computing July 11, 2003 Commission on Jobs Growth and Economic Development 14

Telecom Research at DIMACS • More than $50 M in external funding for research and education program at DIMACS since its inception • NSF, ONR, NSA, NIH, DARPA, ICMIC (intelligence community), Sloan Foundation, Burroughs Wellcome Fund, NJCST, numerous companies. • Solution of Gilbert-Pollak conjecture led to highly efficient heuristics for design of communication networks. • Pioneer in field of computer-aided verification; methods now used widely by Intel, Sun, Motorola, AT&T, Lucent. • Simulation software for the global internet adopted by more than 40 companies/universities. July 11, 2003 Commission on Jobs Growth and Economic Development 15

Telecom Research at DIMACS • Work on error-correcting codes led to new techniques for the design of efficicient encoders and decoders. • A remarkably simple on-line algorithm for bin packing small information packets of varying sizes into bins of fixed capacity. • Powerful cryptographic methods for secure authorized access. • The “players” at DIMACS – 230 scientists from partner universities and companies – partner company scientists directly involved in DIMACS projects – more than 1000 visitors a year July 11, 2003 Commission on Jobs Growth and Economic Development 16

WINLAB • Founded in 1989 • Broad experimental and theoretical expertise in wireless technologies • Broad collaborative experience with industry: – about 20 industry sponsors – major partners brought into NJ include Intel, Nortel, Thomson, Samsung, NTT, Sprint, Motorola, Mitsubishi, … • Implementing technology transfer through both sponsor companies and startups July 11, 2003 Commission on Jobs Growth and Economic Development 17

Telecom Research at WINLAB • • • Freebits -- short range ultra high speed communications NJ Center for Wireless Communication 4 th Generation Radio Resource Management Adaptive Networking for 3 rd Generation Cellular Security in Next Generation Wireless Dynamic Spectrum Management First Generation of MUSE sensor program Research Wireless Testbed Cognitive Network Management July 11, 2003 Commission on Jobs Growth and Economic Development 18

Telecom Research at WINLAB • Pioneer in “hot spot” wireless networking technology now appearing at Starbucks, Mc. Donalds, etc. through its “infostations” program. Going out to startups and Army STTR tech transfer. • ~20 faculty/staff + ~40 -50 students • Currently over $2 M a year in funding. • This year won IEEE Marconi and William R. Bennett Awards July 11, 2003 Commission on Jobs Growth and Economic Development 19

WINLAB: Implications for the Future • Wireless is fastest growing segment of telecom. – – – Almost 500 M cell phones sold/year. 1/3 of calls in US are already wireless, not wired (FCC). 148 M US subscribers, ~ half the population. (FCC). $76 B Wireless revenues in 2002; 30% of telecom (FCC). Wireless Data devices market expected to be $10 B+ by end of 2003. – 21 M American users of Wireless Hot Spots by 2007 (IBM). • 6, 300 global hotspots in 2001; expect 114, 000 by 2006. (IBM). • NJ needs a world-class center of expertise in all major areas of wireless communications. July 11, 2003 Commission on Jobs Growth and Economic Development 20

Economic Impact of Wireless Research in NJ • 100’s of new high tech jobs/year via startups and partnerships. • Retain high tech talent in NJ. • Train and retain the best students. • Diversify the telecom industrial base in NJ through diverse wireless end-user applications, not just the traditional (and now stagnant) core infrastructure. July 11, 2003 21

Outline • What is Telecom? • The research players and their background • A selection of research topics suggested by Rutgers faculty – Short synopsis – Food for thought; aimed at stimulating discussion July 11, 2003 Commission on Jobs Growth and Economic Development 22

Sensor RF Modem/CPU Multimodal Integrated Wireless Sensor on-Silicon (MUSE) (Win. LAB, ECE, BME, CS, UMDNJ, Ga. Tech) • Today, sensors are individual units as transistors once were. – Temperature, pressure, light, chemicals, etc. – Expensive controllers, readouts, and communications – Usually physically large and often hand-made. • With new technology, we should be able to link sensors in complex networks to gather information in new ways. July 11, 2003 Commission on Jobs Growth and Economic Development 23

Multimodal Integrated Wireless Sensor on-Silicon (MUSE) • Networks of Sensors: – Applications to medicine, consumer, environment, security, military, etc. – Need new wireless networking technology – Need ultra-low cost sensors and controllers • New sensor technology that can measure many properties • Ultra low power electronics, algorithms, and protocols • All on one chip, reusing as much of integrated circuit technology as possible July 11, 2003 Commission on Jobs Growth and Economic Development 24

Multimodal Integrated Wireless Sensor on-Silicon (MUSE) • Low Cost, Wireless Networked Sensors – Strongly multidisciplinary program – Draws in all levels of technology from devices to networks to applications and security. – Build on strengths of the partners and ongoing programs. – Too large to tackle without cohesive program with a shared vision and strong core funding. July 11, 2003 Commission on Jobs Growth and Economic Development 25

Multimodal Integrated Wireless Sensor on-Silicon (MUSE) • Economic impact for NJ – Market for integrated sensors estimated at $3 B in 2005 and $10 B in 2010 • This is before security adders or changes in military needs. – Can build on existing industrial partnerships and experience to make the technology transfer happen. July 11, 2003 Commission on Jobs Growth and Economic Development 26

Locating Mobile Users • Estimating the location of wireless communications users attracts huge attention • Applications include – location-aware services • finding the nearest vending machine or printer • finding the nearest buyer or seller in a market of buyers and sellers • in a museum setting, presenting artifact-specific descriptions on a handheld device • locating a misplaced handheld device July 11, 2003 Commission on Jobs Growth and Economic Development 27

Locating Mobile Users • Other applications: – emergency location • identifying the room location of a crime victim • in a prison setting, locating a distressed guard – access control • blocking access to a Wi-Fi network from outside a building • blocking access for specific users from specific locations July 11, 2003 Commission on Jobs Growth and Economic Development 28

Locating Mobile Users • Team from Rutgers (Statistics Dept. ) and Avaya Inc. (wireless expertise) have developed novel and highly sophisticated statistical algorithms unlike any of the existing approaches • Substantially more accurate location estimation with dramatically less training data • Immediate application in enterprise settings July 11, 2003 Commission on Jobs Growth and Economic Development 29

Locating Mobile Users • Every major telecommunications company working on this problem • Tremendous commercial potential • Avaya team has significant experience with wireless technology and markets • Rutgers has a long track record of funding and innovation in statistical methods • Urgent need for seed funding for experimentation and software development July 11, 2003 Commission on Jobs Growth and Economic Development 30

Massive Data Analysis Lab (Mass. DAL) • Agenda: Gather, manage and process massive data logs---Web, IP/wireless traffic data, location trajectories of objects, sensor readings of physical world. July 11, 2003 Commission on Jobs Growth and Economic Development 31

Massive Data Analysis Lab (Mass. DAL) • Key Challenges: – Scale: Beyond the traditional “human” scale. Eg. , IP data at a single router interface for an hour exceeds total yearly worldwide credit card transactions! – Data Collection: probes/sensors with associated data quality and communication problems. • Need breakthroughs in Mathematics, Algorithms, Systems and Engineering, to meet these challenges. • Potential: Major impact in Telecom, Transportation and Society-at-large. July 11, 2003 Commission on Jobs Growth and Economic Development 32

State of Mass. DAL • Engineering: – Consulting in analysis of wireless network logs. Client: AT&T Wireless, 3 rd largest in US, 20 Million customers. Terabytes/month. Current value: $3 M per year. 5 pers. Fully operational, telco-grade! Interest from Cingular wireless. – Incorporated novel algorithms in operational IP network data analysis tools. Current partner: AT&T. Potential partner: Lucent. • Mathematics and Computer Science. – Algorithms, Databases, Statistics, and Data Mining on novel models and algorithms. – Supported by NSF grants. Partners: Rutgers CS, DIMACS, MIT. July 11, 2003 Commission on Jobs Growth and Economic Development 33

State of Mass. DAL (Contd) • Science – Developing wearable sensors for tracking location of objects as well as “interactions” between objects. – Current partner: Telcordia. Their initial investment: $300 K/3 months (est). Potential partner in works: Los Alamos National Lab. – Potential: Analysis of social networks for Epidemiology and Homeland Security. July 11, 2003 Commission on Jobs Growth and Economic Development 34

Future of Mass. DAL • Research: Need breakthrough research in mathematics, systems, databases, algorithms, sensor networking. • Expand data domains. – Potential partners: Google, NJ auto insurance fraud data, USPTO patent data, AWS location trajectories, etc. • Build state-of-art facility at Rutgers. – Secure, 24 x 7, data hosting and analysis infrastructure capable of gathering and processing petabytes of data/month across domains, data sources, etc. Unique in the world! • Potential. – Every wireless, telecom, internet service provider is looking to farm out this crucial piece of their operations. Estimated market for these services: 100’s of millions in US $ per year. Crucial for NJ State. Interest from multiple VCs now. July 11, 2003 Commission on Jobs Growth and Economic Development 35

Visualizing, Monitoring, and Analyzing Network Data (Statistics Dept. and Avaya Labs) • Communication networks are widespread. • Typical data provides a partial view of flow-data (e. g. , on links) • Analyzing network data is important in: – network planning and design – monitoring flaws – measuring reliability parameters – determining suitability of the network for different transmission functions (voice, data, voice over IP…) July 11, 2003 Commission on Jobs Growth and Economic Development 36

Visualizing, Monitoring, and Analyzing Network Data Challenges • Network data is complex and of high dimensionality. • Statistical methods for analyzing network data are few and far between. • Visualizing data helps us to spot trends quickly. • Need is to develop high quality, practical, statistical and data analytic tools for understanding data from partial views and limited measurements. July 11, 2003 Commission on Jobs Growth and Economic Development 37

Visualizing, Monitoring, and Analyzing Network Data • Results potentially useful for other kinds of networks: transportation, social, … • Such tools of great importance in telecom. • Research in this area already funded through NSF and NSA • New methods/products should be very useful to NJ companies. July 11, 2003 Commission on Jobs Growth and Economic Development 38

Self-Healing Dependable Computing • Computer, Heal Thyself – Scientific American, July 2003 • We need systems that – monitor themselves – adjust hardware and software configurations to match demand – predict and diagnose problems and effect repairs – defend against hacker attacks July 11, 2003 Commission on Jobs Growth and Economic Development 39

Self-Healing Dependable Computing Key Concerns • Susceptibility to attack – Do denial-of-service attacks and viruses cause problem? • Performability – Is system available with adequate performance when needed? • Dependability – Can you rely on correct and predictable behavior? • Self-awareness and autonomy – Does your system monitor and repair itself? • Fail-safe uses – Would you trust your computer with a mission-critical task? July 11, 2003 Commission on Jobs Growth and Economic Development 40

Self-Healing Dependable Computing State of the Art needs to be refocused! • Information Technology is predicated on well-behaved, interacting machines • but spam, viruses, and attacks are epidemic • To combat these problems we are pouring valuable resources into firewalls • however firewalls restrict interaction! July 11, 2003 Commission on Jobs Growth and Economic Development 41

Self-Healing Dependable Computing Solutions • Be realistic about computing environments – Errors, both human and computer, will always be present – Machines are only as well-behaved as their owners – Viruses, spam, and attacks ARE part of the environment • Design systems that are self-aware and self-healing – Hardware is fast enough and affordable – Establish self-administered distributed policies – Continuously monitor, diagnose, and adapt July 11, 2003 Commission on Jobs Growth and Economic Development 42

Self-Healing Dependable Computing: Economic Impact • On the global IT sector – System downtime will become increasingly costly – Without self-healing systems salaries will dominate IT costs • On New Jersey – build on strengths – Two of the six NJ growth clusters are related to IT – NJ is center of telecom industry – NJ has the largest number of scientists/engineers per capita • Experienced workforce is available for new initiatives July 11, 2003 Commission on Jobs Growth and Economic Development 43

Self-Healing Dependable Computing: Rutgers Expertise • Active research in several related areas – autonomous agents, change analysis for OO languages, component-based scalable networks, database mining, distributed systems, fault tolerance, peer-to-peer computing, secure services, modeling and simulation • 7 CS faculty are currently working on relevant research • $3. 5 M in external grants awarded over last few years • Active industrial collaboration – Panasonic (Peer-to-peer computing) – IBM (change analysis for OO languages) – Telcordia and Rutgers CS are developing a joint initiative in this area July 11, 2003 Commission on Jobs Growth and Economic Development 44

Multimodal Human-Machine Interface July 11, 2003 45

Multimodal Human-Machine Interface Real-world trial with NJ National Guard Microphone Array Gaze Tracker Force Feedback Glove July 11, 2003 46

User interface for interaction and collaboration with robots and humans NSF Equipment Grant EIA#98 -18313 Center for Advanced Information Processing, Rutgers University, Piscataway, NJ 08854. July 11, 2003 PI: J. L. Flanagan, co-PIs: J. Wilder, I. Marsic, M. Krane 47

Portable Interactive Command Console (PICC) Internet HQ/VEHICLE Loudspeakers Flatpanel display Source locator microphone Gaze tracker Stereo face tracking cameras Steerable microphone array element Light pen Sensors July 11, 2003 Robotic Vehicles Emergency Responder in the 48 Field

Pervasive and Autonomous Computing Win. LAB, ECE, CS July 11, 2003 Commission on Jobs Growth and Economic Development 49

Pervasive and Autonomous Computing Win. LAB, ECE, CS • Communication and computing cost and performance have been improving by 2 x every 18 months or less for decades. • Wireless now makes it possible to complete the last link to people, machines, sensors, etc. , everywhere. • Great opportunity (and challenge) to move from point-to-point communication to pervasive communication, computing and knowledge access. July 11, 2003 Commission on Jobs Growth and Economic Development 50

Pervasive and Autonomous Computing • Computing and communication can be integrated in the environment – Knowledge, information, communication always available, but less obtrusive • Your personal “Radar O’Riley” is there to help, wherever you are (and gone when you want privacy) • Sensors bring realtime data that matters – From your heartbeat to traffic jams and afternoon weather • The world’s knowledge is always available when needed. July 11, 2003 Commission on Jobs Growth and Economic Development 51

Pervasive and Autonomous Computing • Massive (and interesting) Research Challenges – – – – Flexible system integration New approaches to networking at all levels Information-centric parallel and grid computing Energy efficiency at all levels Context awareness for communication and applications Location awareness in routing and computing Effective and user friendly security at all levels • Integrating of Computing and Communication (especially wireless) is already a major corporate thrust at Intel, Microsoft, IBM, and many others. July 11, 2003 Commission on Jobs Growth and Economic Development 52

Pervasive and Autonomous Computing • Rutgers has expertise and ongoing programs in these areas. • Communications and computing affect every aspect of the economy and every individual • Recent events show the limitations of the existing models for Telecom. NJ could take the lead in changing the landscape. New Jersey has the right combination of people, expertise, facilities to make it happen. July 11, 2003 Commission on Jobs Growth and Economic Development 53

Trusted Computing/Authentication Rutgers – Camden (CS) • Security is the fastest growing sector of the telecommunications market today • Security involves: encryption, authentication, access control, identity management, user provisioning, … • Telecom often involves access to remote resources, requiring authentication of users and monitoring of users’ access privileges July 11, 2003 Commission on Jobs Growth and Economic Development 54

Trusted Computing/Authentication Some Research Themes • Authentication of remote users is usually done by passwords. • Traditional (alphanumeric) passwords are not user-friendly and lead to security problems and increased IT costs. • Graphical passwords: userfriendly; provide an extremely large password space (similar to a cryptographic key space) and thus are inherently more secure. • Human-factors analysis of new password schemes July 11, 2003 Commission on Jobs Growth and Economic Development 55

Trusted Computing/Authentication Impact of Research: • Passwords are the most common method for authentication, but also one of the most vulnerable to cyber as well as physical attack. • Improved authentication will impact human-computer interface, security. • Will allow users to directly use passwords as cryptographic keys • Collaborations: Drexel, Brooklyn Poly. , Minnesota • Collaborations: Unisys • Password research is of great interest to software and telecommunications industries. July 11, 2003 Commission on Jobs Growth and Economic Development 56

Trusted Computing/Authentication Other Research Challenges: • Location-aware authentication/provisioning • Dynamically changing access control and inference management • Biometrics July 11, 2003 57

Telecom and Homeland Security • Communication security – wireless security – sharing data – information privacy – identity theft – secure e-commerce • Emergency Communication • Sensor Networks for Bio/Chemical Hazard Monitoring July 11, 2003 Commission on Jobs Growth and Economic Development 58

Telecom and Homeland Security • Rutgers projects in communication security include: – tunable, programmable, adaptive filters for secure communication (Engineering School) – low bit-rate coding of speech signals for secure communications (CAIP) (with Sarnoff) – information privacy (DIMACS) (with HP Labs NJ, Telcordia, AT&T Labs, NEC Labs) – secure e-commerce (CS with Fogbreak Software) • These projects are funded by NSF, DARPA, NJCST July 11, 2003 Commission on Jobs Growth and Economic Development 59

Telecom and Homeland Security • Rutgers projects in emergency communication include: – infostations for rapid wireless communication for first responders (WINLAB) (with Mayflower Radio) – rapid networking at emergency locations (DIMACS with Telcordia) – rapid telecollaboration (CAIP) July 11, 2003 These projects are funded by DARPA Commission on Jobs Growth and Economic Development 60

Telecom and Homeland Security • Rutgers project in sensor networks with application to bio/chemical hazard monitoring: – WINLAB – partnered with Agere, Sarnoff, Semandex, Thomson, J&J, Lucent July 11, 2003 Commission on Jobs Growth and Economic Development 61

Telecom and Homeland Security • Methods used in telecom are coming to be useful in homeland security research. • Provides a great business opportunity for NJ’s telecom industry. • Already, NJ telecom companies are subcontractors to Rutgers federal grants in this area. • Examples are: – surveillance/detection methods – bioterrorism sensor location July 11, 2003 Commission on Jobs Growth and Economic Development 62

Telecom and Homeland Security • Surveillance/detection: – Massive data set methods used in fraud detection, network intrusion detection, etc. are being used in bioterrorist attack detection, emerging disease identification. • DIMACS, $3 M from NSF, ONR, Sloan Foundation, Burroughs Wellcome Fund • cooperating with AT&T, Lucent, Telcordia, Merck, state and local health departments, CDC July 11, 2003 Commission on Jobs Growth and Economic Development anthrax 63

Telecom and Homeland Security • Surveillance/detection: – MDS methods also used in monitoring streams of text messages for “new events” • DIMACS, $1 M from ICMIC (intelligence community) • cooperating with AT&T, Telcordia July 11, 2003 Commission on Jobs Growth and Economic Development 64

Telecom and Homeland Security • Bioterrorism sensor location BASIS bioterroism sensor system July 11, 2003 Commission on Jobs Growth and Economic Development 65

Telecom and Homeland Security • Bioterrorism sensor location – Network design methods useful. – “Equipment placing” algorithms developed for broadband access at Telcordia are candidates for modification for sensor placement problems. – Algorithms developed at Telcordia for placing regenerating equipment in transparent optical networks are also relevant. – Work at DIMACS with partners from AT&T Labs, Telcordia, Industrial Engineering, Environmental and Occupational Health and Safety Institute (joint with UMDNJ), Statistics, CS, and RUTCOR July 11, 2003 Commission on Jobs Growth and Economic Development 66

Telecom and Homeland Security • Thus, homeland security research can put NJ telecom back to work. July 11, 2003 Commission on Jobs Growth and Economic Development 67