CS 6910 Pervasive Computing Section 0 B Opportunistic

CS 6910 – Pervasive Computing Section 0. B: Opportunistic Networks: Specialized Ad Hoc Networks for Emergency Response Applications Dr. Leszek Lilien Wi. Se Lab (Wireless Sensornet Laboratory) http: //www. cs. wmich. edu/wsn Department of Computer Science Western Michigan University Slides are © 2007 by Leszek T. Lilien Requests to use original slides for non-profit purposes will be gladly granted upon a written request.

Outline n n n Part 1. Motivation for Specialized Ad Hoc Networks Part 2. Analogy to a Human Emergency Response Team Part 3. Opportunistic Networks: A New Type of Specialized Ad Hoc Networks n Part 4. Related Research and Research Challenges n Part 5. Conclusions n Part 6. Current and Future Work 2

Part 1. Motivation for Specialized Ad Hoc Networks n n Homeland Security n One of the crucial challenges facing the USA today n Among its 6 mission areas is Emergency Preparedness and Response (EPR) EPR deals with: n Natural disasters n Man-made disasters (incl. accidents, terrorist attacks) [Natl. Strategy for Homeland Security, July 2002] n Mobile ad hoc networks (MANETs) proposed for EPR [Haas, 1999] 3

Motivation for Specialized Ad Hoc Networks (2) n n MANETS are not quite a natural match for EPR n E. g. , human rescue teams can find and mobilize as their helpers local firemen, police, National Guard or even regular citizens n No analogous capability of MANETs to find and “mobilize” devices/networks Let’s find or define a specialization (a subclass) of ad hoc networks more suitable for EPR applications n A new paradigm and a new technology to improve effectiveness & efficiency of EPR 4

This page left blank intentionally. 5

Part 2. Analogy to a Human Emergency Response Team n Important characteristics of a human rescue team n Starts as a seed team (a small group of people) n The seed team grows opportunistically during its operations a) Some people can always be ordered to join § Police, firemen, National Guard or Army Reserve personnel b) Anybody can be ordered to join only in life-or -death situations § Legally required to help saving lives or critical resources c) Anybody can be asked to join in other situations 6

Analogy to a Human Emergency Response Team (2) n n Who is ordered or invited to join? n In some situations any extra pair of hands can help n In others only highly qualified people (e. g. , doctors and nurses) are ordered/asked to join Human helper types n Highly prepared n n Somewhat prepared n n E. g. , National Guard, Army Reserve or state militias n Signed up for service n Undergo training in preparation for duty (even unforeseen) E. g. , people that volunteer for first aid courses n Become pretty valuable helpers in emergencies Not prepared at all 7

Analogy to a Human Emergency Response Team (3) n Benefits of the opportunistic growth of the rescue team n Opportunistic leveraging of all kinds of skills and resources that new helpers can bring n Obtaining a lot of help effectively and efficiently – even for free 8

Analogy to a Human Emergency Response Team (4) Analogous critical requirements for ad hoc network specialized for EPR in the priority order: 1) Minimal starting configuration – a pre-configured seed for EPR operations n Analogy to the seed rescue team 2) High interoperability in terms of communication of diverse devices or networks (Wired/Wi. Fi, Bluetooth, satellite, ham radios, Wi. MAX, . . . ) Analogy to a rescue team’s ability to contact different people, individually or via organizations 3) Highly heterogeneous software (& hardware) Analogy to heterogeneity of rescue teams in terms of members’ skills, communication and other equipment, and other resources 4) Harvesting of diverse resources as needed Analogy to finding people with different skills, equipment, and other resources 5) Persistent connectivity once it is established Analogy to being able to contact (maybe via a chain of others) members of the expanded team, including all helpers 9

Analogy to a Human Emergency Response Team (5) n n If there is no specialized ad hoc network (AHN) matching the requirements, then: n Define a new specialized AHN paradigm n Invent a new specialized AHN technology No known specialized AHN matches the requirements Considered AHNs: n Mobile ad hoc networks (MANETs) n Mesh networks n P 2 P systems n Sensor networks n Spontaneous networks (in the narrow sense, cf. [Feeney et al. 2001]) => Need: - a new specialized AHN paradigm - a new specialized AHN technology Note: Other (than oppnets) specialized AHNs for emergencies are Incident Area Networks (IANs) 10

This page left blank intentionally. 11

Part 3. Opportunistic Networks: A New Type of Specialized Ad Hoc Networks n Paradigm of opportunistic networks (oppnets) n n Based on the analogy to operations of human rescue teams Goals of oppnets: n Opportunistic growth n Opportunistic leveraging of resources aiming at (among others): n Bridging diverse communication media n Offloading computations to additional platforms n Integrating independent sensing systems (enhancing their sensing capabilities) 12

Startup: Seed Oppnet n Oppnet deployed as a seed oppnet Link to the World Seed Nodes Controller (distributed) n Seed oppnet localizes its nodes & self-configures 13

Growth: Expanded Oppnet n Seed oppnet grows into an expanded oppnet by: n n Finding candidate helpers Selecting candidates - they are ordered/asked to join n Add communication, computing, sensing, storage, other resources Satellite Link to the World Appliance Cellphone Tower (refrigerator) Seed Nodes Controller (distributed) Wi. MAX Overturned Vehicle (with On. Star, VAN) Computer Network 1414

Summary of Oppnet Activities n Discovering & identifying candidate helpers n Contacting selected candidates n Inviting or ordering candidates to join n Admitting & integrating helpers that join oppnet n Offloading tasks to helpers n Determining useful colaborative functionalities n Managing offloaded tasks n Cleaning up & releasing each helper when no longer needed n > Need research for each of theses activities < n Think which is most interesting to you as a topic for your CS 6910 Team Project 15

Oppnet Reserve n n A challenge in oppnet growth n Must discover candidate helpers, then contact selected ones n Difficult without facilities provided by candidates or infrastructure A solution: oppnet reserve — facilitating discovery/contacting n Analogy to Army / Navy / Air Force Reserve, etc. n Volunteer helpers sign up for oppnet reserve n Maybe for some incentives (moral, financial, etc. ) n Volunteers „trained” for active duty n Install facilities that make them easier to detect and contact by oppnets n n E. g. , install (future) standard oppnet protocols n Matched to their capabilities: heavy-, medium- and lightweight Available for active oppnet duty whenever must/can help 16

Oppnet Reserve (2) n Oppnet reserve is analogous to having highly prepared human helpers (Recall the classification of human helpers into: highly prepared, somewhat prepared, unprepared) n Oppnet reserve is not required but very helpful n Having highly prepared human helpers is not required either 17

What Oppnet Is … / What Oppnet Is Not … (this slide added after lecture) § Oppnet is … n … a specialized (application-specific or application-classspecific) ad hoc network n … an opportunistic growth network n … an opportunistic use of resources network / an opportunistic use of services network n n § Spectrum, connectivity, services, … … a privacy-preserving & secure technology Oppnet is not … n … a “generic” ad hoc network n … a mesh network n … a grid computing system n … a peer-to-peer (P 2 P) system n … a hybrid network n … 18

Basic Oppnet Categories n n 2 major oppnet categories: n Benevolent oppnets n Malevolent oppnets Corresponding oppnets scenarios: n Benevolent oppnet scenario: „Citizens Called to Arms” n Malevolent oppnet scenario: „Bad Guys Gang Up” 19

Benevolent Oppnet Scenario: „Citizens Called to Arms” (1) n Seed oppnet deployed after an earthquake (unpredictable emergency) n Seed is ad hoc wireless network with very powerful nodes n n 20 More energy, computing and communication resources Seed tries to detect candidate helpers n For help in damage assessment and disaster recovery n Uses any available detection method — including: n Cellphone- or radio-based detection n Including Software Defined Radio (SDR) & Cognitive Radio Searching for nodes using the IP address range for the affected geographic area AI-based visual detection (next slide) 20

Benevolent Oppnet Scenario: „Citizens Called to Arms” (2) n Example: n Helper 1 (supervisor of security cameras) monito- n n n 21 ring a surveillance net views an overturned car Helper 2 (pattern recognition specialist) detects an overturned car Helper 3 (image analysis specialist) asked to recognize car’s license plate Helper 4 (DB manager) finds that the car has On. Star link Helper 5 (On. Star agent) contacts VAN (Vehicle Area Network) and BANs (Body Area Networks) on or within bodies of car occupants via On. Star infrastructure Helper 6 (vital sign evaluator) evaluates obtained info Helper 7 (rescue dispatcher) decides if/when rescuers should be dispatched 21

Benevolent Oppnet Scenario: „Citizens Called to Arms” (3) 22 n n Oppnet selects optimal subset of detected nodes n Inviting devices, clusters & entire networks n Helpers for communicating, sensing, computing Using „hidden” capabilities, e. g. for sensing: n Desktop can „sense” presence of a potential victim at its keyboard n Cellphones can „sense” location n Even ones w/o GPS can be triangulated 22

Benevolent Oppnet Scenario: „Citizens Called to Arms” (4) n Using „hidden” emergency functionalities n Oppnet contacts 2 independent sensornets (SNs): water infrastructure control SN / public space surveillance SN n SNs ordered to abandon normal functions & help in rescue & recovery operations n Water infrastructure SN (with multisensor capabilities, under road surfaces) — ordered to sense vehicular movement and traffic jams n Public space surveillance SN — ordered to search for images of human victims 23 23

Malevolent Oppnet Scenario: „Bad Guys Gang Up” (1) n Scenario 1 — Terrorists create apparently 24 harmless weather monitoring sensornet (SN): n SN becomes a seed of a malevolent opportunistic SN n SN exploits other nodes from many other networks (w/o revealing its true goals) n n “Critical mass” of the opportunistic SN is reached geographical spread and sensing capabilities) (in terms of SN waits for wind patterns that can speed up spread of poisonous chemicals n Collected data used to decide when to start chemical attack 24

Malevolent Oppnet Scenario: „Bad Guys Gang Up” (2) n Scenario 2 — network at home starts spying on you: n Becomes a seed oppnet n Exploits other devices/nets to collect all info on you: n n n n n From your fridge (& RFID-equipped food packaging): what/when you eat From your computer: keylogs your passwords, sensitive data From your cellphone: who you call & when From your networked camera: what photos you take From your home security surveillance system: your private images Cyberfly with camera eyes and microphone ears. . . Huge privacy problem! / Huge security problem! Controls to counteract malevolent oppnets badly needed 25

NOTE: The following slides not required for exam. (Provided as a potential help for project selection and project work. ) 26

Part 4. Related Research & Research Challenges n Examples of Related Research n Interoperability n Among wireless networks - active research area n n Among WANs, MANs, LANs, PANs (Personal Area Networks), etc. Among wired & wireless nets - much less research n Ambient Networks (big European Union project, next-generation Internet—for 2015/2020, smaller networks able to compose themselves into bigger ones) n Localization & self-organization n n Network growth n n n P 2 P systems – search for peers in unstructured systems Trojan Horses - agents spreading in search for helpers Integrating and managing heterogeneous systems, incl. data integration & aggregation n n MANETs / Sensornets Grid systems / MANETs / Sensornets Other n … 27

Research Challenges in Basic Operations n Challenges in seed oppnet deployment n n E. g. , localization, self-configuration, adaptability Challenges in detecting helper systems n E. g. , define primitives to detect candidates, identify and categorize them, evaluate and classify them (e. g. , based on dependability and usefulness) n Challenges in inviting & admitting candidate helpers n n E. g. , select candidates to invite, develop protocols for candidates to accept or reject invitation, devise primitives /methods to manage expanded oppnets Etc. , etc. for remaining oppnet primitives More: Leszek Lilien, Z. Huma Kamal, and Ajay Gupta, "Opportunistic Networks: Research Challenges in Specializing the P 2 P Paradigm, ” Proc. 3 rd International Workshop on P 2 P Data Management, Security and Trust (PDMST’ 06), Kraków, Poland, September 2006. 28

Research Challenges in Privacy & Security (1) 1) Privacy challenges in oppnets n Privacy is critical n Oppnets are pervasive systems n Must face all critical privacy challenges inherent to pervasive computing n Privacy is a „make it or break it” issue for pervasive computing => Privacy is a „make it or break it” issue for oppnets n Basic privacy protection goals in oppnets n Protect helper resources from the host oppnet n Protect oppnet from its helpers n Protect environment from privacy violations by oppnet n Also from malevolent oppnets 29

Research Challenges in Privacy & Security (2) 2) Security challenges in oppnets n Many have privacy aspects/components More: L. Lilien, Z. H. Kamal, V. Bhuse, and A. Gupta, "Opportunistic Networks: The Concept and Research Challenges in Privacy and Security, " book chapter in: "Mobile and Wireless Network Security and Privacy, " ed. by K. Makki et al. , Springer Science+Business Media, Norwell, Massachusetts, 2007 (to appear) – pre-publication version available upon request 30

This page left blank intentionally. 31

Part 5. Conclusions n High-payoff potential for the oppnet initiative 1) Social and economic benefits § Including reduction of human suffering & loss of life 2) Technological benefits 3) Research benefits 4) Educational benefits -- Details below -32

Part 5. Conclusions – cont. 2 1) Social & Economic Impacts n Impacts on Emergency Preparedness and Response operations in Homeland Security (HS) — current app focus n Tremendous leveraging potential in emergencies n n n A wealth of freely available resources Reduction of human suffering & loss of life Increasing safety & efficiency of the first responders Impacts on other applications both in HS and outside HS [SKIP: ] Economic impacts n n Technology transfer & commercialization Benefits for the computer industry n n Production of software / networking equipment Benefits for other industries n n Enhancing many products with standard oppnet interfaces E. g. , the auto industry: cars and trucks as oppnet platforms 33

Part 5. Conclusions – cont. 3 2) Impacts on Technology n n n Advancing the network and pervasive computing know-how n Development of the innovative oppnet technology n Enhancing network/pervasive applications by use of oppnet technologies n Enabling new network/pervasive application niches we can not even foresee Advancing other areas of technology (not only computing) n A side effect of oppnet developments [SKIP: ] Technology impacts speeded up & enhanced by the planned technology transfer plus commercialization activities 34

Part 5. Conclusions – cont. 4 [SKIP: ] 3) Impacts on Research [SKIP: ] 4) Impacts on Education 35

This page left blank intentionally. 36

Part 6. Current and Future Work n n >> Search for good uses for oppnets in all kinds of application areas << Search for applications that. . . n. . . can start with a seed n. . . need/have high interoperability n. . . need/have highly heterogeneous software n. . . can benefit from leveraging diverse resources of helpers n. . . can maintain persistent connectivity once it is established n. . . 37

Part 6. Current and Future Work – cont. n Building an oppnet prototype n Goal: Proof of concept n n To demonstrate technical prowess & economic benefits Designing oppnet architecture n With its associated components: n n Methods, protocols, and algorithms Oppnet prototype implementation n For stimulation and feedback n Necessary for fine-tuning oppnet design 38

Publications, Etc. n Selected Publications (cf. http: //www. cs. wmich. edu/wsn/project_oppnet. html) L. Lilien, A. Gupta, and Z. Yang, "Opportunistic Networks for Emergency Applications and Their Standard Implementation Framework, " Proc. The First International Workshop on Next Generation Networks for First Responders and Critical Infrastructure (Net. Cri 07), New Orleans, Louisiana, April 11 -13, 2007 (to appear) L. Lilien, Z. H. Kamal, V. Bhuse, and A. Gupta, "Opportunistic Networks: The Concept and Research Challenges in Privacy and Security, " book chapter in: "Mobile and Wireless Network Security and Privacy, " ed. by K. Makki et al. , Springer Science+Business Media, Norwell, Massachusetts, 2007 (extended version of the WSPWN 2006 paper; to appear) L. Lilien, Z. H. Kamal and A. Gupta, "Opportunistic Networks: Research Challenges in Specializing the P 2 P Paradigm, " Proc. 3 rd International Workshop on P 2 P Data Management, Security and Trust (PDMST'06), Kraków, Poland, September 4 -8, 2006, pp. 722 -726. L. Lilien, Z. H. Kamal, V. Bhuse and A. Gupta, "Opportunistic Networks: The Concept and Research Challenges in Privacy and Security, " Proc. International Workshop on Research Challenges in Security and Privacy for Mobile and Wireless Networks (WSPWN 2006), Miami, Florida, March 2006, pp. 134 -147. B. Bhargava, L. Lilien, A. Rosenthal and M. Winslett, "Pervasive Trust, " IEEE Intelligent Systems, vol. 19(5), Sep. /Oct. 2004, pp. 74 -77 (just the first brief mention of the oppnet idea, in the form of malevolent opportunistic sensor networks). 39

Publications, Etc. – cont. 1 n Selected Presentations L. Lilien, "Opportunistic Networks: Research Challenges in Specializing the P 2 P Paradigm, " 3 rd International Workshop on P 2 P Data Management, Security and Trust (PDMST'06), Kraków, Poland, September 4 -8, 2006 L. Lilien, "Opportunistic Networks: Specialized Ad Hoc Networks for Emergency Response Applications, " presented for Distributed Systems Research Group, Department of Computer Science, AGH University of Science and Technology, Krakow, Poland, May 22, 2006 L. Lilien, "Opportunistic Networks: Specialized Ad Hoc Networks for Emergency Response Applications, " presented for Section of Information Technology, Institute of Telecomputing, Cracow University of Technology, Kraków, Poland, May 23, 2006 L. Lilien, "Developing Specialized Ad Hoc Networks: The Case of Opportunistic Networks, " Workshop on Distributed Systems and Networks (in conjunction with WWIC’ 06), Bern, Switzerland, May 9, 2006 L. Lilien, "Opportunistic Networks: The Concept and Research Challenges in Privacy and Security, " International Workshop on Research Challenges in Security and Privacy for Mobile and Wireless Networks (WSPWN 2006), Miami, Florida, March 15 -16, 2006 L. Lilien, Z. H. Kamal and A. Gupta (in cooperation with V. Bhuse and Z Yang), "Opportunistic Networks: The Concept and Research Challenges, " Department of Computer Science, Western Michigan University, Kalamazoo, Michigan, February 9, 2006 40

Publications, Etc. – cont. 2 n Selected Posters Leszek Lilien, Zille Huma Kamal, A. Gupta, V. Bhuse and Z. Yang, "Opportunistic Networks, " 3 rd International Conference on Networked Sensing Systems, Chicago, IL, May 31 - June 2 2006. Leszek Lilien, Zille Huma Kamal, Vijay Bhuse and Ajay Gupta, "Opportunistic Networks and Their Privacy and Security Challenges, " The Seventh Annual CERIAS Information Security Symposium - "Negotiating Trust: Security, Privacy, Risk, " CERIAS, Purdue University, West Lafayette, IN, March 21, 2006. Leszek Lilien and Ajay Gupta, "Opportunistic Networks for Emergency Preparedness and Response, " 4 th Annual e-Enterprise Conference, PHSI/RCHE, Purdue University, West Lafayette, IN, March 20, 2006. 41

Wi. Se Lab Experience in Ad Hoc Networks – Selected Projects Since 1/03 § § § § Designing of Wi. Se Security Protocols: DSPS Location Tracker Using Motes (*) RHS: Remote Home Surveillance (*) Directed Diffusion: Attacks & Countermeasures Improving the Accuracy of Mote Measurements by Using Neural Networks SOMS: Smart Occupancy Monitoring System Using Motes (*) Comparative Study of Network Simulators Collaborative Image Processing (*) DENSe: a Development Environment for Networked Sensors Incorporating Mobile-ware in Distributed Computations / Grids (*) Extending the ns-2 Simulator to Satellite and WCN Simulations Smart Antennas for WCNs Energy Efficient MAC Protocols for IEEE 802. 11 x A Wireless Security Testing System (*) Mobile and Self-Calibrating Irrigation System Collective Communications for Sensornets (*) * Results directly useful for oppnets 42

This page left blank intentionally. 43