1. Protocol Engineering Research Center P. M. Melliar-Smith U.C. Santa Barbara M. G. Baker Stanford J. J. Garcia-Luna-Aceves U.C. Santa Cruz J. C. Hou Ohio State K. N. Levitt U.C. Davis L. E. Moser U.C. Santa Barbara D. C. Schmidt U.C. Irvine T. Suda U.C. Irvine

2. The Vision Information and Computation without Boundaries What Matters to the Users – Functionality – Ilities Real-time Performance Reliability Security What Should Be Hidden Engineering – Physical Location – Mobility – Networks and Protocols

3. The Vision Transparency and Ease of Use Major Applications are the most complex things ever built by humans The Network and the Protocols should make applications easier to design and build Application designers should express their requirements for Quality of Service Automated implementation of QoS requirements Predictable network performance

4. Objectives Application Systems Built from Components – The Network is the glue that holds components together Real-Time across Multiple Platforms Mobile Wireless Networks – Full Capabilities – Full Performance – Disconnected Operation Security – Rigorous – Unobtrusive – External intruders and internal subversion

5. Objectives Methodologies allow application designers to express their Quality of Service requirements Networks and Protocols are efficient/predictable over a wide range of topologies and loads Tools help application designers to assess and predict the behaviors of their designs Analysis and Measurement techniques determine whether systems are behaving as they should

6. Synergy in Real Time Doug Schmidt – ACE real-time network executive – Mapping Quality of Service onto the implementation Jennifer Hou – Medusa - meeting QoS requirements Multicasting Routing Flow control and Congestion Michael Melliar-Smith and Louise Moser – Probabilistic analysis of real-time delays J.J. Garcia-Luna-Aceves – Real-time wireless protocols

7. Synergy in Fault Tolerance Louise Moser and Michael Melliar-Smith – Fault-tolerant multicast protocols – Fault tolerance for Java and CORBA (Eternal) – Recovery in real-time – Network resource allocation for fault tolerance Jennifer Hou – Fault-tolerant multicast protocols – Fault-tolerant routing J.J. Garcia-Luna-Aceves and Mary Baker – Fault-tolerant wireless networks

8. Synergy in Security Karl Levitt – Models of attacks – Intrusion detection strategies for detecting attacks and subverted nodes – Response and defense strategies Mary Baker – Routing that exploits trust developed over time – Mechanisms for detecting malicious behavior Louise Moser – Reliable operation in networks that contain subverted nodes

9. Synergy in Multimedia Tatsuya Suda – Scalable multimedia multicasts – Adaptive layered encoding – QoS feedback, with aggregation within the network Jennifer Hou – Multimedia in medical applications J.J. Garcia-Luna-Aceves – Wireless multimedia Michael Melliar-Smith – Least-laxity scheduling for jitter management

10. Synergy in Wireless J.J. Garcia-Luna-Aceves and Mary Baker – Quality of Service at the link level – Quality of Service over multiple paths – Aggregation of flows – Distribution of aggregated flows – Integration of routing and reservations – Transmission of priority packets over multiple paths – CAMP mesh routing Michael Melliar-Smith – Hierarchical clustering for scalable routing in large ad-hoc networks Jennifer Hou – Mobile IP for multicasting

11. ARCH Clustered Routing Protocol Michael Melliar-Smith Hierarchical Clustered Routing allows ad-hoc networks to scale to thousands of nodes while still retaining reasonable overheads

12. Synergy in BioNetworking Tatsuya Suda – Cyber-entities – Hierarchical aggregation into Super-entities – Scalability, adaptability, evolution – Coordination and efficiency Everybody – How can our protocols become Cyber-entities? – How can our strategies be used in BioNetworking?

13. Tools: NetSim^Q Jennifer Hou Java-based network simulation tool Exploits object-oriented programming – Instantiate generic network class – Redefine attributes and methods Trace-driven simulation models real traffic Can be used for – Validation of protocols – Investigation of tradeoffs – Fine tuning

14. Tools: Testing Workbench Karl Levitt Protocol specification in Estelle To reduce the testing burden – Select a component of the specification – Slice the implementation to contain only statements that relate to the specification component – Test the slice against the specification Specifications can also be sliced with respect to a higher level specification Test data can be generated for the slice – Identify code that is vulnerable to race conditions – Generate test cases to explore those vulnerabilities

15. Tools: Verification Workbench Karl Levitt Based on HOL for extensibility and proof checking and PVS for proof power Develop HOL theories for – Refinement – Composition – Liveness and Feedom from Deadlock – Quality of Service – Faults and Attacks Will still require substantial human involvement

16. Tools: Performance Analysis Michael Melliar-Smith and Louise Moser Performance Analyses based on mean durations are not useful for real-time Analyses based on worst-case durations are too conservative We need to estimate the tail of the distribution of durations Analysis is based on pdfs for durations – combined by convolution – allowance for correlations

17. Tools: Performance Analysis Michael Melliar-Smith and Louise Moser

18. Working Together Collaborative research – By professors – By students – Many interesting interactions among our research Weekly seminar series – Conducted across the Internet – Professor or student presents a research project – Discussion of relationship to other research Experimental testbed network based on CALREN-2