Publications

Parallel, Distributed and Fault Tolerant Computing Dependability Modeling

My main area of Research from 1982-1990 was in dependability modeling. The key thrust was to create a capability to construct and solve dependability models of realistic systems. As a result of this effort, significant new model solution techniques were invented, including importance sampling based simulation. All the model construction and solution methods were embodied in a package called System Availability Estimator (SAVE) which is being used by system designers across the industry and has been used as a teaching tool in universities.

Journal Publications

1. A. Goyal, S. S. Lavenberg and K. S. Trivedi, Probabilistic Modeling of Computer System Availability, Annals of Operations Research, vol 8, 1987. S. Albin and C.M. Harris (eds.).
2. A. Goyal and A. N. Tantawi, Evaluation of Performability for Degradeable Computer systems, IEEE Transactions on Computers, C-36, 6, June 1987, pp. 738-744.
3. A. Goyal, V. F. Nicola, A. N. Tantawi and K. S. Trivedi, Reliability of Systems with Limited Repairs, IEEE Transactions on Reliability, Vol. 36, No. 2, June 1987, pp. 202-207.
4. A. Goyal and S. S. Lavenberg, Modeling and Analysis of Computer System Availability, IBM Journal of Research and Development, 31, Nov. 1987, pp. 651-664.
5. A. Goyal and A. N. Tantawi, A Measure of Guaranteed Availability and its Numerical Evaluation, IEEE Transactions on Computers, C-37, 1, Jan 1988, pp. 25-32.
6. A. Goyal, P. Heidelberger, P. Shahabuddin, V. F. Nicola, P. W. Glynn, A Unified Framework for Simulating Markovian Models of Highly Dependable Systems, IEEE Transactions on Computers, 41, No. 1, January 1922, pp. 36-51.
7. V. Nicola and A. Goyal, "Modeling of Correlated Failures and Community Error Recovery in Multiversion Software," IEEE Transaction on Software Engineering, Vol. SE-16, No. 3, pp. 350-359, March 1990.
8. V. F. Nicola, M. K. Nakayama, P. Heidelberger and A. Goyal, Fast Simulation of Highly Dependable Systems with General Failure and Repair Processes, IEEE Transactions on Computers, 42, No. 8, 1993, pp. 1440-1452.
9. M. Nakayama, A. Goyal and P. Glynn, Likelihood Ratio Sensitivity Analysis for Markovian Models of Highly Dependable Systems, Operations Research, 42, No. 1, 1994, pp. 137-157.

High performance and Highly Available Systems

The focus was on high performance and highly available systems for commercial applications - e.g. transaction processing, data-intensive queries, file servers, image and document servers. The goal is to study how these applications can be partitioned to run on parallel machines to provide high availability, modular growth in throughput and capacity, as well as response time reductions in data-intensive applications.

Conference Publications

1. F. Jahanian, A. Goyal, A Formalism for Monitoring Real-Time Constraints at Run-Time, Fault-Tolerant Computing: The Twentieth International Symposium, Newcastle, UK, 1990.

Research Reports

1. A. Bhide, H. Hsiao, A. Jhingran, A. Goyal, Asynchronous Replica Management in Shared Nothing Architectures, IBM Research Report RC 16403, Dec 1990.

Communication Networks

The main focus was to improve the performance and reliability of a token ring network being developed by IBM. Current emphasis is on the architecture and design of an extension of the Token Ring network for optical fiber applications. Other interests have been in the area of voice-data integration and switching. Related publications, disclosure and invited talks are listed below.

Journal Publications

1. A. Goyal, Lookahead Networks, IEEE Transactions on Communications, Nov 1985.

Conference Publications

1. A. Goyal and G.J. Lipovski, Scheduling on a Light Pipe Simplex Ring, 1st Int. Conf. on Dist. Comp. Systems, Huntsville, Alabama, 1979.
2. A. Goyal and G.J. Lipovski, Reconfigurable Hierarchical Rings, Dist. Data Acquisition, Computing, and Control Symposium, Miami Beach, Florida, Dec. 1980.
3. G.J. Lipovski, A. Goyal and M. Malek, Lookahead Networks, National Computer Conference, June 1982.
4. A. Goyal, G.J. Lipovski and M. Malek, Reliability in Ring Networks, CompCon 1982, Washington, DC.
5. D.M. Dias, A. Goyal and M. Kumar, A Dynamically Reconfigurable Switch for Integrated Communications, Proc. Local Communications Systems: LAN and PBX. Toulouse, France 1986 Nov. North-Holland edited by J.P. Cabanel, G. Pujolle, and A. Danthine.
6. A. Goyal and D.M. Dias, Performance of Priority Protocols for High Speed Token Ring Networks, Proc. 3rd Int. Conf. on Data Commun. Sys. and their Perf., Rio de Janeiro, Brazil, June 1987.
7. D. Dias and A. Goyal, A High Speed Token Ring Network, Invited paper in EFOC/LAN88 - European Fiber Optic Conference, Amsterdam, pp.403-407, June 1988.
8. A. Goyal and D. Dias, Correctness and Access Time Properties of an Extended Token Ring Protocol, 9th Int. Conference on Computer Communication, Tel Aviv, Nov. 1988.

Computer Architecture

The main work has been in studying the effect of technology on the computer storage hierarchies, cache organization, machine organization and the instruction set architectures.

Journal Publications

1. A. Goyal and G.J. Lipovski, Light Pipe Implementation of Banyan Networks, Journal of Digital Systems, vol. VI, no. 4, Winter 1982.
2. A. Goyal and T.K.M. Agerwala, Performance Analysis of Future Shared Storage Systems, IBM Journal of Research and Development, Jan. 1984.

Conference Publications

1. W.A. Dees, K.M. Parmar, A. Goyal R.Y. Tsui, B.D. Rathi, and R.J. Smith II, A Computer-Aided VLSI Layout System, National Computer Conference, Chicago, Illinois, May 1981.