Venkita Subramonian

 

I graduated with my doctoral degree in May 2006 and joined AT&T Labs Research, Florham Park, NJ in July 2006. I will update this site with my new website info shortly.

 

Research Interests
Select Publications
Research Projects
Brief Bio
Personal

Research Interests

  • Formal modeling and model checking of middleware
  • QoS-enabled component middleware
  • Composition of scheduling and dispatching mechanisms
  • Middleware customization and configuration
  • Generative programming techniques
  • Formal modeling of Web Services composition

Select Publications

1.      Venkita Subramonian, Christopher Gill, Cesar Sanchez and Henny Sipma, Composable Models for Timing and Liveness Analysis in Distributed Real-Time Embedded Systems, submitted to 26th International Conference on Distributed Computing Systems (ICDCS), July 2006.

2.      Venkita Subramonian and Christopher Gill, "Middleware Design and Implementation for Networked Embedded Systems", Embedded Systems Handbook (Richard Zurawski, ed.), CRC Press, Florida, 2005, Chapter 30, pp. 1-17.

3.      Cesar Sanchez, Henny Sipma, Venkita Subramonian and Christopher Gill, Thread Allocation Protocols for Distributed Real-Time and Embedded Systems, 25th IFIP WG 6.1 International Conference on Formal Techniques for Networked and Distributed Systems, Taipei, Taiwan, October 2-5, 2005.

4.      Tejasvi Aswathanarayana, Venkita Subramonian, Douglas Niehaus and Christopher Gill,  Design and Performance of Configurable Endsystem Scheduling Mechanisms, 11th IEEE Real-Time and Embedded Technology and Applications Symposium, March 7-10, 2005, SanFransisco, USA.

5.      Venkita Subramonian, Liang-Jui Shen, Christopher Gill and Nanbor Wang, The Design and Performance of Dynamic and Static Configuration Mechanisms in Component Middleware for Distributed Real-Time and Embedded Systems, 25th IEEE International Real-Time Systems Symposium, December 5-8, 2004, Lisbon, Portugal.

6.      Nanbor Wang, Chris Gill, Douglas C. Schmidt and Venkita Subramonian, Configuring Real-time Aspects in Component Middleware, Distributed Objects and Applications, Agia Napa, Cyprus, Oct 25-29, 2004.

7.      Venkita Subramonian, Guoliang Xing, Christopher Gill, Chenyang Lu and Ron Cytron, Middleware Specialization for Memory-Constrained Networked Embedded Systems, 9th IEEE Real-Time and Embedded Technology and Applications Symposium, May 25-28, 2004, Toronto, Canada.

8.      Xiaorui Wang, Huang-Ming Huang, Venkita Subramonian, Chenyang Lu and Christopher Gill, CAMRIT: Control-based Adaptive Middleware for Real-time Image Transmission, 9th IEEE Real-Time and Embedded Technology and Applications Symposium, May 25-28, 2004, Toronto, Canada.

9.      Venkita Subramonian and Christopher Gill, A Generative Programming Framework for Adaptive Middleware, Hawai'i International Conference on System Sciences, January 5 – 8, 2004, Big Island, Hawaii (awarded best paper in the Software Technology Track).

Complete list of my publications is available here.

Research Projects

 

Formal Modeling of Middleware
This research is work in progress and is funded by NSF. As part of this research, I am developing timed automata models of canonical lower-level middleware building blocks that have been used to implement a wide range of middleware frameworks. These timed automata models can then be combined with higher-level formal models to provide a faithful model of a system including the middleware platform on which the system is deployed, such that the composite models can be verified for correctness with higher fidelity to the system itself. These models are executable models and can be model checked for constraint violations, at which time a trace will be produced by the model checker as to what sequence of events led to the constraint violation. This research has produced a modeling architecture for modeling middleware infrastructure elements. We have realized this architecture using the UPPAAL and IF toolkit. We have also identified a number of state optimization techniques that make model checking more tractable.

Papers:
Composable Models for Timing and Liveness Analysis in Distributed Real-Time Embedded Systems
Thread Allocation Protocols for Distributed Real-Time and Embedded Systems
A Generative Programming Framework for Adaptive Middleware

Software:

The models described in the papers can be obtained from here. You can obtain the version (based on IFx2.0) of IF-toolkit that I use from here and the version of UPPAAL (3.6 alpha) that I use can be downloaded from here.

Static Configuration Techniques for QoS-enabled Component Middleware

This research was funded by the DARPA Program Composition for Embedded Systems(PCES) program. In this research we considered a  range of issues relevant to component middleware, e.g., static vs. dynamic linking/loading, configuration parsing and component assembly. This research produced a static configuration framework for the Component Integrated ACE ORB (CIAO) which is an implementation of the lightweight CORBA Component Model (CCM). This framework has been integrated into the Deployment and Configuration Engine (DAnCE) for CIAO. Performance comparisons were done between dynamic and static component deployment and configuration mechanisms in CIAO. The resulting performance profiles and analysis help guide DRE system developers in choosing which component deployment and configuration mechanisms to use for particular DRE systems. An empirical case study was done that compares configuration mechanisms in CIAO vs. PRISM, which is an avionics domain-specific component model developed by Boeing. This case study aims at helping guide developers of DRE systems in making trade-offs in performance and flexibility when applying standards-based vs. domain-customized component deployment and configuration solutions. This research was done in collaboration with Dr. Doug Schmidt's group at Vanderbilt University.

 

Papers:

Configuring Real-time Aspects in Component Middleware

The Design and Performance of Dynamic and Static Configuration Mechanisms in Component Middleware for Distributed Real-Time and Embedded Systems

The Design and Performance of Component Middleware for QoS-enabled Deployment and Configuration of DRE Systems

Software:

The static configurator is available as part of CIAO.

Scheduling and Dispatching Mechanisms for Computation Control
This research was funded by the DARPA Program Composition for Embedded Systems (PCES) program. Part of this research was conducted in collaboration with Dr. Doug Niehaus' team at University of Kansas, Lawrence, Kansas. In this research, we developed middleware level thread scheduling and dispatching mechanisms as part of the Kokyu framework. We also investigated the effectiveness of an alternative programming model for scheduling and dispatching computations called Group Scheduling (GS) developed by Dr. Niehaus. Group Scheduling  provides an extremely flexible model that can be used to express a wide range of application and endsystem scheduling semantics. The GS model emphasizes representation of the groups of computation components comprising an application. Furthermore, it recognizes the diversity of scheduling semantics by permitting each group to use the scheduling algorithm most appropriate to the set of components it controls. The DRE system designer then constructs a scheduler for the system as a whole through hierarchic composition of groups. My main contributions to this research are - (1) Design and development of a middleware level group scheduler (2) Design and development of a test harness that uses multiple computation pipelines with specific application semantics that are realistic, but difficult to express directly through a priority-based programming model (3) Design and development of a progress based scheduling decision function (4) Design and development of an elaborate set of post-processing tools for conducting empirical analysis of experimental results (5) Design and development of middleware for scheduling and dispatching distributable threads.

Papers:
Group Scheduling in Systems Software
Design and Performance of Configurable Endsystem Scheduling Mechanisms

Software:
The group scheduling framework is available as part of the LibertOS Linux distribution. The test suite has been developed using ACE and TAO and more information on building the test harness is available here.


Middleware Specialization for Networked Embedded Systems
This research was conducted in the context of the Boeing Open Experimental Platform (OEP) as part of the DARPA Networked Embedded Software Technology (NEST) program. The objective of this research was to develop specialized real-time middleware for the Boeing OEP - a next-generation aerospace application, in which a number of MEMS sensor/actuator nodes are mounted on a surface of a physical structure such as an aircraft wing. These MEMS devices are networked together to form an information system that is used to detect damage of the structure during operation. This research produced a special purpose distributed real-time middleware called nORB that addressed the following key challenges in the Boeing OEP - (1) Reuse existing infrastructure to the extent possible (2) Provide real-time assurances (3) Provide a robust DOC middleware (4) Reduce middleware footprint (5) Support simulation environments that use same application software and middleware intended for deployment on the target. We conducted detailed performance studies of nORB in the context of a distributed constraint satisfaction problem (distributed graph coloring) using 100 nodes. We developed the same application using nORB, ACE and TAO and compared them in terms of performance and footprint. While nORB's footprint was much less  than that of TAO and close to that of ACE, its performance was found to be close to that of TAO.

Papers:
Middleware Design and Implementation for Networked Embedded Systems", Embedded Systems Handbook (Richard Zurawski, ed.), CRC Press, Florida, 2005, Chapter 30, pp. 1-17
Middleware Specialization for Memory-Constrained Networked Embedded Systems. A more recent version based on work done by HuangMing Huang is available.
Towards a Performance Model for Special Purpose ORB Middleware
ORB Middleware Evolution for Networked Embedded Systems

Software: Available here.

Brief Bio

 

Education

* PhD, Washington University, St. Louis, MO. Advisor: Dr. Christopher D. Gill. May 2006.
* Masters in Computer Science, University of Missouri-Rolla, May 2000.
* Bachelors in Computer Science, College of Engineering, Trivandrum, University of Kerala, India, June 1991.

Professional Experience

* Research Associate, Center for Distributed Object Computing, Department of Computer Science, Washington University in St. Louis, August 2001 - present.
* Technical Architect, SBC Communications, St. Louis, Missouri, Dec 1996 - Jul 2001.
* Consultant, Mastech Corporation, Pittsburgh, PA, Aug 1994 - Nov 1996. Worked as a consultant at SBC Communications, St. Louis, Missouri.
* Senior Systems Analyst, Tata Consultancy Services, Madras, India. August 1991 - July 1994.

Software Tools Expertise

* C/C++, Java, ACE, TAO, CIAO, CORBA, IF-toolkit, UPPAAL, Linux kernel, Emulab, Vxworks, Websphere, MQSeries, VisualAge for Java.

Personal


I was born in Thiruvananthapuram, the capital city of Kerala, India. I am proud to say that my state is the most literate state in India. Kerala is also a popular tourist destination in India and well deserves its reputation as God's own country. Thanks to the diversity in India, I speak 3 different languages fluently - Malayalam, Tamil, English - and I understand Hindi. I also have some basic background in Sanskrit. I am deeply interested in the teachings of Ramana Maharishi, one of the greatest saints of India. His teachings are simple but profound and he lived an exemplary life that was very true to his teachings.