Research

CURRENT RESEARCH PROJECTS

• Fingerpointing: Problem Diagnosis in Distributed Systems
  This research project focuses on diagnosing the root cause of a variety of problems (performance problems, configuration faults, crash failures, correlated failures, etc.) in distributed systems. This involves a synergistic combination of runtime instrumentation, offline and online analysis through machine-learning and time-series analysis, as well as proactive recovery mechanisms to preemptively avert the failure.

• Survivable Distributed Systems -- Vajra, Elephant, Thema
  The Elephant work focuses on live updates of intrusion detection systems, such as Snort. The Thema work focuses on Byzantine-fault tolerance for multi-tier distributed applications based on Web Services. Vajra focuses on benchmarking the survivability of various distributed infrastructures (such as Castro-Liskov BFT, Immune, Fleet, etc.) through fault-injection of benign and malicious failures.

• MEAD: Adaptive, Dependable Middleware
  Enhances distributed middleware applications with dependability, including: 1) transparent, yet reconfigurable, fault tolerance at runtime, 2) configuration advice to tailor an application's fault-tolerance to its reliability and resource needs, 3) proactive fault tolerance based on failure prediction, 4) resource-aware system adaptation to failures, and 5) enabling distributed, fault-tolerant applications to live realistically with nondeterminism.

• Middleware for Embedded Systems and Sensor Networks
  This research project focuses on middleware for resource-constrained embedded sensor networks. The purpose of the middleware is to allow sensor applications to be programmed all on one single node, and then for the applications to be automatically relocatable/distributable, without requiring changes to the application. This allows for the rapid prototyping of distributed sensor applications in the context of Sensor Andrew, a new initiative within Carnegie Mellon aiming for a campus-wide living sensor testbed.

• Trinetra: Mobile Assistive Technologies
  This project aims to provide blind people with a greater degree of independence in their daily activies. The overall objective is to improve the quality of life for the blind by harnessing the collective capability of diverse networked mobile and embedded devices to support grocery shopping, transportation, navigation, etc.

• Football Research & Engineering
  This project aims to enhance the viewing, refereeing and broadcasting experience of (American) football games through the strategic use of embedded systems, sensors, mobile devices, network protocols and computer vision. This project conveniently taps into two of my passions, football and embedded systems.

USEFUL RESOURCES

• Fault Tolerance and Survivability for Distributed Systems
  
• Downtime and Outages: Real Incidents & Causes
  
• Software Interceptors
  

PREVIOUS RESEARCH

• Eternal: This was the predecessor to the MEAD system that we are currently developing. MEAD was, in fact, born out of the lessons that we learned in building Eternal. Eternal is a transparent infrastructure that I developed to provide strong fault tolerance to CORBA applications, without requiring any modification to the application, to the operating system, or to the ORB itself. Eternal provides support for active and passive replication, overcomes the non-determinism inherent in multithreaded CORBA applications, and provides for gateways to support external clients. The key contributions of this research work are the support for strong replica consistency, the sanitization of non-deterministic multithreading, and most importantly, the transparency of the fault tolerance. This transparency frees CORBA application programmers from worrying about the difficult issues of reliability, and allows them to focus on their area of expertise - the application. This also leads to considerable savings in terms of development time because, as soon as the application logic is ready, fault tolerance is available to be deployed "out-of-the-box" at run-time. Because it does not require any modification of the ORB, the current implementation of the Eternal system provided transparent fault tolerance to the following implementations of CORBA: VisiBroker (Borland), Orbix (Iona), CORBAplus (Expersoft), TAO (Washington University, St. Louis), e*ORB (Vertel), omniORB2 (AT & T Laboratories, UK), ORBacus (Object-Oriented Concepts) and ILU (Xerox PARC). The understanding and insights gained from this research have impacted industrial practices through significant contributions to the Fault-Tolerant CORBA Specification, formally adopted by the CORBA standards body, the Object Management Group.

• Interceptors: Operating-system "hooks" that allow the transparent insertion of code, at run-time, to modify the application behavior. Software interceptors can provide various enhancements to applications, without requiring the application to be re-linked, re-compiled, or re-written. Some of the possible enhancements include security, debugging, profiling, etc. By developing an interceptor to "attach" a reliable infrastructure to CORBA applications, I developed the interception approach to providing fault tolerance for CORBA.

• Immune: Collaborative research effort with Prof. Kim P. Kihlstrom that led to the development of a survivable infrastructure for CORBA applications. Immune enables CORBA applications to continue operating, despite faults that occur within the system, as well as intrusions or malicious/Byzantine attacks that damage the underlying system. Majority voting on the traffic between replicated CORBA objects, value fault detection, and secure multicast protocols (which employ message digests and digital signatures) are the cornerstones of the Immune System.