Rajive Bagrodia - US grants

This research aims to develop a simulation testbed to study performance issues of distributed algorithms in a comprehensive manner. A number of distributed algorithms have been designed to solve some of the central problems of distributed computing like exclusion, synchronization, and deadlock detection. However, little effort has been devoted to studies that attempt to predict the performance of these algorithms under different operating conditions. The research will identify a common set of performance metrics and the primary factors that affect performance of different algorithms. Based on the above metrics, the simulation testbed will be used to achieve two objectives: for given problem (example, synchronization) and operating conditions (example, fully connected network topology) identify "best" available algorithm; use the performance results to develop a hypothesis to express the complexity of distributed algorithms and to quantitatively compare different algorithms for a given problem. The study will focus on a set of problems, which have been solved by using different combinations and extensions of known algorithms. Performance models will be developed using a modular, "building block" approach to exploit the commonality in the solution of these problems.

This research will focus on the design of parallel algorithms, languages and simulation tools that promote effective utilization of parallel architectures by simplifying the task of programming them efficiently. Also, research in evolutionary software design methodologies based on the use of hybrid models that include simulated and implemented system components for the design of performance-critical distributed systems will be performed.

A special configuration of high performance workstations will be acquired with peripheral equipment and supporting software in order to enhance collaborative problem solving environments. The selected projects to which this equipment shall be dedicated include. * Extension of small group R&D environments. * Universal C programming environments. * CAD for high performance reconfigurable VLSI systems. * High-level factory modeling and simulation system.

A planning grant is requested to establish a multi-disciplinary effort at UCLA to apply wireless, mobile networking to support seamless transition of applications across heterogeneous clients for patient record retrieval in medical settings. The planning grant will be used to establish a very small testbed with heterogeneous devices that include desktop computers, laptops, and hand held devices, to provide an environment in which to deploy an early prototype of the proposed system. The testbed will also allow us to enhance interactions among the researchers from different disciplines and the campus telecommunications facility to prepare for an eventual full throttled deployment of the system. Lastly, the small physical testbed will allow us to cleanly separate the advanced development concerns from the more research-oriented issues such that we can deploy the system using state of the art COTS tools while simultaneously pursuing research ideas to influence the next generation of the proposed environment.

Wide-area distributed computing systems represent the future of scientific and commercial computing.
Such systems will enable a wide range of futuristic applications with the potential for enormous economic
and social impact applications such as distributed multimedia services, Web-based collaboration,
distributed supercomputing, and teleimmersion. Before this vision can be realized, however, extensive
research will be required in virtually all aspects of software system design, implementation, and evaluation.
Discrete-event simulation has been an essential tool for the design and analysis of traditional computational
systems and applications. Realistic simulation of applications executing on wide-area distributed systems,
however, is a challenging task because of the scale of the software, hardware and network subsystems that
must be simulated. Nevertheless, the intricacy of these components and their complex, closed-loop
interactions require the components and their interactions to be modeled in sufficient detail to appropriately
predict their impact on overall system performance.
In recent work, the PIs have collaboratively obtained some exciting but preliminary results showing that
specific compiler information can greatly enhance the efficiency and scalability of simulation of message-passing programs. For instance, it was shown that simulation of a scalable ASCI kernel benchmark
application called SWEEP3D executing on up to 128 processors could be simulated faster than real-time by
using parallel simulations together with the type of compiler optimizations that are the subject of this
proposal. Also, the compiler-optimized simulation can evaluate very large data sets on thousands of
processors: it was possible to simulate the performance of a 40 million-problem size Sweep3D for up to
10,000 processors. There are potentially a number of other strategies to dramatically improve simulation
scalability and performance by using compiler information, none of which have been studied so far. A
comprehensive program of research is required to develop their potential and evaluate their impact on
simulation of real world applications.
The focus of the current proposal is to develop compiler-based techniques for improving the efficiency of
parallel discrete event simulation, and to use these techniques to evaluate application and system software
performance for wide-area distributed systems. There are three key components to this proposal:
1. To explore a range of compiler-supported strategies for highly efficient simulation of dynamic, large-scale systems and applications.
2. To use these strategies in developing a practical performance tool for wide-area distributed systems, by
extending our existing compiler and simulation infrastructure. This requires addressing additional
challenges raised by the dynamic nature of these systems and the lack of well-defined metrics to
measure effective application level performance in such environments.
3. To evaluate the effectiveness of these strategies for real world distributed applications such as SF-Express, a large-scale distributed interactive simulation (DIS) environment, and a distributed video-on-demand server (a distributed multimedia application).
The proposed research program builds on a collaboration of several years between the PIs' research groups,
and brings together key strengths in parallel simulation of large-scale parallel programs, parallel simulation of wide-area networks, and in parallelizing compilers and their use for supporting performance evaluation.
This program of research also complements the ongoing software efforts for wide-area systems that are
aimed at developing operating system services (e.g., Globus, Legion, and WebOS) and programming
environments (e.g., Legion, Globe, and GrADS). As such, the proposed research program represents an
essential third leg of software support for the development and deployment of successful wide-area
distributed systems.

The researchers envision that within the next few years mobile and wireless access to the Internet will very likely become the norm, rather than the exception as is seen today. This proposal describes the plans to develop and deploy iMASH, a network system that supports anytime, anywhere, on any platform access to the electronic patient records database for healthcare providers. The objective is to provide the capability for real-time, multimedia communication, so that a physician may access, on the move, the patients record and other relevant information as filtered by the physician's user profile, and may migrate ongoing application sessions seamlessly to different platforms that range from a high performance diagnostic workstation in the physician's office to hand held PDAs in the examination room. While the proposed techniques are general and extend to a range of mobile applications, the specific target of this project is healthcare applications. To this end, the researchers will develop a clinical testbed, which will serve as a laboratory for developing, testing, and evaluating advanced information technology in the context of patient care. The testbed will provide the user requirements to drive the iMASH architecture design, and will permit direct, realistic validation of the research results.
The researchers expect to make the following contributions from this research and development effort:
1) Development of a middleware infrastructure that provides support for anytime, anywhere, on any platform access to the Internet
2) A suite of wireless networking protocols and algorithms that provide quality of service support in a mobile, heterogeneous networking environment
3) A deployment of iMASH within the UCLA Medial School and a controlled study to evaluate its effectiveness in reducing healthcare costs and improving physician effectiveness
4) A system emulation capability that can be used to evaluate the performance and scalability of the middleware services and protocols across multiple dimensions including number of users, number of devices, types of applications, and geographical area. The emulator will be used to 'test drive' novel protocols and applications prior to deployment on the physical testbed.
The researchers have assembled a strong research and development team to undertake the iMASH effort. The team possesses the necessary expertise in the related areas of networking (Zhang, Gerla), wireless communications (Gerla, Lu), parallel and distributed systems (Bagrodia, Gerla), performance evaluation (Bagrodia), computerized medicine (Valentino, McCoy), clinical evaluation of technological innovations in improving heath care (Fiske), and campus computing and communication technology (Solomon).
A longer term goal of this effort is to deploy iMASH-like technology widely within the UCLA campus to support ubiquitous multimedia access for students and faculty, and to support wireless distance education. To enable appropriate technology transition, the team also includes two key members from the university administration: the CIO for the medical school (McCoy) and the Associate Vice-chancellor of Administrative Services with line responsibility over campus telecommunications (Solomon). The UCLA Hospital has recently embarked on a historical reconstruction with a $1 billion endowment. An integral part of the reconstruction is availability of complete wireless connectivity within the hospital. The UCLA campus is also engaged in a project to upgrade the network connectivity throughout the campus with the aim of providing a minimum of 10Mbps bandwidth from desktop to desktop within any two locations on campus. Planning is underway to further enhance this capability with wireless connectivity. These two technology initiatives provide a unique opportunity to insert the iMASH technology in widespread use within the UCLA campus, and subsequently to other locations.

EIA: 0103764
Rajive Bagrodia
University of California-Los Angeles

The goal of this project is to develop the tools needed for performance-directed integrated design and control of complex applications running on distributed computational systems. Its target computational systems are complex, incorporating the difficult heterogeneity, latency, and adaptive properties of computational grids. Its target applications are at the cutting edge of computational science: very large, complex applications with adaptive characteristics that do not allow their optimal system configurations or computational requirements to be estimated prior to run time. Each application will be viewed as a composition of components, with a formal, high fidelity model of performance to be designed for each component. The approach is to use model-based adaptive run-time control, based on these composed performance models, to control the execution of the application to meet specified performance goals. The control strategy will make real-time changes to parameters that modify the behavior of both application and computational platform.

WHYNET: Scalable Testbed for Next Generation Mobile Wireless Networking Technologies
The next generation of wireless communication technology is likely to rely on cross-layer interactions
that extend from the application layer down to the physical devices. This project proposes to design and
develop WHYNET, a Wireless HYbrid NETwork testbed to facilitate detailed study of such interactions and
their impact on application level performance in heterogeneous wireless systems. The eventual technical
impact of this testbed will be to redefine how specific innovations in wireless communication technologies are
evaluated in terms of their potential to improve application-level performance as well as how alternative
approaches are compared with each other. Its broader impact will be to redefine how students are trained in
wireless technologies by providing a multi-disciplinary 'hands on' environment to complement purely
theoretical classroom training.

WHYNET is envisaged as a hybrid testbed that combines the realism of physical testing with the
scalability and flexibility of simulations. The hybrid testbed will be a networked federation of geographically
distributed, heterogeneous wireless physical testbeds with multiple protocol stacks (CDMA 2000 cellular and IP), next generation physical technologies including UWB (Ultra Wide Band), MIMO (Multiple Inputs,
Multiple Outputs) and SDR (Software Defined Radios), and a parallel & distributed multi-tool simulation
framework. Beyond providing a more accurate & flexible evaluation framework, the hybrid testbed will
facilitate a smooth transition from an abstract simulation model to an operational implementation within a
single framework. For instance, protocol prototypes can communicate with simulated lower layers for
repeatable results, or receive and process variable rate real multimedia application inputs for perceptual
evaluation. Once the physical hardware devices are ready for testing, a portion of the target network system
can be configured with real devices while the rest of the network can still reside in the simulated hardware
domain. The effort will also generate a repository of wireless networking scenarios, measurements, models
and implementations. A representative set of studies will be used to demonstrate the unique contributions of
WHYNET for cross-layer optimization studies in particular, and mobile wireless networking in general. These include sensor networks, energy-aware networking, protocols & middleware for multi-access networking, and adaptive transport and security protocols. The testbed itself will be accessible by the research community via a web-based mechanism that will allow remote uploading of models, implementations, and configurations.

The proposed research is likely to have a broader impact on two fronts: the training of future generation
of wireless engineers and wireless technology standards. Wireless engineers will need significant technical
depth to contribute to a rapidly developing technology and significant technical breadth to understand how this
technology fits into a market driven economy. The latter category requires engineers who are trained in
insystemslt aspects with an in-depth understanding of trade-offs and interactions across layers of a wireless
communication system. The current course structure is not designed to produce well-trained engineers of the
second type. The project team feels strongly that broad systems training can only be accomplished in i.hands-
onlo experimental courses or projects where the students see the tradeoffs involved in real system design. The proposed testbed can enable these types of courses across the curriculum. Even though today wireless is a
vertical technology, 4-5 years from now, the most interesting and challenging problems will be those related to wireless systems, so we believe that an inter-disciplinary yet closely-knit engineering program such as ours is well suited for the training of wireless engineer of tomorrow. By providing a scalable platform, methodology, and tools to support objective and accurate evaluation of protocol and technology alternatives, we expect that the testbed will also play an important role in shaping standards activity in IETF and related bodies.

A multi-disciplinary, multi-institution team has been formed to achieve the ambitious objectives of the
WHYNET project. The team members have substantial expertise in design and management of physical and
simulation testbeds (Bagrodia, Gerla, Rao, Takai), development of novel radio technology (Daneshrad, Fitz,
Mitra), wireless systems (Mitra, Rao, Srivastava), protocol design (Gerla, Krishnamurthy, Mohapatra, Royer, Shen, Srivastava, Tripathi) and performance evaluation (Bagrodia, Gerla, Molle, Rao, Tripathi). Many of thePIs have successfully worked together on previous collaborative projects. We have also received strong support from a number of companies that play a critical role in this space including Microsoft, Hughes
Research Laboratories (now part of Boeing), ST Electronics, HP, Ericsson, Intel, and Xtreme Spectrum.