Welcome

 

Monaco, Francisco José

monaco.fj@acm.org

University of São Paulo

at São Carlos, Brazil

Biographic notes

After obtaining the Doctorate degree in Electrical Engineering in 2002 by the University of São Paulo, Brazil, I joined ICMC-USP where I currently hold a tenure faculty position since 2003.

At the Department of Computer Sciences, SSC-ICMC, my work includes teaching(1) in undergraduate and graduate courses, supervising Msc and PhD theses and carrying out research work.

My main investigation area lays in the fields of Real-Time and Distributed Systems.

I am also a researcher of the Brazilian National Institute for Research and Technology – Critical Embedded Systems sector, INCT-SEC(2).

Other scientific activities relates to Free Open Source Software(3) and Computer in Education.

Personal interests include walking trails and landscape photography.

Current activities (2009)
  1. Disciplines: Computer Networks, Operating Systems, Pervasive Computing, Real-Time Systems, Fault-Tolerant Systems.
  2. The INCT-SEC (FAPESP/CNPq) carries out research in impacting areas of national economical and social development, with emphasis on robotic systems and autonomous vehicles.
  3. Steering committee of the ICMC Competence Center for Free Software / Open Source, a USP center devoted to research, development and education in open source software and open knowledge generation.

    Researcher of the ICMC staff in the QualiPSo Project - Quality Platform for Open Source Software. Involving European, Brazilian and Chinese ICT industry players, SMEs, governments and academics, QualiPSo is funded by the European Commission’s FP6 under IST initiative, aiming at fostering trustworthiness in industrial adoption of Open Source sofware.

Research outline

Keywords: Real-Time, Distributed Systems, QoS, SOA, Pervasive Computing

QoS-Aware Real-Time Distributed Systems

As the envisioned pervasive computing age is being brought into reality by the ubiquitous dissemination of networked computer systems in our surrounding environment in the form of ever-diffusing devices — from smart hand-held and home appliances, to mission-critical systems such as vehicular, industrial and medical equipment — the integration of on-line resources into the implementation of services we rely upon for our daily activities raises concerns on performance and dependability requirements.

Beyond the demands of traditional Internet regular file-transfer applications, originally meant for a context of asynchronous transactions, those novel computer services are supposed to respond synchronously to real-world events and meet temporal constraints dictated by the dynamics of their operational environment. This condition elicits a real-time system approach.

Most of the classical theory on real-time, nonetheless, arises from the domain of automation engineering, where the deterministic timing characteristics have allowed for the development of sound analytical techniques. Contrasting with this scenario, the typically stochastic load patterns and poorly predictable event-driven dynamics of network interactive services render performance guarantees exceedingly more complex in large-scale distributed systems. While ensuring strict temporal requirements may often be technically or costly unfeasible under realistic assumptions, responsiveness is still a key quality-of-service (QoS) metric for real-time interactive services.

The extension of methodological and technical results from the real-time theory to address the non-deterministic features of interactive services, under a systemic cross-layer perspective (from network to application level), is a research-demanding area. The scientific investigation theme aligned under this perspective aims at the challenges posed by QoS-aware real-time distributed systems. Particularly, the study focuses on self-adaptive resource management strategies for the handling of the unpredictable dynamics and complexity of internal and external factors impacting performance, under a systemic cross-layer. This approach is explored in emerging domains within the distributed systems field such as real-time service-oriented architectures (SOA), novel QoS and service-level agreement (SLA) models and performance evaluation models and metrics.

Currently active topics:

Real-Time Service Computing
One important paradigm in the design of large-scale distributed computer applications is given by the service-oriented architecture (model, based on the composition of complex application out of independently developed, loosely coupled, component services. When deployed as operational components of production systems, SOA systems targeted to real-time business processes must fulfill the application's response time requirements. Monitoring systems, on-line computer-supported collaborative tools, mobile context-aware applications and action management engines, to name a few examples, follow into this categories of services with real-time constraints. The fulfillment of both the functional requirements concerning the overall business logic and the non-functional requirements concerning the QoS levels expected from the SOA system is a challenging goal in view of the stochastic dynamic of the network infrastructure. This research work addresses run-time control strategies of service selection and request scheduling in composite applications supporting real-time business processes. The investigation aims at the the definition of methodologies, techniques and tools for enabling self-adaptive QoS provision mechanisms in large-scale distributed systems.

Self-Adaptive Real-Time Resource Control
One emerging concept to cope with the unpredictability of large-scale distributed computer applications is the use of feedback adaptive principle. It grounds on the fundamentals of Control Theory, which has developed as a groundbraking field in Engineering and in some natural science branches, and counts on a rich collection of mathematical modeling tools to describe the behavior of dynamic systems in terms of how it respond to different stimuli, and to verify stability, observability and controllability properties. Feedback rationale bases on the continuous monitoring of the the difference between the system output and the meant reference value, and its use as an input signal which is injected back into the system so as to force it contrarily to the deviation iself. By manipulating the deviation signal through a properly designed control function it's possible to shape the dynamic properties of the system and tailor the response characteristics to meet specific requirements. The application of Control Theory in computer systems architectures is nevertheless recent, and considerably less explored than in other domains. It has been employed, for instance, in resource partitioning problems and congestion control. This research theme is devoted to the investigation of an analytical foundation for the study of real-time distributed systems and the modeling of their dynamic properties. It's outcome shall include approaches and techniques to analyze the steady-state and transient response characteristics with respect to inner and outer disturbances and to design adaptive run-time QoS control mechanisms.

Recent and ongoing outcomes in these research lines include the development of novel real-time scheduling, load-balance and admission-control algorithms, feedback-adaptive resource allocation architectures and QoS and SLA approaches based on aggregate performance measures.

Recent related publication

Real-Time, QoS, SOA

Other fields