© Photo by Gonçalo João

| Home | Curriculum | Research | Teaching | Publications | Students |

João Leitão's Homepage : Research

Research

Research Statement:

Research Statement: Systems are becoming more and more relevant in the everyday life of a broader number of users. This leads system's design and architectures to become more and more complex at the same time that user's expectations concerning responsiveness, availability, and dependability grow. In this context, my research strives to answer the broad question of: How to scale systems, and distributed systems in particular, in number of components, geographic distribution, and user base, while ensuring high availability, efficiency, and fault tolerance?

To answer this question I focus on the following research vectors: i) devising new algorithms and solutions to scale distributed architectures, such as those of distributed databases, both within the scope of a data center, as well as across multiple geo-distributed sites, minimizing the use of centralized components that hinder performance and fault-tolerance; ii) understanding fundamental aspects of consistency semantics in the context of data replication and proposing new algorithms that explore relevant trade- offs between consistency and user perceived performance; iii) extending the horizon of replication techniques, in the context of geo-replicated data stores, as to include clients, while enabling direct browser-to-browser communication and collaboration, in an intuitive way for system architects; and iv) devise new security mechanisms to increase the dependability of cloud-based large-scale systems.

This research is entiwned with the strategic plan of the NOVA LINCS Laboratory to develop new solutions and algorithms to support large-scale distributed systems.

Research Areas:

  • Cloud Computing
  • Geo-Distributed and Geo-Replicated Systems
  • Peer-to-Peer Systems
  • Gossip protocols
  • Overlay networks
  • Large scale distributed systems
  • Fault tolerance
  • Dynamic systems

Research Projects:

I am currently part of the research team of the following funded research projects.

LightKone - Lightweight computations for networks at the edge.
Technical Lead at FCT UNL. FCT UNL Local Co-coordinator. Workpackage Leader (WP5).
Coordinator: Peter van Roy, Université Catholique de Louvain (UCL) in Louvain-la-Neuve, Belgium
Financed by the European Union under the H2020 grant agreement ID 732505.
[LightKone Webpage]

NG-STORAGE - New Generation of data STORage And manaGement systEms.
Principal Researcher.
Financed by FCT under contract PTDC/CCI-INF/32038/2017 (starting October 2018).

Secure and Scalable Platform for Massive-scale Mobile Applications.
Co-Principal Researcher. Principal Researcher: Nuno Preguiça (NOVA LINCS & FCT UNL).
Financed by FCT under contract PTDC/CCI-INF/32662/2017

HiPSTr: High-Performance Serializable Transactions.
Co-Principal Researcher. Principal Researcher: João Lourenço (NOVA LINCS & FCT UNL).
Financed by FCT under contract PTDC/CCI-COM/32456/2017

Cosmos: Causal Consistency on the Network Edge.
NOVA FCT UNL Local Coordinator. Principal Researcher: Luís Rodrigues (INESC-ID & Técnico UL)
Financed by FCT under contract PTDC/EEI-COM/29271/2017

HYRAX - Crowd-Sourcing Mobile Devices to Develop Edge Clouds, FCT/CMU (FCT CMUP-ERI/FIA/0048/2013)
[HYRAX Homepage]

Past Research Project:

In the past I participated in the following funded research projects.

SyncFree - Large-scale computation without synchronisation (SyncFree is a European research project taking place for 3 years, staring October 2013, and is funded by the European Union, grant agreement n 609551.)
[SyncFree Homepage]

ADAAS - Assuring Dependability in Architecture-based Adaptive Systems FCT (CMU-PT/ELE/0030/2009)
[ADAAS Homepage]

REDICO - Dynamic reconfiguration of communication protocols
[REDICO Homepage]

HPCI - High-Performance Computing over the Large-Scale Internet
[HPCI Homepage]

P-SON: Probabilistically-Structured Overlay Networks
[P-SON Homepage]

Available Software/Prototypes:

Legion Framework
This is a javascript framework that enables writting web applications where clients (transparently) replicate fractions of the application state and synchronize direcly among them.
The software and a demo are available in the Legion web page here.

Overnesia Overlay
I have been developing a novel cluster-based overlay named Overnesia. Here you can find the first prototype of Overnesia, which was already deployed in PlanetLab for preliminary tests. If you want to execute this code or have any double please contact me.

Other Research Activities:

Appia (collaborator)
[since September 2005][Appia Homepage]