Home » Gruppi Di Ricerca » 18322
Distributed Systems

The Distributed Systems group has developed, in the last fifteen years, a solid worldwide reputation in the context of theory and practice of distributed, pervasive and p2p computing, middleware platforms, data processing, and information systems infrastructures. On these topics, the group has created strong relationships with the most influential research groups in the world.

We developed several theories and practical experiences in various topics including checkpointing, causal and total ordering theory, distributed replication, group communication, distributed agreement, publish subscribe systems, dynamic systems, byzantine fault tolerance, distributed stream processing, etc.

The distributed systems group has participated and successfully coordinated several important EU projects in the context of e-government, security and dependability of large scale systems, and protection of critical infrastructures. It has developed remarkable connections with the major Italian ICT industries and Public Administrations for creating innovative solutions and prototypes transferring the latest results from research area into practice.

Current research areas include:

Byzantine fault-tolerant algorithms: in the past few years the group has proposed several solutions in the area of BFT focussing, in particular, on algorithms for basic distributed abstractions in both static and dynamic settings and algorithms for robust lattice agreement algorithms. In this context, the group is also investigating solutions able to deal with the so called Mobile Byzantine Failure model.

Distributed stream processing systems: since 2003 the group has regularly proposed novel solutions for improving the efficiency of distributed stream processing systems. In particular, we focussed our efforts on designing solutions to dynamically adapt the system runtime to changes in the input load distribution to tackle different goals (e.g. latency reduction, throughout maximization, efficient resource usage, etc.)

Dynamic networks and population protocols: The group has a keen interest in the study of dynamic networks, especially the one composed by anonymous processes. In this area, it has designed the first known terminating counting algorithms for rooted interval-connected networks, bootstrapping the research in the field. Regarding, population protocols the group has been the first to investigate computability under faulty interactions increasing the understanding of fault-tolerance for population protocols.  The group also provided contribution to the analysis of theoretical aspect of distributed systems affected by continuous churn i.e., the phenomenon of continuously changing the set of processes participating in to the distributed system.  

Mobile agents and robots: The DS group has strong expertise in the field of mobile agents (autonomous entities inhabiting a graph) and mobile robots (autonomous entities inhabiting an euclidean space). Regarding mobile agents, it has been the first to investigate, with a distributed perspective, the problems of exploration, gathering, patrolling, and black hole search on dynamic interval connected graphs. While in the field of robots it has been the first to study the computational power of luminous robots in the obstructive model, and it has given general contributions in understanding the computational power of oblivious robots in the setting of restricted visibility.

DLT and Blockchain: in the fast few years the group started to investigate the theoretical foundations of Blockchains and (more in general) of DLT and how to efficiently take advantage of such technologies to support applications behind cryptocurrencies.

Fog and Edge Computing: The DS group also has experience in designing distributed, cooperative, and decentralized algorithms that target the problems of load balancing and scheduling problems in the Edge and Fog Computing environments. With the former, we intend to optimize the load among all the nodes involved in the system to avoid saturation and consequently increase the number of successfully served tasks and minimize the latency, while in the latter, we deal with tasks that have specific deadlines that they necessarily need to meet. We have expertise in mathematical modeling (probabilistic models, linear optimization, and discrete/continuous time systems) of the problem and modern technologies like Docker and Kubernetes. Moreover, we also rely on Reinforcement Learning to design adaptive and resilient strategies that cope with unpredictable changes in the environment in which the algorithm runs. In general, our main objective is to start from the problem analysis, then model the system and the solution, and finally implement a working approach both in simulation and on real and pseudo-real environments, such as clusters of Raspberry Pis. For this reason, the group also developed an open-source framework called P2PFaaS (https://p2p-faas.gitlab.io) which allows the implementation of cooperative and decentralized scheduling and load-balancing algorithms on Fog and Edge nodes which are Docker-enabled.

The Distributed Systems group is also strongly involved in the activities of the Research Center of Cyber Intelligence and Information Security (CIS). CIS does leadership research in the context of cyber security, information assurance, critical information infrastructure protection, trend prediction, malware analysis, open-source intelligence, cyber physical systems and smart complex systems. Advanced capabilities in cyber intelligence will be indeed essential in the next years due to the pervasiveness of cloud, social computing and mobility technologies, that lower the control that organizations and governments have over systems, infrastructure and data. CIS aims at designing better information security methodologies, threat profiles and at elaborating defense strategies taking into account the economic and legal impact in a unique framework. Research results are applied to real world contexts such as cyberwarfare, fraud detection, stock market stability, detection of tax evasion, monitoring of mission-critical systems, early warning systems and smart environments.



© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma