Brice NÉDELEC

UNTIL: March 2021
FROM: September 2019

A software architect position at Sigma Informatique in La-Chapelle-sur-Erdre, France. This project aimed at enhancing microservices architectures in order to include end users in the decision processes about their energy consumption.

UNTIL: December 2018
FROM: March 2017

A post-doctoral position at LINA laboratory in Nantes, France. The topics were about causality tracking in unstructured networks, and broadcasting as a communication primitive for large scale networks subject to dynamic changes.

UNTIL: October 2016
FROM: October 2012

As a Ph.D student and member of the GDD team, at LINA laboratory in Nantes, France, under the supervision of Pr. Pascal Molli and Pr. Achour Mostéfaoui. The topics ranged from replicated data structures, to decentralized network membership protocols, and large scale collaborative editing in Web browsers.

UNTIL: June 2014
FROM: October 2013

I was teaching at the Institute of Technology of Nantes in France to first year students. The courses included Swing, AWT, as well as JavaScript, CSS, and HTML.

UNTIL: August 2012
FROM: March 2012

An internship that would later lead to a Ph.D! Logoot is a distributed data structure that allows writers to collaboratively and concurrently edit text documents. This data structure needs identifiers to avoid the difficult and error-prone task of solving editing conflicts. I studied ways to reduce the size of such identifiers.

UNTIL: June 2009
FROM: April 2009

Nothing new: monitoring is essential. This was an internship at IUT, site de la Fleuriaye, Carquefou, France. Developing such metrology tool to monitor software licences implied data aggregation; export and alert; a graphical user interface using Qt.

DATE: December 2018
AUTHOR: Brice Nédelec
AUTHOR: Pascal Molli
AUTHOR: Achour Mostéfaoui
PROCEEDINGS: 22nd OPODIS.

Causal broadcast constitutes a fundamental communication primitive of many distributed protocols and applications. However, state-of-the-art implementations fail to forget obsolete control information about already delivered messages. They do not scale in large and dynamic systems. In this paper, we propose a novel implementation of causal broadcast. We prove that all and only obsolete control information is safely removed, at cost of a few lightweight control messages. The local space complexity of this protocol does not monotonically increase and depends at each moment on the number of messages still in transit and the degree of the communication graph. Moreover, messages only carry a scalar clock. Our implementation constitutes a sustainable communication primitive for causal broadcast in large and dynamic systems.

DATE: October 2018
AUTHOR: Brice Nédelec
AUTHOR: Pascal Molli
AUTHOR: Achour Mostéfaoui
PROCEEDINGS: 37th SRDS

Many distributed protocols and applications rely on causal broadcast to ensure consistency criteria. However, none of causality tracking state-of-the-art approaches scale in large and dynamic systems. This paper presents a new non-blocking causal broadcast protocol suited for dynamic systems. The proposed protocol outperforms state-of-the-art in size of messages, execution time complexity, and local space complexity. Most importantly, messages piggyback control information the size of which is constant. We prove that for both static and dynamic systems. Consequently, large and dynamic systems can finally afford causal broadcast.

DATE: 2017
AUTHOR: Brice Nédelec
AUTHOR: Pascal Molli
AUTHOR: Achour Mostéfaoui
PROCEEDINGS: CCPE 4108

Distributed real-time editors made real-time editing easy for millions of users. However, main stream editors rely on Cloud services to mediate sessions raising privacy and scalability issues. Decentralized editors tackle privacy issues but scalability issues remains. We aim to build a decentralized editor that allows real-time editing anytime, anywhere, whatever is the number of participants. In this paper, we propose an approach based on a massively replicated sequence data structure that represents the shared document. We establish an original trade-off on communication, time and space complexity to maintain this sequence over a network of browsers. We prove a sublinear upper bound on communication complexity while preserving an affordable time and space complexity. In order to validate this trade-off, we built a full working editor and measured its performance on large scale experiments involving up till 600 participants. As expected, the results show a traffic increasing as O((log I)² ln R) where I is the number of insertions in the document, and R the number of participants.

DATE: 2017
AUTHOR: Brice Nédelec
AUTHOR: Julian Tanke
AUTHOR: Davide Frey
AUTHOR: Pascal Molli
AUTHOR: Achour Mostéfaoui
PROCEEDINGS: WWWJ

Peer-sampling protocols constitute a fundamental mechanism for a number of large-scale distributed applications. The recent introduction of WebRTC facilitated the deployment of decentralized applications over a network of browsers. However, deploying existing peer-sampling protocols on top of WebRTC raises issues about their lack of adaptiveness to sudden bursts of popularity over a network that does not manage addressing or routing. Spray is a novel random peer-sampling protocol that dynamically , quickly, and efficiently self-adapts to the network size. Our experiments show the flexibility of Spray and highlight its efficiency improvements at the cost of small overhead. We embedded Spray in a real-time decentralized editor running in brow-sers and ran experiments involving up to 600 communicating web browsers. The results demonstrate that Spray significantly reduces the network traffic according to the number of participants and saves bandwidth.

DATE: October 2016
AUTHOR: Brice Nédelec
PROCEEDINGS: Ph.D manuscript

Collaborative editors allow users to distribute the writing of a document across space and time. Thanks to their ease of use, real-time collaborative editors working in Web browsers vastly contributed to the adoption of such tools. However, current editors are centralized: a service provider's server hosts an editing session. It raises privacy and scalability issues.

Recently, the enabling of browser-to-browser connection establishments opened new opportunities in favor of a decentralized Web. Decentralized real-time collaborative editors working in Web browsers must efficiently handle highly dynamic groups of different size.

Contritributions of this thesis are threefold: (i) To represent the document, we propose a replicated data structure for sequences using metadata the size of which scales sub-linearly compared to the number of inserted characters. (ii) To efficiently propagate the changes to all editors involved in the collaborative writing, we propose a random peer sampling protocol that supports Web browsers constraints and self-adjusts its functioning to the variations of network membership. (iii) To demonstrate the feasibility of a decentralized real-time collaborative editors running in Web browsers, we propose an editor using (i) and (ii), and we highlight its scalability.

DATE: April 2016
AUTHOR: Brice Nédelec
AUTHOR: Pascal Molli
AUTHOR: Achour Mostéfaoui
PROCEEDINGS: WWW'16 Companion

Real-time collaborative editors are common tools to distribute work across space, time, and organizations. Unfortunately, mainstream editors such as Google Docs rely on central servers and raises privacy and scalability issues. CRATE is a real-time decentralized collaborative editor that runs directly in web browsers thanks to WebRTC. Compared to state-of-the-art, CRATE is the first real-time editor that only requires browsers in order to support collaborative editing and to transparently handle from small to large groups of users. Consequently, CRATE can also be used in massive online lectures, TV shows or large conferences to allow users to share their notes. CRATE's properties rely on two scientific results: (i) a replicated sequence structure with sub-linear upper bound on space complexity; this prevents the editor from running costly distributed garbage collectors, (ii) an adaptive peer sampling protocol; this prevent the editor from oversizing routing tables, hence from letting small networks pay the price of large networks. This paper describes CRATE, its properties and its usage.

DATE: September 2017
AUTHOR: Brice Nédelec
AUTHOR: Pascal Molli
AUTHOR: Achour Mostéfaoui
AUTHOR: Emmanuel Desmontils
PROCEEDINGS: 13th DocEng

Distributed collaborative editing systems allow users to work distributed in time, space and across organizations. Trending distributed collaborative editors such as Google Docs, Etherpad or Git have grown in popularity over the years. A new kind of distributed editors based on a family of distributed data structure replicated on several sites called Conflict-free Replicated Data Type (CRDT for short) appeared recently. This paper considers a CRDT that represents a distributed sequence of basic elements that can be lines, words or characters (sequence CRDT). The possible operations on this sequence are the insertion and the deletion of elements. Compared to the state of the art, this approach is more decentralized and better scales in terms of the number of participants. However, its space complexity is linear with respect to the total number of inserts and the insertion points in the document. This makes the overall performance of such editors dependent on the editing behaviour of users. This paper proposes and models LSEQ, an adaptive allocation strategy for a sequence CRDT. LSEQ achieves in the average a sub-linear spatial-complexity whatever is the editing behaviour. A series of experiments validates LSEQ showing that it outperforms existing approaches.

DATE: September 2013
AUTHOR: Brice Nédelec
AUTHOR: Pascal Molli
AUTHOR: Achour Mostéfaoui
AUTHOR: Emmanuel Desmontils
PROCEEDINGS: DChanges

Distributed collaborative editors such as Google Docs or Etherpad allow to distribute the work across time, space and organizations. In this paper, we focus on distributed collaborative editors based on the Conflict-free Replicated Data Type approach (CRDT). CRDTs encompass a set of well-known data types (sets, graphs, sequences, etc.). CRDTs for sequences model a document as a set of elements (character, line, paragraph, etc.) with unique identifiers, providing two commutative update operations: insert and delete. The identifiers of elements can be either of fixed-size or variable-size. Recently, a strategy for assigning variable-size identifiers called LSEQ has been proposed for CRDTs for sequences. LSEQ lowers the space complexity of variable-size identifiers CRDTs from linear to sub-linear. While experiments show that it works locally, it fails to provide this bound with multiple users and latency. In this paper, we propose h-LSEQ, an improvement of LSEQ that preserves its space complexity among all collaborators, regardless of the latency. Ultimately, this improvement allows to safely build distributed collaborative editors based on CRDTs. We validate our approach with simulations involving latency and multiple users.

OS: Unix, Windows

Languages: C, C++, Java, LaTeX, NodeJS, Python, Rust

Web: Javascript, CSS, HTML, WebRTC

Groupeware: git (with Github and CI), svn

Design: UML

French: 5/5

English: 4/5

Spanish: 2/5

Guitar: 2/5