Start date: October 2017

Type of contract: Inria fixed term contract (CDD) for 3 years.

Salary: About 1600 EUR net per month (1980 EUR gross), health insurance included (French Social Security system).

Location: The thesis project takes place within an academic/ industrial collaboration between Inria and Nokia Bell Labs and in this common lab context, the thesis location will be balanced between the Inria Montbonnot site, at about 10 kilometers from Grenoble, and the Nokia Bell Labs site in the south of Paris. On average, the PhD candidate is expected to spend about two years at Inria Grenoble and one year at Nokia Bell Labs Nozay.

Subject: « Softwarization of Everything: IoT Service Composition »

Objectives:

The Softwarization of Everything tends to gather under the same roof computing (clouds/fogs), networking (SDN) and complex networked services oriented to the user (apps, IoT). Such systems exhibit common features like distribution, dynamic/on demand (re)configuration, usage of multi-techno multi-domain cross-layer resources, and the need to automate most management operations like configuration, trouble-shooting, repair, optimization, etc. The current trend to control everything by software should help in managing all the connected objects and the network elements. It enables a large scope, both horizontal (end-to-end, multi-tenants) and vertical (from physical layer to applications), to manage the connected objects and the network elements together, in order to render the best services to the end users.

The French Institute for Research in Computer Science and Automation (Inria) is a national research institution focusing on computer science and applied mathematics. Inria promotes scientific excellence for technology transfer and society in fields as diverse as healthcare, transport, energy, communications, security and privacy protection, smart cities and the factory of the future. In particular, CONVECS is an Inria Grenoble research team with areas of expertise on formal modelling and formal verification techniques and tools for asynchronous concurrent systems. CONVECS is involved in several research directions: providing formal specification languages for describing concurrent systems; enhancing temporal logics and verification tools; fighting state explosion; designing generic components for verification, test, and performance evaluation; and demonstrating the applicability of its methods and tools on real-life applications (cloud computing, networks, communication protocols, embedded systems, etc.).

The Softwarization of Everything trend is important for Nokia future development and corresponds to the vision of the company. Nokia is often described as a global leader in the technologies that connect people and things. Nokia weaves together the networks, data, and device technologies to create the universal fabric of our connected lives - where new applications flow without constraint, where services and industry automate and run seamlessly, where communities and businesses can rely on privacy, security, and near instant response times, connecting through the Cloud. Bell Labs drives this vision within Nokia by inventing future networks : "We are at the dawn of an era in networking that has the potential to define a new phase of human existence. This era will be shaped by the digitization and connection of everything and everyone with the goal of automating much of life, effectively creating time by maximizing the efficiency of everything we do..." (M. Weldon, The Future X Network: A Bell Labs Perspective, CRC press, 2016).

The objective of the Nokia / Bell Labs partnership is to help in making this vision a reality, by carrying out common research on an integrated framework for softwarized networks, IoT and clouds/fogs that aims at facilitating the delivery of the best services to people, and on the management of such complex and dynamic systems.

Work proposed and expected results:

The general objective of this PhD is to investigate and propose (i) abstract models for the description of things and their composition, including the associated pieces of software, and (ii) techniques for re/configuring such applications privileging automation, decentralization, and robustness. More precisely, the main tasks of the PhD thesis will be as follows:

• Study the state of the art on SDN, NFV, IoT service composition, etc., as well as existing solutions for deployment and reconfiguration of distributed software.


• Based on this study, define abstract models for managing distributed software.


• Propose dynamic reconfiguration algorithms keeping in mind the properties of interest (e.g., service availability) that applications must respect during the online update and modification of the software being reconfigured. Compatibility notions can also be defined for formally describing what must be preserved by the reconfiguration process.


• Implement the algorithmic solutions proposed above in the context of concrete frameworks and case studies.


• Evaluate the solutions from a quantitative perspective (performance, response time, etc.).

Required skills and profile:

• Knowledge of networks, information/data models, distributed systems, and algorithmics

• Knowledge of formal methods (concurrency theory) and verification is a plus

• Candidates who enjoy programming would be appreciated, as the work will include software development

• Education: MSc/Master 2 Recherche in Computer Science

• Good command of English as the working language, French is a plus

• Proven communication and interpersonal relationships skills, attention to detail, methodical approach, autonomy, team player

Contacts:

Gwen Salaün: Gwen.Salaun@inria.fr
Michel Le Pallec: michel.le_pallec@nokia-bell-labs.com

Application content:

• Letter of application

• Curriculum vitæ

• School report

• References or letters of recommendation, if any

• Scientific or technical publications, if any

Application submission:

Applications should be addressed directly to Gwen Salaün and Michel Le Pallec, by e-mail, mentioning the position number #2017A. Applications received after July 18, 2017 might not be considered if a candidate has been selected already.