Faites connaître cette offre !
Reference : UMR8049-RAMLAN-002
Workplace : MARNE LA VALLEE
Date of publication : Tuesday, November 10, 2020
Type of Contract : FTC Scientist
Contract Period : 12 months
Expected date of employment : 1 December 2020
Proportion of work : Full time
Remuneration : Between 2500 et 2700€ gross salary per month (before taxes)
Desired level of education : 5-year university degree
Experience required : Indifferent
Next generation mobile networks (5G / 5G +) must support a wide variety of devices (drones, autonomous vehicles, IoT, etc.) and uses with divergent characteristics in terms of mobility, security, latency, reliability, etc. According to 5G-PPP , the 5G network is capable of supporting three categories of Network Slice, referred to as: i) Enhanced Mobile Broadband (eMBB), ii) Massive Object Communications (mMTC) and iii) Ultra-reliable, low latency communications (uRLLC).
In this context, the SNR team (Software, Networks, Real-time) of the LIGM laboratory first developed a BeC3 platform (Behavior Crowd Centric Composition)  which allows the creation and deployment of distributed applications of Internet of Things. Secondly, it developed a 5G network prototype based on SDN (Software Defined Networking) and using Open Air Interface (OAI)  software, which allows the creation and configuration of several network slices at the same time. request in a Cloud radio environment [4, 5].
Presentation of the BeC3 platform:
The BeC3 platform consists on the one hand of a web application (developed with the Framework Play in java ). The interface is in AngularJS. On the other hand, there are several clients, one for smartphones (in Ionic / cordoba) and a version in python for any other type of object (home automation, and even pc). The whole is articulated around XMPP, which allows objects to recognize each other, and to communicate. It supports different protocols: Blutooth Low Energy, EnOcean, ZWave, and types of objects: home automation, smartphones and even Twitter type services.
The specific approach consists in proposing choreographies of services: the objects included in the platform are programmable: it is possible to deposit code in the form of XMPP messages. These messages contain the description of the actions to be performed, and the object begins to obey them. Thus, it is possible, for a user, to draw from a list of pre-written behaviors, to deploy them via XMPP on the objects, and then to let them exchange the messages dictated by these behaviors.
The platform adds to these basic mechanisms checks (to avoid deploying combinations of codes that would make no sense) and a tool to write compatible codes (very simple codes, the behavior of each object is described in the form of 'automata).
Presentation of the 5G prototype:
The SNR team has developed an open-source 5G software platform based on Open Air Interface (OAI)  software to implement radio access network (RAN) nodes and the EPC (Evolved Packet Core) network. The team has developed a virtualized version of RAN components which is based on Docker container technology . As a result, the RAN consists of the Remote Radio Unit (RRU) node, which includes a remote radio transceiver using a USRP B210 type software radio card, which is interconnected via a link interface to the Radio Cloud Center (RCC) node, which performs the baseband processing functions.
Docker RAN containers run on an “Ubuntu 14.04” laptop computer, characterized by “Linux kernel” version 3.19.0-61-lowlatency SMP PREEMPT, i7 - 4-core CPU and 16 GB RAM (Random Access Memory). The EPC functions work on a second “Ubuntu 16.04” laptop, characterized by “Linux kernel” version 4.7.0-1, i5 vPro, a 4-core processor and 4 GB RAM. The EPC is interconnected to the RAN via a Gigabit Ethernet interface, called "S1", while it is connected to the Internet via a second Ethernet interface.
The prototype comprises a total of 5 RRU nodes, several smartphones and uses the SDN FlexRAN controller , which allows the orchestration of radio resources via a specific southern interface (SBI) based on the Google Protobuf protocol. For more information, please visit the website mentioned in reference .
Existing work :
The SNR team set up a scenario involving the two prototypes mentioned above and which allows the orchestration of two types of network slices: eMBB transporting video traffic and mMTC transporting BeC3 services while taking into account their requirements for quality of service (QoS). To do this, the team has developed a Network Slicing SDN application in Python, called "Slicing APP", which runs on top of the FlexRAN controller allowing users to perform network slicing experiments in a user-friendly and abstract. The 5G / IoT platform obtained is completed by three Android smartphones, which broadcast YouTube videos, and a set of physical and virtual EnOcean  objects based on the BeC3 platform (eg, smart plugs, smart lamps, Raspberry PI, etc.) allowing to generate dense IoT type traffic.
The recruited engineer should perform the following tasks:
• Takeover of the current installation (certificates, validity of exchanges) of the 5G / IoT platform including the BeC3 platform .
• Evaluation of the transition to a more recent version of OAI and XMPP, change of server.
• Once the versions are stabilized, changes will have to be studied:
o Should we keep XMPP or switch to other types of instant messaging or other protocols?
o Add new behaviors and events supported by IoT objects,
o Add other platform management methods (automatic deployment, feedback).
• Develop new APIs allowing the interconnection of the updated BeC3 platform with our 5G / IoT prototype.
• Test and implement other SDN controllers in our 5G / IoT platform while respecting the new O-RAN reference architecture .
• Evaluate the scaling up of the platform by increasing the number of supported nodes. Three new software radio cards of the USRP X310 type  and several IoT / smartphone objects will be acquired by the SNR team as part of the FlashRAN platform project funded by the DIM RFSI .
 “5G PPP 5G Architecture”, White Paper, version 2.0, Dec 2017.
 “BeC3: Behaviour Crowd Centric Composition for IoT applications”, available online at http://bec3.com/en/
 N. Nikaien, “OpenAirInterface Simulator/Emulator”, available on line at : http://www.openairinterface.org/
 S. Costanzo I. Fajjari, N. Aitsaadi, and R. Langar, "A Network Slicing Prototype for a Flexible Cloud Radio Access Network", IEEE CCNC, Las Vegas, USA, Jan. 2018.
 X. Foukas, N. Nikaein, M. M. Kassem, M. K. Marina and K. Kontovasilis, “FlexRAN: A Flexible and Programmable Platform for Software-Defined Radio Access Networks”, ACM CoNEXT. Dec. 2016.
 S. Costanzo, S. Cherrier, and R. Langar, "Network Slicing Orchestration for IoT-BeC3 applications and eMBB services in C-RAN", IEEE INFOCOM 2019, Paris, France, May 2019.
 N. Salhab, R. Rahim, R. Langar, and R. Boutaba, "Machine Learning based Resource Orchestration for 5G Network Slices", IEEE GLOBECOM 2019, Waikoloa, Hawaii, USA, Dec. 2019.
 O-RAN Alliance, https://www.o-ran.org/
 “Universal Software Radio Peripheral (X310),” http://www.ettus.com/all-products/x310-kit/
The candidate must hold a Ph.D. or Master in computer science or computer engineering with advanced
knowledge of cellular networks, cloud architectures and virtualization techniques.
Advanced programming skills in C++/Java, XMPP.
Knowledge of SDN controllers is a plus.
Knowledge of French is not required.
The recruited candidate will collaborate with the members of the SNR team of the LIGM laboratory, more particularly with a professor, a lecturer and a doctoral student.
Interested candidates should send the following documents:
- Detailed CV including publications. Please, provide a link to the documents to download rather than attach them to your application email.
- Names and contact details of at least two referees, who are willing to provide detailed
recommendation letters about the candidate.
We talk about it on Twitter!