The MOSAICO project proposes to design, implement and validate a global and multi-layer orchestration solution, able to control several underlying network programmability technologies to compose micro-services forming the overall network service, focusing on low-latency services.

The hypothesis made by the MOSAICO project is that multi-level and multi-technology orchestration of high- performance micro-services is the key to augment the security and QoS of networks to make them ready for immersive cyber-physical services. Such services have low-latency and secured constraints from two perspectives: from the user when sending control actions, and from the controlled remote device sending real-time high-quality live video. This imposes two flows with low- latency constraints but with different traffic patterns (small packets vs. large packets), while also having very high security requirements for obvious reasons (any alteration or degradation of the control actions or video streams may result in high damages in the real world). Many services share those characteristics but we can mention in particular two use-cases: telesurgery, when a surgeon can remotely control a robot in a hospital, and drone operation where users can remotely control the movement of their drone. Our goal is to bypass the current networking limitations preventing these services thanks to our contributions and make one of them real as a proof-of-concept.
 

Project organization

The project will follow an experimental research methodology organized in 4 tasks.
  • The first task will enable the identification and design of the micro-services and define the overall architecture.
  • The second task will consist of benchmarking the identified micro-services with the different network programmable technologies (container, SDN, P4-based) in an environment close to reality, in order to evaluate the performance of these programmable environments and assess to what extent they meet the strong requirements imposed by the project use-case.
  • The third task will be related to the orchestration of the full architecture, with the design and implementation of a multi-layer and multi-technology orchestrator. The latter will be able to select the right programmable technology and the appropriate location where to deploy micro-services, depending on the service requirements.
  • Finally, the fourth task will be related to the evaluation of the solution against the project use-case. A testbed with wired and wireless access networks, using the Open Air Interface (OAI) solution for taking into account the constraints and uncertainties of the 5G mobile networks, will be implemented. The use of real hardware equipment offering a P4 environment (versus soft-switches) to show the feasibility of the project solution will also be considered.
Task Objectives Subtask Leader Start Duration
Task 1: Identification of micro-services and global architecture Identification and featuring of the micro-services of the project use-case, namely the performance and security of low latency immersive services leveraging encrypted traffic Identification and design of micro-services Orange T0 24 months
Global Architecture Orange T0+6 36 months
Task 2: Development and Benchmark of the micro-services develop the selected micro-services and perform unitary tests to assess their performance regarding the supporting network-programmable technology chosen for each of them (as a micro-service, P4 function, Openflow function) Development of the micro-services CNRS-LORIA T0+6 18 months
Benchmark of the micro-services CNRS-LORIA T0+12 18 months
Task 3: Multi-layer Orchestration design and implementation of a multi-layer and multi-technology orchestrator. The latter will be able to select the right programmable technology and the appropriate location where to deploy micro-services, depending on the service requirements. Multi-layer Orchestration algorithms ICD-UTT T0+12 18 months
Multi-layer Orchestration implementation ICD-UTT T0+12 24 months
Task 4: testbed and Demonstrator evaluation of the solution against the project use-case. A testbed with wired and wireless access networks, using the Open Air Interface (OAI) solution for taking into account the constraints and uncertainties of the 5G mobile networks, will be implemented. Testbed setup Montimage T0+6 30 months
Integration and Evaluation Montimage T0+18 24 months

Partners

Partner Name First name Current position Role & responsibilities in the project (4 lines max) Contact
Orange MATHIEU Bertrand Senior Researcher Project Coordinator, Task 1 Leader Contributor to QoS, SDN, P4, Segment Routing and Testbed, ONAP bertrand2.mathieu@orange.com
DUGEON Olivier Senior Researcher olivier.dugeon@orange.com
CADRO Philippe R&D Engineer philippe.cadro@orange.com
TUFFIN Stéphane R&D Engineer stephane.tuffin@orange.com
OLLIVIER Cédric R&D Engineer cedric.ollivier@orange.com
Montimage MONTES DE OCA Edgardo CEO, Senior Researcher Task Leader 4, leader of the development of the micro-services and orchestrator, integration in the SW and HW testbeds edgardo.montesdeoca@montimage.com
MALLOULI Wissam Senior Researcher wissam.mallouli@montimage.com
NGUYEN Huu Nghia Research engineer huunghia.nguyen@montimage.com
ICD-UTT DOYEN Guillaume Associate Professor Task Leader 3, Contribution to Low-rate DoS detection, secure multi-criteria orchestration guillaume.doyen@utt.fr
COGRANNE Rémi Associate Professor remi.cogranne@utt.fr
PRODHON Caroline Associate Professor caroline.prodhon@utt.fr
AYED Samiha Associate Professor samiha.ayed@utt.fr
PLOIX Alain Associate Professor alain.ploix@utt.fr
NDJORE Boris Research Engineer kouame_boris.n_djore@utt.fr
CNRS-LORIA CHOLEZ Thibault Associate Professor Task Leader 2, contributor to micro-services architecture and security functions, network management and orchestration thibault.cholez@loria.fr
FESTOR Olivier Professor olivier.festor@loria.fr

Funding

The MOSAICO project is funded by French Nation Research Agency (ANR), project number ANR-19-CE25-0012-01. The project is also supported by partner UTT's co-funding from FEDER funds of Region Grand Est, France, project number CA0023762.