This contract is running from 2009 to 2011. I am the representative for INRIA in the consortium.
Summary
The need for bandwidth and the incitation to reduce power consumption leads to the reduction of cells' size in cellular networks. This allows reducing the distance between mobile and base station thus increasing the capacity. This reduction implies problems for cells organization and network deployment, especially due to the cells interference. The classical solutions used in cellular networks are not satisfactory in this context, due to the scale change. The density of base stations needs also to re-design the backhauling networks such to support a huge capacity and huge variability in traffic.
The ECOSCells (Efficient Cooperating Small Cells) project aims at developing the algorithms and solutions required to allow Small Cells Network (SCN) deployment. The theoretical studies will define the models needed to understand the behaviour of radio channels, and will define the algorithms which will allow to exploit the diversity (user, spatial, interferences, etc.) of such networks.
The novelty of the project is not to consider the interferences as a drawback anymore, but to exploit these in order to offer an optimal resource usage. The algorithms will be based on most recent development in distributed algorithms, game theory, reinforcement learning. The architecture and the algorithms for the backhauling network will be proposed. The high density of a SCN (one base station every 10 or 100 meters) needs to consider the backhaul as a dynamic system which has to face to a large number of variability: traffic, links quality, outage, etc. The developed solutions in the project will be validated thanks to simulations and will be implemented on the top of an industrial development platform with the objective to deliver a prototype of SCN base station. Several instances of the prototype will be used to conduct an experimental validation.
The SCN envisaged in ECOSCells provide a transparent plug-and-play infrastructure, which can be deployed in an instantaneous manner and which can evolve on its own (energy and configuration-wise), without any human intervention. However, the theoretical foundation of small cells networks faces much broader constraints:
- The system changes rapidly and the small cells network needs to adapt fast and even anticipates future evolutions.
- As density of base stations increases, interference becomes bottleneck. Classical techniques for frequency and space reuse or power control are not able to resolve themselves the complexity of the system.
- The co-existence of different systems may lead to conflicted communications, then decisions have to be made based on such distributed information.
Organization and objectives
The work is composed of 6 tasks.
The Task 1 will gather the activities related to coordination of the activities (Sub-Task 1.1) as well as the efforts for dissemination and valorisation (ST1.2). It will aim of achieving efficiency management and high quality dissemination and will also support the valorisation of the outcome of the project.
In the Task 2, the foundations of the work will be established. The four sub-tasks will define the models and the architectures that will be adopted in the rest of the project. It will output a general framework for the other tasks.
The Task 3 will address the design and performance analysis of algorithms for a dynamic allocation of the resources, which are needed to achieve an efficient, fair, and environmental-aware distribution of the resources in SCN.
The key point of the backhauling of the SCN will be addressed in the Task 4. It is dedicated to the development of optimization algorithms, distributed,
hopefully localized, protocols, and theoretical bounds for the access network infrastructure.
The Task 5 is dedicated to the prototyping activities. Its objective is to release a SCN Base Station including the algorithms and solutions designed in Task 3 and 4.
The output of the Task 5 will be taken as one the inputs of the Task 6 where the experimental validation will be conducted. A first validation reference will be done y simulation and will help to define the experimental process that will conducted, in order to demonstrate the value and the efficiency and the selected and developed solution.
These are some of the issues that this project addresses and intends to solve. Other issues are related to mobility management, resource allocation, multihoming, cross-layer optimization, and performance tuning. In particular, we are interested in:
- Focused on optimization and fairness issues that concern the cell size as well as the location of base stations o Soft handover and hard handover in small cells
- Establishing trade-offs between centralized and distributed spectrum access for small cells networks
- Establishing an intelligent adaptation of medium access control (e.g. energy adaptive, interference adaptive, ect.) in dense self-organizing networks.
- Create and analyze distributed algorithms exploiting cross-layer design. To exploit the full potentials of different layers (e.g. logical vs. physical), cross-layer design intrinsically requires cooperation and flow of information and know how among experts of different layers.
- Apply advanced game theoretical tools such as stochastic dynamic games and games with partial information to cross-layer design in cooperative and non-cooperative scenarios. In fact, cross-layer optimization algorithms for wireless networks are heavily based on decomposition techniques in optimization and distributed approaches in game theory.
- Demonstrating the business value of SCN concept and the performance of algorithms, thanks to a prototype based on industrial platform and associated simulation and experiment based validation.
ECOSCells outcome will deliver the means to exploit the density of small cells : innovative concepts and algorithms on one hand, simulation results and equipments prototype demonstrating their validity on the other hand. Ambition is to address the SCN concept as a whole, from theory to implementation, allowing a fast market acceptance.
Partners
The consortium gathers the 2 mains industrial groups inthetelecommunicationdomain,together with 3 leading SME and 6 research institutes:
- Alcatel-Lucent Bell Labs (leader)
- INRIA : Teams involved are