Cooperation in Cellular Networks: Taking a new look at interference

David GESBERT - PRofesseur
Communication systems

Date: -
Location: Eurecom

As it is well known to wireless system designers, fading limits the reliability of point to points radio links while interference limits their spatial reusability. As the performance of practical point to point links is now approaching its fundamental log(1+SNR) rate limit, increasing further the capacity of wireless systems will require the adoption of a new truly multi-terminal view of the network, i.e. a view where the impact of interference is fully accounted for in the design of transmission algorithms. This is in contrast with traditional designs, where interference has been dealt with a combination of resource allocation protocols at the link layer (with soft to hard orthogonality constraints) together with robust coding and detection schemes at the PHY layer. Recent results in the domain of cooperative communications have revealed however that interference can be efficiently tackled, or even exploited, via a proper treatment at the transmitter side, involving the concept of coordination among the interfering transmitters, as in so-called network MIMO or interference alignement to name a few examples of popular methods. In this talk we emphasize the promise of such methods, yet also how the coordination gains come at a substantial price in terms of exchanging information across the devices. After giving an overview comparing such techniques, we are addressing the coordination gain/cost trade-off. Finally we look at some promising research avenues in the domain of distributed interference avoidance and exploitation. (Note: Based on a keynote speech at the ISWCS Conference, York, Sept. 2010)