Section: Research Program

Agile Radio Resource Sharing

Participants : Jean-Marie Gorce, Claire Goursaud, Nikolai Lebedev, Perlaza Samir, Leonardo Sampaio-Cardoso.

The second research axis is dealing with the resource sharing problem between uncoordinated nodes but using the same (wide) frequency band. The agility represents the fact that the nodes may adapt their transmission protocol to the actual radio environment. Two features are fundamental to make the nodes agile : the first one is related to the signal processing capabilites of the software radio devices (middle circle in Fig 3 ), including modulation, coding, interference cancelling, sensing... The set of all available processing capabilites offers the degrees of freedom of the system. Note how this aspect relies on the two other research axes: radio front-end and radio programming.

But having processing capabilities is not enough for agility. The second feature for agility is the decision process, i.e. how a node can select its transmission mode. This decision process is complex because the appropriateness of a decision depends on the decisions taken by other nodes sharing the same radio environment. This problem needs distributed algorithms, which ensure stable and efficient solutions for a fair coexistence.

Beyond coexistence, the last decade saw a tremendous interest in cooperative techniques that let the nodes do more than coexisting. Of course, cooperation techniques at the networking or mac layers for nodes implementing the same radio standard are well-known, especially for mobile ad-hoc networks, but cooperative techniques for sdr nodes at the phy layer are still really challenging. The corresponding paradigm is the one of opportunistic cooperation, let us say on-the-fly, further implemented in a distributed manner.

We propose to structure our research into three directions. The two first directions are related to algorithmic developments, respectively for radio resource sharing and for cooperative techniques. The third direction takes another point of view and aims at evaluating theoretical bounds for different network scenarios using Network Information Theory.

The second research axis is dealing with multi-user communications focusing on resource sharing between uncoordinated nodes but using the same spectral resources. The agility relies on the nodes capability to adapt their transmission protocol to the actual radio environment. Centralized and decentralized approaches are investigated and the group is targeting fundamental limits as well as feasible and even practical implementations.

To make agile radio resource sharing a reality, two research directions are investigated. The first one aims at increasing the signal processing capabilities of software radio devices (middle circle in Fig 3 ), including modulation, coding, interference cancelation, sensing. The objective is to broaden the set of available processing capabilities thus offering more degrees of freedom. Note how this aspect relies on the two other research axes: radio front-end and radio programming.

Processing capabilities is not enough for agility. The second research direction concerns the decision process, i.e. how a node can select its transmission mode. This decision process is complex because the appropriateness of a decision depends on the decisions taken by other nodes sharing the same radio environment. In some cases, centralized solutions are possible but distributed algorithms are often required. Therefore, the target is to find distributed solutions ensuring stability, efficiency and fairness. Beyond coexistence, the last decade saw a tremendous interest in cooperative techniques that let the nodes do more than coexisting. Of course, cooperation techniques at the networking or mac layers for nodes implementing the same radio standard are well-known, especially for mobile ad-hoc networks, but cooperative techniques for sdr nodes at the phy layer are still challenging. The corresponding paradigm is referred to as opportunistic cooperative transmissions. We structure our research into three directions:

• Establishing theoretical limits of cooperative wireless networks in the network information theory framework.

• Designing coding and signal processing techniques for optimal transmissions (e.g. interference alignment).

• Developing distributed mechanisms for distributed decision at layer 1 and 2, using game theory, consensus and graph modeling.