Resource allocation in dense wireless communication networks

This PhD thesis focuses on interference mitigation techniques for wireless communication networks to limit detrimental effects of in-band interference. First, we consider cooperative communication systems and study the trade-off between cooperation benefits and increased level of interference. Cooperation in wireless networks is like a crowded cocktail party with a cacophony of conversations all around. The more people repeat the same information, the more likely you understand it. However, neighbour repeaters act as interferers which harm your understanding. We propose to coordinate and adapt the activation of cooperation and the resource allocation of neighbour cells to time, frequency and space variations of communication context. Second, we propose to classify interference a destination senses on a given frequency band by differentiating three regimes of interference: noisy, intermediate and very strong. This classifier ensures an adaptive and effective processing of in-band interference adapted to time-varying nature of channel. Then, we combine this classifier with QoS constraints to derive centralized and distributed algorithms for power allocation. Both approaches aim at allocating the minimal transmit power vector while meeting QoS requirements of each user, whatever the communication scenario may be. Our simulations show how an adaptive handling of in-band interference may notably reduce the power budget without affecting transmission reliability.