Resource Allocation in Wireless Networks: Energy Efficiency and Spectrum Assignment

Yuedong XU

INRIA Sophia Antipolis - Méditerranée


Résumé:

This presentation concentrates on the resource allocation of three types of wireless networks, IEEE 802.11 wireless networks, orthogonal frequency selective multiple access (OFDMA) networks and cognitive radio networks (CRN) that can be differentiated by their medium access control (MAC) mechanisms. The generalized formulation is to maximize specific objectives subject to the constraints of airtime, channel frequency as well as power budget. In IEEE 802.11 systems, one of our work falls in the class of energy efficiency issue. We considers how to reduce energy consumption by exploiting multi-rate diversity in single-cell 802.11 wireless networks. We have an important observation that ``probabilistic rate combination'' (PRC) in transmission can significantly reduce power consumption. To realize the PRC strategy, we propose a distributed algorithm that each node adjusts the combination strategy and determines the transmission probability based on the average interface queue length. In the OFDMA relay networks, we consider how to apply the practical network coding scheme via cross-layer optimization. Two efficient coding-aware relay strategies are presented to select forwarding paths with both fixed and dynamic power allocations. In order to exploit the network capacity in slow frequency selective fading channels, we propose channel-aware coding-aware subcarrier assignment algorithms for an arbitrary traffic pattern. In dynamic spectrum markets, we present an oligopoly pricing framework for dynamic spectrum allocation in which the primary users sell the excessive spectrum to the secondary users for monetary return. Two approaches, the strict constraints and the QoS penalty, are presented to model the realistic situation that primary users have limited spectrum capacities. The existence of unique Nash Equilibrium (NE) is proved in both cases. We further analyze the dynamic pricing strategies when the market information is incomplete for primary users.


[Yuedong XU]
INRIA Sophia Antipolis - Méditerranée