Wideband resource allocation for future cellular networks

Toufik, Issam

Most of wireless communication systems, including cellular ones, are converging to wideband solutions (OFDMA, MIMO-OFDMA like) with dynamic channel allocation. The optimal allocation, in terms of system throughput, uses Multiuser Diversity concept which is known to allow high system performance but results on a system highly unfair. Also, it is widely believed that guaranteeing users’ requested QoS comes at the cost of a significant capacity penalty. We show in this presentation that this belief is not true for wideband systems with intelligently allocated resources. We first consider the single antenna case and show through numerical results that both optimal and orthogonal signaling with delay limited rates (i.e. zero-outage rates) achieve performance approaching that of the optimal unfair strategy (i.e. maximizing the system throughput). For Multiple antennas, we consider the downlink of a simple multi-user wideband system where only the knowledge of channel gain amplitudes is required. We propose different hard fairness and delay limited rate allocation algorithms. In a single cell system and for symmetric users, the proposed schemes perform almost as well as the well kwon opportunistic scheduling with random beamforming but the gap with the optimal DPC performance is considerable. This gap decreases with increasing number of sub-channels. Our system has the advantage of simplicity, no need for channel phase estimates and fairness. Distributed antenna benefit in both SE and fairness in multi-cell systems is highlighted. Another important issue considered is the feedback load. We present different feedback reduction algorithms for wideband MIMO systems.

Communication systems
Eurecom Ref:
© ENST Paris. Personal use of this material is permitted. The definitive version of this paper was published in Thesis and is available at :
See also:

PERMALINK : https://www.eurecom.fr/publication/2056