Traffic-Aware Resource Allocation and Feedback Design in Wireless Networks

Publication Type:



Supélec (2014)


Wireless systems are facing an increase in the data demands, and this trend is
expected to continue in the future. This increase is mostly due to demand of
video and data services. The most prominent approaches proposed to deal with
this problem, namely the use of multiple antennas and OFDMA modulations
(already part of the 3GPP LTE standards) and Small Cell Networks have mostly
been analyzed from a pure physical layer perspective, focusing on metrics like
total system throughput. However, the traffic pattern of video and data requests
as well as the individual requests of the users have to be also taken into account
when designing resource allocation algorithms. The objective of this thesis
is, therefore, to study the impact of physical layer resource algorithms (power
control, precoding, scheduling) and CSI feedback on the behaviour of the queues
of the users. In particular, we study the problems of precoding and power
control to regulate the behaviour of the users' queues in the interference channel,
as well as joint feedback/training and user selection and scheduling in order to
stabilize the queues for a large area of traffic demands in the MISO and OFDMA
broadcast channels. To this end, we use tools from heavy traffic asymptotic
modelling of communication networks and stochastic stability theory.