att_abstract={{We study optimal power allocation for a set of parallel channels when transmitter has a
limited knowledge of the Channel State Information (CSI). This case is typical for
mobile OFDMA systems like LTE and LTE-A where transmitter relies on CSI feedback from a
receiver which is subject to measurement errors, sparse sampling, quantization errors
(especially noticeable in MIMO systems), processing and transmission delays.  Based upon
per-channel targeted Packet Error Rates (PER) we construct a power allocation scheme
that maximizes the sum capacity. The optimal algorithm can be cast into a modified 3-D
(Volumetric) version of the Water-Filling algorithm.  Additionally, the solution
obtained here improves performance of the real-time delay sensitive traffic such as
voice and real-time video by favoring the traffic with lower target PER. Simulations
show that the gain in sum capacity over traditional Water-Filling solution can reach as
high 90% in some cases.
	att_authors={lr7762, rx2732},
	att_copyright={{Academy Publisher}},
	att_copyright_notice={{The definitive version was published in  2012   {, Volume 7}}{{, Issue 9}}{{, 2012-09-15}}
	att_tags={OFDM,  RRM,  Waterfilling,  CSI},
	author={Leonid Razoumov and Robert Miller},
	institution={{Journal of Communications (JCM)}},
	title={{Power Allocation under Transmitter Channel Uncertainty and QoS Constraints: Volumetric Water-Filling Solution}},