Outage Behavior of Slow Fading Channels With Power Control Using Partial and Erroneous CSIT

Abstract

The topic of this study is the outage behavior of multiple-antenna slow fading channels with quantized feedback and partial power control. A fixed-rate communication system is considered. It is known from the literature that with error-free feedback, the outage-optimal quantizer for power control has a circular structure. Moreover, the diversity gain of the system increases polynomially with the cardinality of the power control codebook. Here, a similar system is studied, but when the feedback link is error-prone. We prove that in the high-SNR regime, the optimal quantizer structure with noisy feedback is still circular and the optimal Voronoi regions are contiguous non-zero probability intervals. Furthermore, the optimal power control codebook resembles a channel optimized scalar quantizer (COSQ), i.e., the Voronoi regions merge with erroneous feedback information. Using a COSQ, the outage performance of the system is superior to that of a no-feedback scheme. However, asymptotic analysis shows that the diversity gain of the system is the same as a no-CSIT scheme if there is a non-zero and non-vanishing feedback error probability. Index termsOutage probability, noisy feedback, channel optimized quantizers, bit-mapping, diversity.

Multiuser diversity in correlated Rayleigh-fading channels

“…It is assumed that each receiver has perfect CSI about its corresponding channel gain which is an acceptable assumption in quasi-static condition, e.g., [17,18,[34][35][36][37][38]. However, the transmitter may be provided with imperfect (Section 3) or perfect (Section 4) CSI about the fading channels.…”

“…However, as stated in the following, the water-filling properties imply that higher powers are allocated to the higher quantization regions (see, e.g., (14), (19) and [34][35][36][37][38]). Therefore, the SNR increases with the quantization index and scheduling based on the quantization indices works the same as scheduling based on the SNRs.…”

“…which, as discussed in [34][35][36][37][38], [[41], Section 9.4], is a convex problem in terms of transmission powers T i . Therefore, the optimal transmission powers can be determined based on the Lagrange multiplier function ϒ =R − λT which leads to the water-filling equations…”

“…A simple average rate optimization algorithm: In contrast to transmission power parameters, the power-limited average rate optimization problem of quantized CSI-based systems, e.g., (13), is not a convex optimization problem in terms of quantization parametersĝ i ,g i ∀ i [34][35][36][37][38]. Therefore, although implementable, gradient-based algorithms are not efficient in determining the optimal quantization parameters.…”

A dynamic resource allocation model along with adaptive power control in a multi-rate direct sequence CDMA with MIMO

“…However these studies [12,13,14] do not the leverage the bene fits of power control while allocating data rates for max imum throughput. Efficient power control is of great imp0l1ance in COMA system to minimize the outage probability and optimize the system performance [15]. So, the resource allocation strategies should involve appropriate power control algorithms to achieve incre mental capacity enhancement.…”