In this paper, the iterative water-filling (IWF) algorithm is considered when the noise-plus-interference levels fed back from the receivers are quantized. A sufficient condition for this algorithm to have a unique Nash equilibrium is derived. This condition is shown to approach the corresponding condition for standard IWF when the quantization resolution or the transmission power are sufficiently high.

Additional Metadata
Keywords Convergence, Nash equilibrium, noncooperative games, nonnegative matrix, power allocation
Persistent URL dx.doi.org/10.1109/TSP.2012.2188715
Journal IEEE Transactions on Signal Processing
Citation
Gohary, R, & Yanikömeroǧlu, H. (2012). Convergence of iterative water-filling with quantized feedback: A sufficient condition. IEEE Transactions on Signal Processing, 60(5), 2688–2693. doi:10.1109/TSP.2012.2188715