Noise Power Estimation in SC-FDMA Systems

Abstract

This article addresses the problem of noise power estimation for single-carrier frequency-division-multiple-access (SC-FDMA), which is used in long-term evolution (LTE). If channels are known, signal-to-noise ratio (SNR) can be derived from the noise power. A maximum-likelihood (ML) estimator is first derived to obtain a temporal noise power and then a minimum-variance unbiased (MVU) estimator is adopted to obtain an average noise power over the time-frequency plane. Since there is no analytical solution for the MVU estimator, an iterative approach based on fixed-point iteration (FPI) is employed to solve MVU estimation.