Spectrum sharing employs dynamic allocation of frequency resources for a more efficient use of the radio spectrum. Despite its appealing features, this technology inevitably complicates the synchronization task, which must be accomplished in an environment that may involve interference. This paper considers the uplink of an orthogonal frequency-division multiple-access (OFDMA)-based spectrum sharing system and provides solutions for estimating the frequency and timing errors of multiple unsynchronized users. In doing so, we exploit suitably designed training blocks and apply maximum likelihood methods after modeling the interference power on each subcarrier as a random variable with an inverse gamma distribution. The resulting frequency estimator turns out to be the extension to a multiuser scenario of a scheme that has previously been proposed for single-user spectrum-sharing systems. Timing recovery is more challenging and leads to a complete search over a multidimensional domain. To overcome such a difficulty, two alternative approaches are proposed. The first one relies on a simplifying assumption about the interference distribution in the frequency domain, while the second scheme operates in an iterative fashion according to the expectation-maximization algorithm.
Uplink Synchronization in OFDMA-Spectrum-Sharing Systems
SANGUINETTI, LUCA;MORELLI, MICHELE;
2010-01-01
Abstract
Spectrum sharing employs dynamic allocation of frequency resources for a more efficient use of the radio spectrum. Despite its appealing features, this technology inevitably complicates the synchronization task, which must be accomplished in an environment that may involve interference. This paper considers the uplink of an orthogonal frequency-division multiple-access (OFDMA)-based spectrum sharing system and provides solutions for estimating the frequency and timing errors of multiple unsynchronized users. In doing so, we exploit suitably designed training blocks and apply maximum likelihood methods after modeling the interference power on each subcarrier as a random variable with an inverse gamma distribution. The resulting frequency estimator turns out to be the extension to a multiuser scenario of a scheme that has previously been proposed for single-user spectrum-sharing systems. Timing recovery is more challenging and leads to a complete search over a multidimensional domain. To overcome such a difficulty, two alternative approaches are proposed. The first one relies on a simplifying assumption about the interference distribution in the frequency domain, while the second scheme operates in an iterative fashion according to the expectation-maximization algorithm.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
