Channel Estimation in Uplink of Long Term Evolution

Abstract

Long Term Evolution is considered to be the fastest spreading communication standard in the world.To live up to the increasing demands of higher data rates day by day and higher multimedia services,the existing UMTS system was further upgraded to LTE.To meet their requirements novel technologies are employed in the downlink as well as uplink like Orthogonal Frequency Division Multiple Access (OFDMA) and Single Carrier- Frequency Division Multiple Access (SC-FDMA).For the receiver to perform properly it should be able to recover athe transmittedadata accurately and this is done through channel estimation.Channel Estimation in LTE engages Coherent Detection where a prior knowledge of the channel is required,often known as Channel State Information (CSI).This thesis aims at studying the channel estimation methods used in LTE and evaluate their performance in various multipath models specified by ITU like Pedestrian and Vehicular.The most commonly used channel estimation algorithms are Least Squarea(LS) and Minimum MeanaSquare error (MMSE) algorithms.The performance of these estimators are evaluated in both uplink as well as Downlink in terms of the Bit Error Rate (BER).It was evaluated for OFDMA and then for SC-FDMA,further the performance was assessed in SC-FDMA at first without subcarrier Mapping and after that with subcarrier mapping schemes like Interleaved SC-FDMA (IFDMA) and Localized SC-FDMA (lFDMA).It was found from the results that the MMSE estimator performs better than the LS estimator in both the environments.And the IFDMA has a lower PAPR than LFDMA but LFDMA has a better BER performance.