Improved Channel Estimation Algorithms based on a Proposed Signal Model for PUCCH Format 3 in LTE-A

Abstract

SORTD (Space Orthogonal-Resource Transmit Diversity) has already become a better way of transmit diversity for PUCCH format 3 in LTE-A [1]. SORTD technology lies in the characteristic that each antenna port transmits signal by using different cyclic shift [2] to make orthogonal each other. Using the SORTD's feature of incoherence in different antennas, this paper proposes a special form of signal model. Based on this proposed model, we propose a low complexity channel estimator based on LS and cubic spline interpolation, and then we put a LS and DCT algorithm to improve the interpolation performance. Finally, for more accurate estimator we drive the third approach based on MMSE and DCT. A series of theoretical researches and simulation results show that all of them are suitable for the proposed signal model even in the environment of high-speed movement; especially the algorithms with DCT have a better performance. 228Copyright ⓒ 2015 SERSC Currently, channel estimation to use more interpolation algorithms consist of linear interpolation, curvilinear Interpolation and transform-domain Interpolation processing, while the algorithm based on discrete cosine transform (DCT) [5] has achieved outstanding performance in the OFDM system. For given a sequence of N points [6] [7], the principle of DCT is to extend the sequence of N points into the sequence of 2N points, thus reducing the high frequency component of the signal and making interpolation more effect. Based on the analyses above, this paper proposes three channel estimation algorithms appropriate for in the LTE-A system, PUCCH adopting SORTD transmits uplink control information, namely the three algorithms, respectively, are based on LS and cubic spline interpolation, LS and DCT interpolation and MMSE and DCT interpolation, which are applied to PUCCH format 3 on the two transmitting portsIn order to be closer to the real environment of mobile communication, we have simulated in the case of ETU( Extended Typical Urban) where the mobile speed are 3 km/h and 350 Km/h, respectively, which represents the slowest speed namely people walking and the fast speed namely high-speed rail on the ground. The result shows that the third algorithm has the better performance.