Summary
International Symposium on Antennas and Propagation
2010
Session Number:3TF3
Session:
Number:3TF3-4
PAPR Reduction Method Based on Significant-Bit Scrambling for MIMO Vector-Coding Systems
Yasuto Ando, Osamu Muta, Hiroshi Furukawa,
pp.-
Publication Date:2010/11/23
Online ISSN:2188-5079
Summary:
In high speed wireless transmission, inter-symbol interference (ISI) caused in a frequency-selective channel severely degrades bit error rate (BER) performance. As a method to overcome frequency selectivity in wireless channel, vector coding (VC) transmission has been studied [1]. The VC is a kind of code-division multiplexing (CDM) where eigenvectors of channel autocorrelation matrix are used as spreading code, i.e., the basic principle of VC is the same as the eigen-beam transmission in Multi-Input Multi-Output (MIMO) systems. The VC can be extended to MIMO transmission called MIMO-VC that achieves orthogonal parallel transmission over eigen-paths in frequency-selective MIMO channels, provided that channel state information (CSI) is known to both transmitter and receiver. Each eigen path has different channel gain and thus the VC needs to adopt an adaptive modulation to enhance transmission performance. A major drawback of the VC is that the transmit signal exhibits high peak-toaverage power ratio (PAPR) which causes nonlinear distortion and/or power efficiency degradation at power amplifier, similarly to OFDM and CDM systems. Therefore, it is required to reduce PAPR of the transmit signal in VC systems. As a solution to the PAPR problem, several techniques have been proposed such as clipping and filtering (C&F) and selected mapping (SLM) [2]. The C&F is a simple technique. However, it causes nonlinear distortion which destroys the orthogonal property of eigen channel in VC transmission and consequently degrades transmission performance. The SLM is a distortion-less PAPR reduction method which reduces the probability of high peak power occurrence by controlling the constellation data mapping of modulated symbols so as to minimize PAPR of the transmit signal. However, on the receiver side, it is needed to restore the original constellation data mapping to obtain information data. As a method to solve this problem, an SLM method with a self-synchronized scrambler (SLM-SS) has been proposed [2]. In SLM-SS, the transmit data sequence is randomized by a self-synchronized scrambler with a given generation polynomial using an initial bit pattern which is adaptively selected to minimize the PAPR. On the receiver side, original information data sequence is obtained by descrambling the received data, where a de-scrambler employs the same generation polynomial as the scrambler. However, when transmission errors occur, the abovementioned descrambling calculation causes error propagation which degrades BER performance. In this paper, we propose an SLM-SS method with the significant-bit scrambling for adaptive modulated MIMOVC transmission systems, where only most significant bits of I- and Q-phase of each modulated symbol are scrambled to reduce PAPR and thus BER degradation caused by error propagation is effectively mitigated, while achieving almost the same PAPR reduction performance as conventional SLM-SS method.