Thomas Hindelang

Iterative source and channel decoding for GSM

When transmitting speech signals, there is residual redundancy in the signal after source coding, due to the limited complexity of the coding algorithms and delay constraints. This redundancy is expressed in correlations inside one frame as well as in a time correlation of subsequent speech frames. The method of iterative channel and source decoding applied in this paper is based on exploiting this redundancy in terms of a priori knowledge of the source to improve decoding of the transmitted parameters. The used source a priori information is obtained and exploited directly on a parameter level. We show the application of this theory to a real world mobile communication system. Here GSM speech transmission was chosen, but the presented method could also he applied to any other system which leaves a certain amount of redundancy in the speech coded signal.

SC-FDMA Waveform Design, Performance, Power Dynamics and Evolution to MIMO

We study waveform design, performance and power dynamics of the localized form of the single-carrier frequency-division multiple access (SC-FDMA) considered for uplink in the 3GPP long-term evolution (LTE). A classical single-carrier scheme with root-raised-cosine filtering and two DFT-based schemes are considered, one with and another without a novel spectral expansion. It is shown how to design waveforms similar to the classical scheme by using this spectral expansion. The advantages are that now the transmitter and receiver can be fully specified in frequency domain and that any fraction of the total bandwidth can be assigned to a user. For constant-amplitude modulation (e.g. BPSK, QPSK), the power dynamics is reduced by 3 dB with spectral expansion. Of course, this is bought for 25% more bandwidth. The inherent power dynamics of complex QAM constellations covers part of this gain. Spectral expansion may sometimes be dispensable, e.g. for 16-QAM in the 3GPP LTE uplink. The evolution to MIMO follows directly from the receiver concept based on virtual antennas.

Quality enhancement of coded and corrupted speeches in GSM mobile systems using residual redundancy

There is often residual redundancy remaining in coded speech data, even if a powerful speech codec (e.g. the full rate coder used in GSM mobile communications) is employed. By using such redundancy together with the information provided by the channel decoder, such as soft output (L-value), the number of channel bits inverted by the decoder, or a cyclic redundancy check, the bit error rate can be further reduced and a more graceful degradation of speech quality can be achieved, especially under bad channel conditions. In this paper, we report on the study with regard to this aspect for GSM full rate speech transmission and error concealment. The algorithms developed can be easily implemented with a currently available DSP designed for GSM mobile phones.

A low complexity algorithm for frame synchronization in TDD-LCR system

Time synchronization is required in mobile terminals operating in any digital radio system to detect the radio frame start of the received signal. The main operation performed in the synchronization procedure is the correlation between the received signal and the known synchronization sequences transmitted in each frame. The UMTS standard for TDD Low Chip Rate (TDD-LCR) option (integrated by 3GPP in Release Four of the UTRA standard) defines 32 different synchronization codes (SYNC sequences). Each base station transmits one of these sequences always at the same position of a radio frame and the used sequence defines already a group of base stations to which this sequence belongs. In UMTS FDD or in the high chip rate TDD one common primary synchronization sequence is used which does not exist in the TDD-LCR. This forces the mobile set to calculate the correlation between the received signal and all 32 SYNC sequences and thus increases the computational complexity enormously. In this paper we propose a low complexity algorithm for the detection of the radio frame start and of the sequence number using an approximation with lowpass filtering and sample rate reduction in the frequency domain. The performance loss of this approximation is quite small and an implementation in currently available signal processors is possible. Furthermore, the proposed algorithm could be generalized to any case where a problem of computational complexity arises due to the calculation of the correlation between two signals.