Improving the performance of digital radio mondiale plus (DRM+) by LDPC channel coding

Abstract

The performance of a DRM+ broadcast system using a modified Digital Video Broadcasting (DVB-T2/S2) Low Density Parity-Check (LDPC) code is compared against the standardized system, which uses convolutional coding. By using this modified LDPC code the datarate can be increased by 33 % at a similar bit error rate (BER) performance. Alternatively, the broadcast coverage can be increased. For the purpose of integration, a signaling scheme is proposed, which enables the broadcaster to decide between convolutional coding and LDPC coding without affecting the compatibility of the system. The presented approach allows for using existing DVB-T2/S2 LDPC codec implementations inside a DRM+ broadcast chain.Index Terms-DRM+, LDPC codeword puncturing, LDPC channel coding, VHF channel simulation, DVB-T2/S2 I. MOTIVATION DRM+ is a new standard for digital audio broadcasting in the very high frequency (VHF) domain. In 2004 the design of the system was initiated in order to supplement FM radio and Digital Audio Broadcasting (DAB) in the VHF band. The DRM+ system specification was published by the European Telecommunications Standards Institute (ETSI) in 2009 [1].Like most of the existing digital audio broadcasting standards DRM+ uses a punctured convolutional channel code, which can optionally be combined with a Reed Solomon code for packet data transmissions. Albeit having a moderate decoding complexity, the coding performance of convolutional codes is rather limited compared to state-of-the-art channel coding schemes.In contrast to DRM+, the second generation of Digital Video Broadcasting (DVB) standards use enhanced Forward Error Correction (FEC) mechanisms like LDPC coding for both satellite DVB-S2 [2] as well as terrestrial DVB-T2 broadcasting [3] in order to increase the system performance.In the following sections we will show that by using these LDPC codes inside a DRM+ transmission chain, a significant gain in coding performance can be achieved without violating the framing of the system. II. PRIOR ARTConvolutional channel coding in DRM+ is almost equal to channel coding in DRM30, the short wave counterpart of DRM+. In [4] the performance of a punctured turbo-code was compared against the performance of the convolutional code using a 16 QAM alphabet for DRM30. In [5] the applicability of turbo codes for DRM30 was evaluated using iterative decoding techniques in a diversity reception scenario.Our work targets the performance of LDPC codes within the DRM+ specific VHF channel environment. III. DIGITAL RADIO MONDIALE PLUSIn order to provide a basic understanding of the DRM+ system architecture, the key-parameters of the transmission chain are depicted in the subsequent sections. A. OFDM FramingThe OFDM framing in DRM+ is strictly hierarchical. The top level framing unit is the transmission super frame (TSF) which contains a sequence of N TF = 4 transmission frames (TF) as shown in Figure 1. Each TF wraps N s = 40 OFDM symbols. With an OFDM symbol duration of T s = 2.5 ms, the transmission duration of a TF is T s · N s = 100 ...

DRM+ performance measurement in laboratory test environment

“…One super frame is built by four frames. DRM+ uses convolutional coding with a mother code rate of 1/6 and a constraint length of 7, where the generated code is terminated by an all-zero input sequence [9]. Different puncturing patterns can be applied in order to select the desired output code rate.…”