CVR: A Continuously Variable Rate LDPC Decoder Using Parity Check Extension for Minimum Latency
- PDF / 734,284 Bytes
- 8 Pages / 595.276 x 790.866 pts Page_size
- 55 Downloads / 190 Views
CVR: A Continuously Variable Rate LDPC Decoder Using Parity Check Extension for Minimum Latency Sina Pourjabar 1
&
Gwan S. Choi 1
Received: 15 November 2019 / Revised: 28 August 2020 / Accepted: 7 September 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract This paper presents a novel IEEE 802.16e (WiMAX) based decoder that performs close to the 5G code but without the expensive hardware re-development cost. The design uses an extension of the existing WiMAX parity check code to reduce the initial decoding latency and power consumption while keeping the decoder throughput at maximum. It achieves similar Frame Error Rate (FER) compared to 5G (0.1 dB off), and most notably the error curves trend down like 5G instead flooring. At FER= 10−3 there is 0.1 dB gain in the FER code performance compared to WiMAX. An implementation of the design is a modified version of the existing fully-parallel WiMAX decoder that supports multi-rate codeword size and reduces the initial latency by 33%. Additionally, for SNR greater than 3 dB, decoding only the shorter code reduces the power consumption by 34%. Keywords Decoder architecture . Parity-check matrix (PCM) extension . Low latency . Low-density parity-check (LDPC) codes . Very large-scale integration (VLSI)
1 Introduction Low-density parity-check (LDPC) codes were first introduced by Gallager in 1960 [1]. One major advantage of these codes is their fully parallel architecture that significantly improves the throughput in theory. However, due to the complex structure of these codes, they were difficult to implement until the rediscovery of these codes in 1996 by MacKay [2]. With the further increase of the data rate in newer telecommunication protocols such as 5G NR, LDPC codes are playing the key role in reaching the Shannon channel capacity while allowing more flexibility on power consumption. A disadvantage known to block codes such as LDPC codes in comparison to convolutional codes is that in block codes the decoder should stall until the whole frame is received before starting to decode, whereas in convolutional codes the decoder can decode on the fly. The mentioned stall time is known as latency. The stall time becomes more evident when the
* Sina Pourjabar [email protected] Gwan S. Choi [email protected] 1
Department of Electrical and Computer Engineering, Texas A&M University, College Station, TX 77843, USA
decoder has just started decoding the first frame and there is no previous frame that the pipelined structure can decode; and therefore, there is no output. [3] showed that convolutional codes outperform LDPC codes in terms of latency and bit error rate (BER) for short length codes. However, for code lengths greater than 500 bits, LDPC performance starts to dominate. In other studies, quasi-cyclic LDPC convolutional codes (QC-LDPC-CC) and spatially coupled LDPC convolutional codes were evaluated in [4, 5]. Both studies showed improvement in latency and BER. In [4] latency of the code based on the sum-product (SP) decoding was cal
Data Loading...
