The optimization scheme for joint iterative detection and decoding of polar coded SCMA system

Highlights

• The PC-SCMA technology can enhance the performance of communication system.

• JIDD scheme can make full use of the intermediate messages to improve performance.

• S-JIDD can effectively reduce complexity and achieve ideal error performance.

• Adding CRC early stopping criterion can further reduce the BER of the system.

Abstract

Polar coding and sparse code multiple access (SCMA) technique are excellent candidates for 5G, and have a promising prospect in optical communication, the joint design of polar coded sparse code multiple access (PC-SCMA) system is important to enhance the overall system performance. In this paper, a serial joint iterative detection and decoding scheme is proposed by passing message between serial max-log message passing algorithm and soft cancellation (SCAN) algorithm, the strategy can accelerate the algorithm convergence speed by reducing the number of required iterations, decrease the computational complexity and achieve ideal bit error rate (BER) performance. In order to further improve the BER performance, a cyclic redundancy check (CRC) is added in the joint iterative process as an early stopping criterion to avoid decoding divergence caused by a low signal to noise ratio (SNR). Simulation results demonstrate that the addition of a termination mechanism can obtain an excellent performance gain compares with the conventional JIDD algorithm.

Introduction

In 2009, Erdal Arian proposed the polar codes, which is a channel coding theory that has been proved to be able to achieve the Shannon capacity in the binary-discrete memoryless channel (B-DMC) and has the advantage of low coding and decoding complexity [1]. In reference [2], an ultraviolet communication system based on polar coding was proposed. The experimental results show that the polar code scheme is superior to the LDPC scheme in improving bit error rate performance under the condition of low complexity. Reference [3] verified the error control performance can be obviously improved by polar codes with lower code rates in visible light communication (VLC) system. Sparse code multiple access (SCMA) system directly maps the incoming bits to a multi-dimensional codeword of an SCMA codebook set by combining multi-dimensional modulation design and sparse spreading spectrum, so as to solve the system overload of massive connectivity [4], [5], [6]. Reference [7] combined SCMA with orthogonal frequency division multiplexing (OFDM) modulation, and proposed a new generation of passive optical network (PON) architecture, which can provide 150% overload gain, and experiments verified that the power loss caused by dispersion can be ignored after 20 km and 80 km of fiber transmission. SCMA also used to realize real-time transmission of VLC multi- medical data, which can achieve higher bandwidth efficiency [8].

In previous separate detection and decoding schemes for the polar coded SCMA system, the SCMA detection algorithm is mainly the message passing algorithm (MPA)[9], [10]. There are mainly two kinds of schemes for polar decoding: soft-input-hard-output (SIHO) algorithm [11], [12] and soft-input-soft-output (SISO) algorithm [13], [14]. The SIHO can be successive cancelation (SC) algorithm and successive cancelation list (SCL) algorithm, reference [15] verified that SCL decoding algorithm can further improve the effective communication distance of ultraviolet communication system. The SISO such as belief propagation (BP) algorithm and soft cancellation (SCAN) algorithm. The design of a joint detection and decoding algorithm (IDD) which combines the message passing algorithm (MPA) and BP decoding can improve system performance by making full use of the internal messages [16]. BP decoding algorithm reduces greatly the decoding delay due to the use of parallel decoding, but the performance of bit error rate (BER) is poor and suffers high complexity. SCAN algorithm is also parallel propagation algorithms, but its convergence speed is faster and the BER performance is better. In reference [17], the authors proposed a joint iterative detection and decoding (JIDD) algorithm whose each outer iteration consists of MPA and SCAN algorithm. In order to improve the performance and reduce the computational complexity of the joint system, in this paper, we firstly propose a serial joint iterative detection and decoding(S-JIDD) scheme whose each outer iteration is performed between the serial max - log message passing algorithm and the soft cancellation (SCAN) algorithm. The algorithm can effectively speed up the convergence and save a large part of the storage space since the update of the intermediate variables are merged into the resource nodes updating process during the iteration. The method can also achieve an ideal BER performance while the iterations is small. During the process of iterative decoding, decoding divergence might be caused by a low SNR level [18]. To overcome this problem and further reduce the BER performance, a cyclic redundancy check (CRC) is added in the joint iterative process as an early stopping condition to lock in the correct codeword before decoding diverges and terminate the decoding iteration process. Simulation results demonstrate that the joint algorithm S-C-JIDD who combines the two proposed schemes can effectively improve the performance of BER while reducing the complexity of joint detection and decoding, which provides a good balance between BER and complexity.