“…13 except for the fact that the SN includes four more 1B ECNs (i.e., the HB(10, 4, 2) CN architecture). The T-MM performance of a code of same length and coding rate provided in [8] is also presented in Fig. 15.…”
Section: A Performancementioning
confidence: 99%
“…However, the performance gap may come from the use of different parity check matrices. In [8] the authors consider a d c = 24, d v = 3 code with an array dispersion construction while we use a d c = 16, d v = 2 code constructed using optimized coefficients for the parity checks [25].…”
Section: A Performancementioning
confidence: 99%
“…For a fair comparison with the T-MM architecture [8] [9], the full decoder is considered since the T-MM-base decoder uses a different VN architecture. We implemented a complete decoder architecture based on an HB(10,4,2) CN over GF(64) which is connected to 16 VNs.…”
Section: Comparison With the T-mm Architecturementioning
confidence: 99%
“…The synthesis results are given in Table VI. We considered the same clock frequency as [8] F clk = 350 MHz to compute the HB efficiency and throughput for a fair comparison. Those results triggers us few comments, first, this comparison is not really fair for T-MM since a higher variable and check node degree is used for T-MM (d v = 3, d c = 24) compared with HB architecture (d v = 2, d c = 16).…”
Section: Comparison With the T-mm Architecturementioning
confidence: 99%
“…The Trellis-EMS (T-EMS) introduced in [6] avoids the long latency of the FB computation but its hardware complexity highly increases with q when a parallel implementation is considered. The complexity of the T-EMS was reduced with the one-minimum T-EMS [7] and the Trellis Min-Max (T-MM) [8] [9] algorithms.…”
This paper proposes a unified framework to describe check node architectures of Non-Binary LDPC decoders. Forward-Backward, Syndrome-Based and Pre-sorting approaches are first described. Then, they are hybridized in an effective way to reduce the amount of computation required to perform a check node. This work is specially impacting check nodes of high degrees (or high coding rates). Results of 28 nm ASIC post-synthesis for a check node of degree 12 (i.e. code rate of 5/6 with a degree of variable equal to 2) are provided for NB-LDPC over GF(64) and GF(256). While simulations show almost no performance loss, the new proposed Hybrid implementation check node increases the hardware and the power efficiency by a factor of six compared to the classical Forward-Backward architecture. This leads to the first ever reported implementation of a degree 12 check node over GF(256) and these preliminary results open the road to high decoding throughput, high rate, and high order Galois Field NB-LDPC decoder with reasonable hardware complexity.
“…13 except for the fact that the SN includes four more 1B ECNs (i.e., the HB(10, 4, 2) CN architecture). The T-MM performance of a code of same length and coding rate provided in [8] is also presented in Fig. 15.…”
Section: A Performancementioning
confidence: 99%
“…However, the performance gap may come from the use of different parity check matrices. In [8] the authors consider a d c = 24, d v = 3 code with an array dispersion construction while we use a d c = 16, d v = 2 code constructed using optimized coefficients for the parity checks [25].…”
Section: A Performancementioning
confidence: 99%
“…For a fair comparison with the T-MM architecture [8] [9], the full decoder is considered since the T-MM-base decoder uses a different VN architecture. We implemented a complete decoder architecture based on an HB(10,4,2) CN over GF(64) which is connected to 16 VNs.…”
Section: Comparison With the T-mm Architecturementioning
confidence: 99%
“…The synthesis results are given in Table VI. We considered the same clock frequency as [8] F clk = 350 MHz to compute the HB efficiency and throughput for a fair comparison. Those results triggers us few comments, first, this comparison is not really fair for T-MM since a higher variable and check node degree is used for T-MM (d v = 3, d c = 24) compared with HB architecture (d v = 2, d c = 16).…”
Section: Comparison With the T-mm Architecturementioning
confidence: 99%
“…The Trellis-EMS (T-EMS) introduced in [6] avoids the long latency of the FB computation but its hardware complexity highly increases with q when a parallel implementation is considered. The complexity of the T-EMS was reduced with the one-minimum T-EMS [7] and the Trellis Min-Max (T-MM) [8] [9] algorithms.…”
This paper proposes a unified framework to describe check node architectures of Non-Binary LDPC decoders. Forward-Backward, Syndrome-Based and Pre-sorting approaches are first described. Then, they are hybridized in an effective way to reduce the amount of computation required to perform a check node. This work is specially impacting check nodes of high degrees (or high coding rates). Results of 28 nm ASIC post-synthesis for a check node of degree 12 (i.e. code rate of 5/6 with a degree of variable equal to 2) are provided for NB-LDPC over GF(64) and GF(256). While simulations show almost no performance loss, the new proposed Hybrid implementation check node increases the hardware and the power efficiency by a factor of six compared to the classical Forward-Backward architecture. This leads to the first ever reported implementation of a degree 12 check node over GF(256) and these preliminary results open the road to high decoding throughput, high rate, and high order Galois Field NB-LDPC decoder with reasonable hardware complexity.
This paper presents an ultra-high-throughput decoder architecture for NB-LDPC codes based on the Hybrid Extended Min-Sum algorithm. We introduce a new processing block that updates a check node and its associated variable nodes in a fully pipelined way, thus allowing the decoder to process one row of the parity check matrix per clock cycle. The work specifically focuses on a rate 5/6 code of size (N, K) = (144, 120) symbols over GF(64). The synthesis results on a 28-nm technology show that for a 0.789 M NAND-gates complexity complexity, the architecture reaches a decoding throughput of 0.9 Gbps with 30 decoding iterations. Compared to the 5G binary LDPC code of the same size and code rate, the proposed architecture offers a gain of 0.3 dB at a Frame Error Rate of $$10^{-3}$$
10
-
3
.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.