Gou Hosoya

Copyright © 2020 The Institute of Electronics, Information and Communication Engineers. We propose a coding/decoding strategy that surpasses the symmetric information rate of a binary insertion/deletion (ID) channel and approaches the Markov capacity of the channel. The proposed codes comprise inner trellis codes and outer irregular low-density parity-check (LDPC) codes. The trellis codes are designed to mimic the transition probabilities of a Markov input process that achieves a high information rate, whereas the LDPC codes are designed to maximize an iterative decoding threshold in the superchannel (concatenation of the ID channels and trellis codes).

Copyright © 2020 The Institute of Electronics, Information and Communication Engineers Over the past two decades, irregular low-density parity-check (LDPC) codes have not been able to decode information corrupted by insertion and deletion (ID) errors without markers. In this paper, we bring to light the existence of irregular LDPC codes that approach the symmetric information rates (SIR) of the channel with ID errors, even without markers. These codes have peculiar shapes in their check-node degree distributions. Specifically, the check-node degrees are scattered and there are degree-2 check nodes. We propose a code construction method based on the progressive edge-growth algorithm tailored for the scattered check-node degree distributions, which enables the SIR-approaching codes to progress in the finite-length regime. Moreover, the SIR-approaching codes demonstrate asymptotic and finite-length performance that outperform the existing counterparts, namely, concatenated coding of irregular LDPC codes with markers and spatially coupled LDPC codes.

© 1965-2012 IEEE. Racetrack memories (RMs) are prone to alignment faults called position errors (PEs), which manifest as insertions and deletions of stored data bits. Conventional coding schemes for PEs demonstrate the promising results by employing low-density parity-check (LDPC) codes with an iterative detection and decoding algorithm. However, the computational complexity of detection is relatively high. In this article, for channels with PEs, we present a new coding scheme that can effectively decode corrupted data, even if the detection is executed only once (i.e., a non-iterative detection and decoding scenario). The proposed code consists of a concatenation of an inner 2-D marker code, which is specialized for PEs to mitigate the effect of insertion and deletion (ID) errors, and an outer irregular LDPC code. We also provide tractable design methodologies for these constituent codes. First, we identify the 2-D-marker code structures that offer higher achievable information rates in a non-iterative scenario and then optimize irregular LDPC codes to ensure good decoding properties. Through asymptotic-performance analysis and finite-length simulation, we confirm the effectiveness of the proposed coding scheme. Ultimately, the proposed coding scheme has the capability to reduce code rate loss and to provide excellent decoding performance under a non-iterative scenario, which also helps in understanding the reliability of RM when a low-complexity decoding algorithm is used to correct ID errors caused by PEs.

Copyright © 2020 The Institute of Electronics, Information and Communication Engineers. For insertion and deletion channels, there are many coding schemes based on low-density parity-check (LDPC) codes, such as spatially coupled (SC) LDPC codes and concatenated codes of irregular LDPC codes and markers. However, most of the previous works have problems, such as poor finite-length performance and unrealistic settings for codeword lengths and decoding iterations. Moreover, when using markers, the decoder receives log-likelihood (LLR) messages with different statistics depending on code bit position. In this paper, we propose a novel concatenation scheme using protograph-based LDPC code and markers that offers excellent asymptotic/finite-length performance and a structure that controls the irregularity of LLR messages. We also present a density evolution analysis and a simple optimization procedure for the proposed concatenated coding scheme. For two decoding scenarios involving decoding complexity, both asymptotic decoding thresholds and finite-length performance demonstrate that the newly designed concatenated coding scheme outperforms the existing counterparts: the irregular LDPC code with markers, the SC-LDPC code, and the protograph LDPC code, which is optimized for an additive white Gaussian noise channel, with markers.

© 2019 IEEE. Over the past two decades, irregular low-density parity-check (LDPC) codes have been hardly able to decode information corrupted by insertion and deletion (ID) errors without markers. Surprisingly, in this paper, we bring to light the existence of irregular LDPC codes that approach the theoretical limit of the channel with ID errors even without markers. These codes contain high fractions of low-degree check nodes that do not appear in irregular codes for other channels. This motivates us to investigate the contribution of low-degree check nodes to correcting ID errors. The investigation provides the following interesting result: degree-2 check nodes are critical to approaching the theoretical limit even without markers, codes with only degree-3, 4, or more check nodes provide moderate decoding performance, and codes with only degree-5 or more check nodes can hardly correct ID errors. Finally, we present simulation results that confirm the excellent decoding performance of the irregular codes without markers.

An all-optical AND gate using photonic-crystal quantum dot semiconductor optical amplifiers is designed and its performance is evaluated. The input–output characteristics of the gate are simulated using a rate equation model and it is found that the gate can achieve a maximum of ∼9 dB extinction ratio at 160 Gb/s. The proposed gate is compared with quantum dot semiconductor optical amplifiers AND gate to evaluate its effectiveness with regard to signal quality, device size, and power consumption.

We thoroughly explore the characteristics of an ultrafast all-optical NOR gate for 160 Gb/s return-to-zero Gaussian data signals using a single quantum-dot semiconductor optical amplifier (QD-SOA) and an optical filter (OF). In this proposed scheme, we employ an optical clock signal as a probe in addition to data signals as pumps between which the Boolean NOR function is executed. By conducting numerical simulations, we investigate and evaluate the effects of various critical factors on the extinction ratio and Q2-factor. This enables us to specify the margins of clock wavelength, peak power of data and clock signals, current density, electron relaxation time from the excited state to the ground state, linewidth enhancement factor, small signal gain of QD-SOA, OF bandwidth and order, the permissible extent of arrival time difference between data signals and clock, and the effect of amplified spontaneous emission. Moreover, we demonstrate that the proposed device can be applied to a multiple-input NOR gate. The results show that the proposed NOR gate can be achieved with both logical correctness and high quality when the specified conditions are satisfied.

We propose an ultrafast all optical NOR gate for 160 Gb/s return-to-zero Gaussian data signals using a single quantum-dot semiconductor optical amplifier (QD-SOA) and optical filter (OF). By conducting numerical simulations, we investigate and evaluate the impacts of probe wavelength and power of the involved signals on the extinction ratio (ER) and Q2-factor. Results show that the proposed NOR gate can be achieved both with logical correctness and high quality when the specified conditions are satisfied.

We propose an ultrafast all optical NOR gate for 160 Gb/s return-to-zero Gaussian data signals using a single quantum-dot semiconductor optical amplifier (QD-SOA) and optical filter (OF). By conducting numerical simulations, we investigate and evaluate the impacts of probe wavelength and power of the involved signals on the extinction ratio (ER) and Q(2) -factor. Results show that the proposed NOR gate can be achieved both with logical correctness and high quality when the specified conditions are satisfied.

A signal shaping for the bit-interleaved coded modulation (BICM) is considered. Previous work on signal shaped BICM (SBICM) has two problems: (i) It requires an optimization for the linear factors of superposition modulation. The optimization is valid around target SNR. (ii) The binary labeling of signal constellation is not Gray code. To overcome these problems, non-uniform signal labeling for the input signal with Gaussian distribution is considered. We implement the reduction of peakto-average power ratio (PAPR) by clipping for the largest signal points. From the numerical results, we show that achievable rate of the proposed modulation is larger than that of the previous modulation of SBICM for wide range of SNR, while keeping the expansion of PAPR small. Moreover it is quite close to the coded modulation capacity for wide range of SNR, especially in some SNR region, it is larger than the coded modulation capacity.

We investigate a synchronization method for channels that are impaired by insertion, deletion, and substitution (IDS) errors and include outer low-density parity-check (LDPC) codes with error-correction capabilities and inner marker codes for synchronization. We improve synchronization using fixed-symbols, and evaluate the achievable rate by a specific fixed-symbol assignment. We demonstrate via simulation that the bit error rate of the proposed method over an IDS channel is lower than that of the conventional method.

We study the signal shaping based on the truncated Gaussian distribution. Since the possibility of reducing peak-to average power ratio (PAPR) by the previous study with clipped Gaussian signal constellation is limited, we study PAPR reduction for one-dimensional signal constellation using the truncated Gaussian distribution. From the numerical results, we show that achievable rate of the truncated Gaussian distribution is larger than that of the clipped Gaussian signal constellation for wide range of SNR, while keeping the expansion of PAPR small. Moreover it is quite close to the uniform signal constellation for high SNR.

In this study, we present new methods for piece-wise linear approximation of Min-Sum (MS) decoding algorithm with good trade-off between performance and complexity. By analysis based on density evolution, the performance of the proposed method is identical to that of the Belief-Propagation decoding algorithm. The increment of the complexity of the proposed algorithm is small compared with the MS decoding algorithm. Moreover simulation result also shows effectiveness of the proposed algorithm.