Decoder Output Research Articles

A Coding Approach With Key-Channel Randomization for Physical-Layer Authentication

We propose a physical-layer challenge-response authentication approach in this paper based on combined shared secret key and channel state information between two legitimate nodes in an orthogonal frequency division multiplexing system. The proposed approach can be used even if the correlation of channel coefficients exists, which can be exploited to extract the shared secret key in conventional approaches. Furthermore, channel coding is employed to mitigate the difference between the two estimated channels as well as channel fading and background noise. As a result, in the proposed physical-layer authentication approach, the decoder's output can be used for authentication and provide a reliable decision under active attack.

Stack-Captioning: Coarse-to-Fine Learning for Image Captioning

The existing image captioning approaches typically train a one-stage sentence decoder, which is difficult to generate rich fine-grained descriptions. On the other hand, multi-stage image caption model is hard to train due to the vanishing gradient problem. In this paper, we propose a coarse-to-fine multi-stage prediction framework for image captioning, composed of multiple decoders each of which operates on the output of the previous stage, producing increasingly refined image descriptions. Our proposed learning approach addresses the difficulty of vanishing gradients during training by providing a learning objective function that enforces intermediate supervisions. Particularly, we optimize our model with a reinforcement learning approach which utilizes the output of each intermediate decoder's test-time inference algorithm as well as the output of its preceding decoder to normalize the rewards, which simultaneously solves the well-known exposure bias problem and the loss-evaluation mismatch problem. We extensively evaluate the proposed approach on MSCOCO and show that our approach can achieve the state-of-the-art performance.

Design and analysis of LDPC codes for joint source‐channel decoding of two correlated sensors

This study is concerned with the design of ensembles of systematic low-density parity-check (LDPC) codes to increase the lifetime of wireless sensors by taking advantage of the inherent correlation between the transmitted data of the sensors. The authors consider two correlated sensors where the data is encoded independently at each sensor through a systematic LDPC encoder and sent over two independent channels. At the receiver, a joint source-channel decoder consisting of two component LDPC decoders is considered where the encoded bits at the output of each component decoder are used as the a priori information at the other decoder. The authors first perform asymptotic performance analysis using the concept of extrinsic information transfer (EXIT) charts. Then, the developed modified EXIT charts are used to design ensembles for different values of correlation. Our results show that as the average check node degree of the designed ensembles grow, the decoding thresholds corresponding to the designed ensembles approach the theoretical limit. Finite block-length performance evaluation indicates that for larger values of correlation, deploying the designed ensembles through the joint decoder can almost double the sensor's lifetime without increasing the complexity of the encoder.

Design of a high bitrate optical decoder based on photonic crystals

This paper presents a triangular lattice of two-dimensional photonic crystal in the design of a novel all-optical $$1\times 2$$ decoder. To control and conduct light, linear and point defects were developed in curved directions. In addition to the main entrance, a controlling entrance was assumed to control light using light interference theory when the input source is active and to activate the intended output in its absence. Logical values of “0” and “1” are introduced based on the ratio of output-to-input power. Plane wave expansion is used to calculate band structure of the lattice and the finite difference time domain method is used to calculate the optical power distribution in waveguide paths. The key feature of the proposed structure are low size and contrast enhancement, which will reduce the error in the decoder output. The results show that the structure can be used as a decoder with a bit rate of about 2 Tbit/s and time delay of 0.1 ps.

Modal beamformer analysis of ambisonics decoders

Ambisonic recording and playback is experiencing a revival with the introduction of higher-order Ambisonics (HOA) and is included in the recent MPEG-H and ATSC 3.0 standards. Higher-Order Ambisonics processing results in an improved spatial resolution compared to the original first-order Ambisonics proposed in 1970s. From a signal processing point of view, Ambisonics is a two-step process. In the first step spherical harmonic signals are generated by an Ambisonic encoder. Based on these spherical harmonic signals an Ambisonic decoder generates the output signals in the second step. Many different decoder flavors have emerged over time. Various decoder designs are motivated by different optimization criterion for the resulting soundfield. In this presentation we will analyze the Ambisonic decoders based on the fact that the two-step Ambisonics processing is closely related to spherical harmonic modal beamforming. It is shown that the decoder output signals can be interpreted in terms of beampatterns. The resulting beampatterns for the different decoders are computed and some performance criteria are presented.

An Improved $A^{\ast}$ Decoding Algorithm With List Decoding

Comparing with hard decision decoding algorithms, soft decoding has a lower probability of bit error but a higher computational complexity. As a maximum-likelihood soft decoding method, the $A^{\ast }$ algorithm is the most basic and widely used to minimize bit error probability. However, its average computational complexity strongly depends on a seed codeword and a heuristic function utilized during the decoding process. To efficiently reduce the computational complexity while maintaining the decoding accuracy theoretically and practically, this paper proposes an improved $A^{\ast }$ decoding algorithm consisting of two phases. The first phase applies the greedy list decoding to the linear block code to obtain a seed codeword. According to the seed, the second phase applies the improved $A^{\ast }$ algorithm to obtain the final decoding output. The heuristic function used in the $A^{\ast }$ algorithm is modified in two aspects: 1) use more information of partial decoded symbols to improve the accuracy of the function and 2) take advantage of Hamming distance to reduce the search space. Simulations on the $RM(5,2)$ Reed–Muller codes and [128, 64] binary extended BCH code show that this improved $A^{\ast }$ algorithm is more efficient in average decoding complexity than many other algorithms while maintaining the decoding accuracy.

A low complexity soft decision decoder for linear block codes

The Chase-2 decoding algorithm is an efficient soft input hard output decoder that uses a list of most likely error patterns. The main idea behind this decoder is to use a hard decision decoder HD in 2t times, where t is the error correcting capability of the used code. The complexity of Chase-2 decoder is then of order 2t.O(HD). The choice of the HD decoder impacts considerably the complexity. In this paper, we present the Chase-HSDec decoder where we have integrated the low complexity hard decision decoder HSDec that we have recently developed as component decoder. Given its low temporal complexity comparing to their competitors, the use of HSDec in the Chasing technique, decrease the temporal complexity. Chase-HSDec is applied to decode some BCH and QR codes; the simulation results show that the proposed solution yield to good error correcting performances. The comparisons between the proposed decoder and some competitors have shown that it exceeds them. The main advantage of this decoder is that it guarantees the same performances of Chase-Berlekamp Massey decoder with reduced temporal complexity. Contrary to the Chase-Berlekamp Massey decoder which is generally used for BCH codes; it is possible to apply Chase-HSDec on other linear codes defined with their generator matrix.

A cryptography-based approach for movement decoding.

Brain decoders use neural recordings to infer the activity or intent of a user. To train a decoder, one generally needs to infer the measured variables of interest (covariates) from simultaneously measured neural activity. However, there are cases for which obtaining supervised data is difficult or impossible. Here, we describe an approach for movement decoding that doesn’t require access to simultaneously measured neural activity and motor outputs. We use the statistics of movement—much like cryptographers use the statistics of language—to find a mapping between neural activity and motor variables, and then align the distribution of decoder outputs with the typical distribution of motor outputs by minimizing their Kullback-Leibler divergence. By using datasets collected from the motor cortex of three non-human primates performing either a reaching task or an isometric force-production task, we show that the performance of such a distribution-alignment decoding algorithm is comparable with the performance of supervised approaches. Distribution-alignment decoding promises to broaden the set of potential applications of brain decoding.

Turbo equalization based on joint Gaussian, SIC-MMSE and LMMSE for nonlinear satellite channels

The nonlinear distortion of wideband signal due to the filtering and efficiently operated high power amplifiers limits the performance of satellite communications. Volterra series can be used to describe the nonlinear satellite channels effectively. Most existing equalizers simply ignore the nonlinear terms or treat all the nonlinear combinations of symbols as interference. In this study, by properly exploiting information from nonlinear terms, we propose three turbo equalizers for nonlinear satellite channels, namely, joint Gaussian (JG), soft interference cancellation-minimum mean square error (SIC-MMSE) and linear minimum mean square error (LMMSE) equalizers. In JG and SIC-MMSE-based equalizers, both the linear and nonlinear terms that contain the symbol of interest are considered as desired signals. Accordingly, the required statistics are calculated based on the a priori probabilities of coded bits from output of channel decoder. For LMMSE-based equalizer, we propose to calculate the extrinsic information from output of equalizer by excluding the prior information in both the linear and nonlinear terms. Simulation results demonstrate that the proposed equalizers significantly outperform the method which ignores the presence of nonlinear interferences. Moreover, the nonlinear terms that contain the symbol of interest can be exploited to further improve the performance of turbo equalization.

A Study on Relationship Between Recording Pattern and Decoding Reliability in SMR

The shingled magnetic recording (SMR) system is influenced by not only jitter-like medium noise from adjacent bits in the down-track direction but also intertrack interference (ITI) from the adjacent tracks. In this paper, we investigate the relationship between the recording pattern and the decoding reliability by classifying the log-likelihood ratio at the a posteriori probability decoder output according to the recording patterns under an SMR R–W channel specification of 4 Tb/in2 by computer simulation. The result shows that the decoding reliability for the recording pattern “010” heavily affected by transition jitter is the lowest and little improved even if the influence of ITI is mitigated by a 2-D finite-impulse response filter.

An iterative detection/decoding algorithm of correlated sources for the LDPC-based relay systems

An iterative detection/decoding algorithm of correlated sources for the LDPC-based relay systems is presented. The signal from the source-destination (S-D) link is formulated as a highly correlated counterpart from the relay-destination (R-D) link. A special XOR vector is defined using the correlated hard decision information blocks from two decoders and the extrinsic information exchanged between the two decoders is derived by the log-likelihood ratio (LLR) associated with the XOR vector. Such the decoding scheme is different from the traditional turbo-like detection/decoding algorithm, where the extrinsic information is computed by the side information and the soft decoder outputs. Simulations show that the presented algorithm has a slightly better performance than the traditional turbo-like algorithm (Taking the (255,175) EG-LDPC code as an example, it achieves about 0.1dB performance gains around BLER=10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-4</sup> ). Furthermore, the presented algorithm requires fewer computing operations per iteration and has faster convergence rate. For example, the average iteration of the presented algorithm is 33 at SNR=1.8dB, which is about twice faster than that of the turbo-like algorithm, when decoding the (961,721) QC-LDPC code. Therefore, the presented decoding algorithm of correlated sources provides an alternative decoding solution for the LDPC-based relay systems.

DVB-S3기반 (1+7)PSK 변조방식에서 FTN 신호의 효율적인 복호 모델

위성방송 시스템의 표준안인 DVB-S3 에서는 LDPC 부호화된 비트를 (1+7) PSK 변조방식과 전송률을 증가 시키기 위해 FTN 기법의 적용 방안을 고려하고 있다. 기존의 복호 방식은 FTN 간섭으로 인한 성능 감소를 위해 BICM-ID 과 BCJR 복호 방식의 반복 복호 기법을 적용하고 있으나, BICM-ID방식은 심볼 LLR 적용후 LDPC 복호기의 출력값을 이용하여 나머지 비트에 대해 복호를 하는 비트 LLR기반인데 이는 많은 계산량을 요구하며, BCJR 복호 방식 또한 순방형 역방향 메트릭을 구하는 과정에서 마놓은 계산량을 요구한다. 따라서 본 논문에서는 FTN 기법의 적용으로 인해 열화된 성능을 반복 복호 기법을 통하여 성능 및 계산량 관점에서 개선하는 효율적인 복호방법을 제안한다. 계산량을 감소시키기 위해 동일 성능향상에서 복호기의 출력 값을 다시 FTN 매핑하여 FTN으로 인한 간섭량을 제거 후 심볼 LLR 계산만 하여 LDPC 복호기로 입력한다. 본 논문에서 제안한 (1+7) PSK 변조방식이 적용된 DVB-S3 시스템에 가우시안 채널에서 성능 향상 및 계산량이 감소됨을 확인 하였다. In DVB-S3 standard of satellite broadcasting systems, FTN technique is applied to LDPC codes with (1+7) PSK modulation. In standard, BICM-ID and BCJR decoding method are considered to alleviate performance degradation due to FTN processing. BICM-ID method improves performance by calculating a new LLR from hard-decision value of decoder output. BCJR also improves performance by calculating forward and backward matrix each other. However these two methods require high computational complexity. Therefore this paper proposed modified decoding method in order to reduce computational complexity without performance degradation.

Code-Aided DOA Estimation From Turbo-Coded QAM Transmissions: Analytical CRLBs and Maximum Likelihood Estimator

In this paper, we address the problem of direction of arrival (DOA) estimation from turbo-coded square-QAM-modulated transmissions. We devise a new code-aware direction finding concept, derived from maximum likelihood (ML) theory, wherein the soft information provided by the soft-input soft-output decoder, in the form of log-likelihood ratios, is exploited to assist the estimation process. At each turbo iteration, the decoder output is used to refine the ML DOA estimate. The latter is in turn used to perform a more focused receiving beamforming thereby providing more reliable information-bearing sequences for the next turbo iteration. In order to benchmark the new estimator, we also derive the analytical expressions for the exact Cramer-Rao lower bounds (CRLBs) of code-aided (CA) DOA estimates. Simulation results will show that the new CA direction finding scheme lies between the two traditional schemes of completely non-data-aided and data-aided (DA) estimations. Huge performance improvements are achieved by embedding the direction finding and receive beamforming tasks into the turbo iteration loop. Moreover, the new CA DOA estimator reaches the new CA CRLBs over a wide range of practical SNRs thereby confirming its statistical efficiency. As expected intuitively, its performance further improves at higher coding rates and/or lower modulation orders.

Pareto-optimal multi-objective dimensionality reduction deep auto-encoder for mammography classification

Background and objectiveFeature reduction is an essential stage in computer aided breast cancer diagnosis systems. Multilayer neural networks can be trained to extract relevant features by encoding high-dimensional data into low-dimensional codes. Optimizing traditional auto-encoders works well only if the initial weights are close to a proper solution. They are also trained to only reduce the mean squared reconstruction error (MRE) between the encoder inputs and the decoder outputs, but do not address the classification error. The goal of the current work is to test the hypothesis that extending traditional auto-encoders (which only minimize reconstruction error) to multi-objective optimization for finding Pareto-optimal solutions provides more discriminative features that will improve classification performance when compared to single-objective and other multi-objective approaches (i.e. scalarized and sequential). MethodsIn this paper, we introduce a novel multi-objective optimization of deep auto-encoder networks, in which the auto-encoder optimizes two objectives: MRE and mean classification error (MCE) for Pareto-optimal solutions, rather than just MRE. These two objectives are optimized simultaneously by a non-dominated sorting genetic algorithm. ResultsWe tested our method on 949 X-ray mammograms categorized into 12 classes. The results show that the features identified by the proposed algorithm allow a classification accuracy of up to 98.45%, demonstrating favourable accuracy over the results of state-of-the-art methods reported in the literature. ConclusionsWe conclude that adding the classification objective to the traditional auto-encoder objective and optimizing for finding Pareto-optimal solutions, using evolutionary multi-objective optimization, results in producing more discriminative features.

Coded Eight-Dimensional QAM Technique Using Iterative Soft-Output Decoding and Its Demonstration in High Baud-Rate Transmission

We proposed coded eight-dimensional (8-D) quadrature amplitude modulation (QAM) and experimentally demonstrated in high baud-rate transmission. Both high baud rate and multilevel modulation techniques are necessary for large capacity long haul transmission without increasing the number of transponders. However, optical signal-to-noise ratio (OSNR) tolerance of higher order modulation is low due to the short minimum Euclidean distance. Therefore, to relax OSNR tolerance by enlarging minimum Euclidean distance, we proposed 8-D modulation formats. The spectral efficiency of the proposed coded 8-D-16QAM, 8-D-32QAM, and 8-D-64QAM is equivalent to polarization division multiplexed (PDM)-8QAM, PDM-16QAM, and PDM-32QAM, respectively. To confirm the applicability of the proposed 8-D modulation to high baud rate transmission, we experimentally demonstrated 96-GBaud 8-D-16QAM wavelength division multiplexed transmission over 5,-252 km using iterative soft output decoding. The line data rate is 576-Gbps and a net rate 478-Gbps. At the same spectral efficiency, 8-D-16QAM using iterative soft output decoding can expand the transmission reach by 1212 km compared to conventional PDM-8QAM.

Signal-independent noise in intracortical brain–computer interfaces causes movement time properties inconsistent with Fitts’ law

Objective. Do movements made with an intracortical BCI (iBCI) have the same movement time properties as able-bodied movements? Able-bodied movement times typically obey Fitts’ law: (where MT is movement time, D is target distance, R is target radius, and are parameters). Fitts’ law expresses two properties of natural movement that would be ideal for iBCIs to restore: (1) that movement times are insensitive to the absolute scale of the task (since movement time depends only on the ratio ) and (2) that movements have a large dynamic range of accuracy (since movement time is logarithmically proportional to ). Approach. Two participants in the BrainGate2 pilot clinical trial made cortically controlled cursor movements with a linear velocity decoder and acquired targets by dwelling on them. We investigated whether the movement times were well described by Fitts’ law. Main results. We found that movement times were better described by the equation , which captures how movement time increases sharply as the target radius becomes smaller, independently of distance. In contrast to able-bodied movements, the iBCI movements we studied had a low dynamic range of accuracy (absence of logarithmic proportionality) and were sensitive to the absolute scale of the task (small targets had long movement times regardless of the ratio). We argue that this relationship emerges due to noise in the decoder output whose magnitude is largely independent of the user’s motor command (signal-independent noise). Signal-independent noise creates a baseline level of variability that cannot be decreased by trying to move slowly or hold still, making targets below a certain size very hard to acquire with a standard decoder. Significance. The results give new insight into how iBCI movements currently differ from able-bodied movements and suggest that restoring a Fitts’ law-like relationship to iBCI movements may require non-linear decoding strategies.

Chaos based blood glucose noninvasive measurement: new concept and custom study

Background. Non invasive monitoring of Blood Glucose (BG) has been a challenge calling for new accurate and fast measurement methods. Objective. To propose new concept of chaos based BG non invasive test aiming at personal customization requirements. Methods. First to build the compact RC model of tissue BG through impedance precision measuring Kit, then to simulate and soft-test BG by Boolean chaotic Codec circuits in soft tool Multisim 13.0, The third to capture the chaotic decoding outputs with the Kit plus PC in calculated signatures of resistor and phase of the tested impedance at the subjects’ left wrist in synchronous test by Bayer BG meter. Results. All in controlled trials of Bayer BG meter, the chaotic BG modelling had gained three new compared formulae in merits of errors less than 1mmol/L and latency less than 1minute. Conclusion. During further verification of this chaotic test paradigm, the opened logic route of above methods will boost measurement experts’ confidence in overcoming future problems of blood glucose monitoring in vivo.

Neural mechanisms underlying catastrophic failure in human–machine interaction during aerial navigation

Objective. We investigated the neural correlates of workload buildup in a fine visuomotor task called the boundary avoidance task (BAT). The BAT has been known to induce naturally occurring failures of human–machine coupling in high performance aircraft that can potentially lead to a crash—these failures are termed pilot induced oscillations (PIOs). Approach. We recorded EEG and pupillometry data from human subjects engaged in a flight BAT simulated within a virtual 3D environment. Main results. We find that workload buildup in a BAT can be successfully decoded from oscillatory features in the electroencephalogram (EEG). Information in delta, theta, alpha, beta, and gamma spectral bands of the EEG all contribute to successful decoding, however gamma band activity with a lateralized somatosensory topography has the highest contribution, while theta band activity with a fronto-central topography has the most robust contribution in terms of real-world usability. We show that the output of the spectral decoder can be used to predict PIO susceptibility. We also find that workload buildup in the task induces pupil dilation, the magnitude of which is significantly correlated with the magnitude of the decoded EEG signals. These results suggest that PIOs may result from the dysregulation of cortical networks such as the locus coeruleus (LC)—anterior cingulate cortex (ACC) circuit. Significance. Our findings may generalize to similar control failures in other cases of tight man-machine coupling where gains and latencies in the control system must be inferred and compensated for by the human operators. A closed-loop intervention using neurophysiological decoding of workload buildup that targets the LC-ACC circuit may positively impact operator performance in such situations.

An interleave division multiple access receiver for multi-user underwater acoustic communication

A downlink interleave division multiple access (IDMA) receiver, employing sparse channel estimation and interference canceller (IC), is proposed for multi-user underwater communications. To eliminate inter-symbol interference (ISI) and multiple access interference (MAI), an equalizer, a decoder, and an interference canceller are implemented in the receiver. The equalizer is used to eliminate ISI. The channel parameters used in the equalizer are estimated by a modified OMP sparse channel estimation algorithm. The performance of channel estimation is further improve ed using the decoder soft output information during the receiver iteration. MAI, reconstructed by the decoder output and channel parameters, can be eliminated in IC section of other users. The receiver’s performance is verified in shallow water acoustic channels during a sea-trail in the South China Sea. Three different ranges between the transmitter and the receiver, i.e., 4.5 km,6 km, and 7 km, were investigated with a maximum of four users at the data rate of 500 bps per user. The results demonstrated that the proposed IDMA receiver provides good performance with low bit error rate (BER). Furthermore, the receiver has been applied in a real-time underwater communication network.

Enhancing the Performance of the Quasi-ML Receiver (Detector Plus Decoder) for Coded MIMO Systems Via Statistical Information

We propose a statistical information-based approach to enhance the performance of the noniterative quasi-maximum-likelihood (QML) receiver (detector plus decoder) recorded at the output of the channel decoder for coded multiple-input–multiple-output (MIMO) systems. As an illustrative example, the QR decomposition with an $M$ -algorithm (QRD-M)-based noniterative QML receiver is employed in this paper. We first illustrate that there is a large discrepancy between the amplitude of the raw log-likelihood ratio (LLR) generated by the complexity-constrained QRD-M detector and its reliability, particularly for LLRs with high amplitudes. The discrepancy is shown to be the main underlying cause of degraded performance in coded MIMO systems. Then, we propose a reliability-aware transformation technique, which exploits the statistical relationship between the raw LLR's amplitude and its reliability to adjust the raw LLR's amplitude to a level commensurate with its reliability. By using the proposed statistical knowledge-based technique, the performance of the receiver (detector plus decoder) recorded at the output of the channel decoder is significantly enhanced for high-order modulations. The extension of the proposed method to other QML receivers such as the list sequential (LISS) detector and list sphere detector (LSD)-based receivers is straightforward.