This paper proposes a layer-wise tunable retraining method for edge FeFET Computation-in-Memory (CiM) to compensate the accuracy degradation of neural network (NN) by FeFET device errors. The proposed retraining can tune the number of layers to be retrained to reduce inference accuracy degradation by errors that occur after retraining. Weights of the original NN model, accurately trained in cloud data center, are written into edge FeFET CiM. The written weights are changed by FeFET device errors in the field. By partially retraining the written NN model, the proposed method combines the error-affected layers of NN model with the retrained layers. The inference accuracy is thus recovered. After retraining, the retrained layers are re-written to CiM and affected by device errors again. In the evaluation, at first, the recovery capability of NN model by partial retraining is analyzed. Then the inference accuracy after re-writing is evaluated. Recovery capability is evaluated with non-volatile memory (NVM) typical errors: normal distribution, uniform shift, and bit-inversion. For all types of errors, more than 50% of the degraded percentage of inference accuracy is recovered by retraining only the final fully-connected (FC) layer of Resnet-32. To simulate FeFET Local-Multiply and Global-accumulate (LM-GA) CiM, recovery capability is also evaluated with FeFET errors modeled based on FeFET measurements. Retraining only FC layer achieves recovery rate of up to 53%, 66%, and 72% for FeFET write variation, read-disturb, and data-retention, respectively. In addition, just adding two more retraining layers improves recovery rate by 20-30%. In order to tune the number of retraining layers, inference accuracy after re-writing is evaluated by simulating the errors that occur after retraining. When NVM typical errors are injected, it is optimal to retrain FC layer and 3-6 convolution layers of Resnet-32. The optimal number of layers can be increased or decreased depending on the balance between the size of errors before retraining and errors after retraining.

It is very important to design an embedded real-time system as a fault-tolerant system to ensure dependability. In particular, when a power failure occurs, restart processing after power restoration is required in a real-time system using a conventional processor. Even if power is restored quickly, the restart process takes a long time and causes deadline misses. In order to design a fault-tolerant real-time system, it is necessary to have a processor that can resume operation in a short time immediately after power is restored, even if a power failure occurs at any time. Since current embedded real-time systems are required to execute many tasks, high schedulability for high throughput is also important. This paper proposes a non-stop microprocessor architecture to achieve a fault-tolerant real-time system. The non-stop microprocessor is designed so as to resume normal operation even if a power failure occurs at any time, to achieve little performance degradation for high schedulability even if checkpoint creations and restorations are performed many times, to control flexibly non-volatile devices through software configuration, and to ensure data consistency no matter when a checkpoint restoration is performed. The evaluation shows that the non-stop microprocessor can restore a checkpoint within 5µsec and almost hide the overhead of checkpoint creations. The non-stop microprocessor with such capabilities will be an essential component of a fault-tolerant real-time system with high schedulability.

This paper describes an in-depth analysis of crosstalk in a high-bandwidth 3D-stacked memory using a multi-hop inductive coupling interface and proposes two countermeasures. This work analyzes the crosstalk among seven stacked chips using a 3D electromagnetic (EM) simulator. The detailed analysis reveals two main crosstalk sources: concentric coils and adjacent coils. To suppress these crosstalks, this paper proposes two corresponding countermeasures: shorted coils and 8-shaped coils. The combination of these coils improves area efficiency by a factor of 4 in simulation. The proposed methods enable an area-efficient inductive coupling interface for high-bandwidth stacked memory.

Series-connection of resonant-tunneling diodes (RTDs) has been considered to be efficient in upgrading the output power when it is introduced to oscillator architecture. This work is for clarifying the same architecture also contributes to increasing oscillation frequency because the device parasitic capacitance is reduced M times for M series-connected RTD oscillator. Although this mechanism is expected to be universal, we restrict the discussion to the recently proposed multiphase oscillator utilizing an RTD oscillator lattice loop. After explaining the operation principle, we evaluate how the oscillation frequency depends on the number of series-connected RTDs through full-wave calculations. In addition, the essential dynamics were validated experimentally in breadboarded multiphase oscillators using Esaki diodes in place of RTDs.

This paper presents a hippocampal cognitive prosthesis chip designed for restoring the ability to form new long-term memories due to hippocampal system damage. The system-on-chip (SOC) consists of a 16-channel micro-power low-noise amplifier (LNA), high-pass filters, analog-digital converters (ADCs), a 16-channel spike-sorter, a generalized Laguerre-Volterra model multi-input, multi-output (GLVM-MIMO) hippocampal processor, an 8-channel neural stimulator and peripheral circuits. The proposed LNA achieved a voltage gain of 50dB, input-referred noise of 3.95µVrms, and noise efficiency factor (NEF) of 3.45 with the power consumption of 3.3µW. High-pass filters with a 300-Hz bandwidth are used to filter out the unwanted local field potential (LFP). 4 12-bit successive approximation register (SAR) ADCs with a signal-to-noise-and-distortion ratio (SNDR) of 63.37dB are designed for the digitization of the neural signals. A 16-channel spike-sorter has been integrated in the chip enabling a detection accuracy of 98.3% and a classification accuracy of 93.4% with power consumption of 19µW/ch. The MIMO hippocampal model processor predict output spatio-temporal patterns in CA1 according to the recorded input spatio-temporal patterns in CA3. The neural stimulator performs bipolar, symmetrical charge-balanced stimulation with a maximum current of 310µA, triggered by the processor output. The chip has been fabricated in 40nm standard CMOS technology, occupying a silicon area of 3mm2.

This study proposes a low-complexity permittivity estimation for ground penetrating radar applications based on a contrast source inversion (CSI) approach, assuming multilayered ground media. The homogeneity assumption for each background layer is used to address the ill-posed condition while maintaining accuracy for permittivity reconstruction, significantly reducing the number of unknowns. Using an appropriate initial guess for each layer, the post-CSI approach also provides the dielectric profile of a buried object. The finite difference time domain numerical tests show that the proposed approach significantly enhances reconstruction accuracy for buried objects compared with the traditional CSI approach.

High-performance deep neural network (DNN)-based systems are in high demand in edge environments. Due to its high computational complexity, it is challenging to deploy DNNs on edge devices with strict limitations on computational resources. In this paper, we derive a compact while highly-accurate DNN model, termed dsODENet, by combining recently-proposed parameter reduction techniques: Neural ODE (Ordinary Differential Equation) and DSC (Depthwise Separable Convolution). Neural ODE exploits a similarity between ResNet and ODE, and shares most of weight parameters among multiple layers, which greatly reduces the memory consumption. We apply dsODENet to a domain adaptation as a practical use case with image classification datasets. We also propose a resource-efficient FPGA-based design for dsODENet, where all the parameters and feature maps except for pre- and post-processing layers can be mapped onto on-chip memories. It is implemented on Xilinx ZCU104 board and evaluated in terms of domain adaptation accuracy, inference speed, FPGA resource utilization, and speedup rate compared to a software counterpart. The results demonstrate that dsODENet achieves comparable or slightly better domain adaptation accuracy compared to our baseline Neural ODE implementation, while the total parameter size without pre- and post-processing layers is reduced by 54.2% to 79.8%. Our FPGA implementation accelerates the inference speed by 23.8 times.

In industry, automatic speech recognition has come to be a competitive feature for embedded products with poor hardware resources. In this work, we propose a tiny end-to-end speech recognition model that is lightweight and easily deployable on edge platforms. First, instead of sophisticated network structures, such as recurrent neural networks, transformers, etc., the model we propose mainly uses convolutional neural networks as its backbone. This ensures that our model is supported by most software development kits for embedded devices. Second, we adopt the basic unit of MobileNet-v3, which performs well in computer vision tasks, and integrate the features of the hidden layer at different scales, thus compressing the number of parameters of the model to less than 1 M and achieving an accuracy greater than that of some traditional models. Third, in order to further reduce the CPU computation, we directly extract acoustic representations from 1-dimensional speech waveforms and use a self-supervised learning approach to encourage the convergence of the model. Finally, to solve some problems where hardware resources are relatively weak, we use a prefix beam search decoder to dynamically extend the search path with an optimized pruning strategy and an additional initialism language model to capture the probability of between-words in advance and thus avoid premature pruning of correct words. In our experiments, according to a number of evaluation categories, our end-to-end model outperformed several tiny speech recognition models used for embedded devices in related work.

Since the dark channel prior (DCP)-based dehazing method is ineffective in the sky area and will cause the problem of too dark and color distortion of the image, we propose a novel dehazing method based on sky area segmentation and image fusion. We first segment the image according to the characteristics of the sky area and non-sky area of the image, then estimate the atmospheric light and transmission map according to the DCP and correct them, and then fuse the original image after the contrast adaptive histogram equalization to improve the details information of the image. Experiments illustrate that our method performs well in dehazing and can reduce image distortion.

In the massive machine-type communication scenario, aiming at the problems of active user detection and channel estimation in the grant-free non-orthogonal multiple access (NOMA) system, new sets of non-orthogonal spreading sequences are proposed by using the zero/low correlation zone sequence set with low correlation among multiple sets. The simulation results show that the resulting sequence set has low coherence, which presents reliable performance for channel estimation and active user detection based on compressed sensing. Compared with the traditional Zadoff-Chu (ZC) sequences, the new non-orthogonal spreading sequences have more flexible lengths, and lower peak-to-average power ratio (PAPR) and smaller alphabet size. Consequently, these sequences will effectively solve the problem of high PAPR of time domain signals and are more suitable for low-cost devices in massive machine-type communication.

For dichromats to receive the information represented in color images, it is important to study contrast improvement methods and quantitative evaluation indices of color conversion results. There is an index to evaluate the degree of contrast improvement and in this index, the contrast for dichromacy caused by the lightness component is given importance. In addition, random sampling was introduced in the computation of this index. Although the validity of the index has been shown through comparison with a subjective evaluation, it is considered that the following two points should be examined. First, should contrast for normal trichromacy caused by the lightness component also be attached importance. Second, the influence of random sampling should be examined in detail. In this paper, a new index is proposed and the above-mentioned points are examined. For the first point, the following is revealed through experiment. Consideration of the contrast for normal trichromacy caused by a lightness component that is the same as that for dichromacy may or may not result in a good outcome. The evaluation performance of the proposed index is equivalent to that of the previous index overall. It can be said that the proposed index is superior to the previous one in terms of the unity of evaluating contrast. For the second point, the computation time and the evaluation of significant digits are shown. In this paper, a sampling number such that the number of significant digits can be considered as three is used. In this case, the variation caused by random sampling is negligible compared with the range of the proposed index, whereas the computation time is about one-seventh that when the sampling is not adopted.

Uncovered muck truck detection aims to detect the muck truck and distinguish whether it is covered or not by dust-proof net to trace the source of pollution. Unlike traditional detection problem, recalling all uncovered trucks is more important than accurate locating for pollution traceability. When two objects are very close in an image, the occluded object may not be recalled because the non-maximum suppression (NMS) algorithm can remove the overlapped proposal. To address this issue, we propose a Location First NMS method to match the ground truth boxes and predicted boxes by position rather than class identifier (ID) in the training stage. Firstly, a box matching method is introduced to re-assign the predicted box ID using the closest ground truth one, which can avoid object missing when the IoU of two proposals is greater than the threshold. Secondly, we design a loss function to adapt the proposed algorithm. Thirdly, a uncovered muck truck detection system is designed using the method in a real scene. Experiment results show the effectiveness of the proposed method.

In this paper, we present a novel discriminative dictionary learning (DDL) method for image classification. The local structural relationship between samples is first built by the Laplacian eigenmaps (LE), and then integrated into the basic DDL frame to suppress inter-class ambiguity in the feature space. Moreover, in order to improve the discriminative ability of the dictionary, the category label information of training samples is formulated into the objective function of dictionary learning by considering the discriminative promotion term. Thus, the data points of original samples are transformed into a new feature space, in which the points from different categories are expected to be far apart. The test results based on the real dataset indicate the effectiveness of this method.

Underwater acoustic channels (UWA) are usually sparse, which can be exploited for adaptive equalization to improve the system performance. For the shallow UWA channels, based on the proportional minimum symbol error rate (PMSER) criterion, the adaptive equalization framework requires the sparsity selection. Since the sparsity of the L0 norm is stronger than that of the L1, we choose it to achieve better convergence. However, because the L0 norm leads to NP-hard problems, it is difficult to find an efficient solution. In order to solve this problem, we choose the Gaussian function to approximate the L0 norm. Simulation results show that the proposed scheme obtains better performance than the L1 based counterpart.

In this paper, we propose a heuristic planning method to efficiently accommodate dynamic multilayer path (MLP) demand in multilayer networks consisting of a Time Division Multiplexing (TDM) layer and a Wavelength Division Multiplexing (WDM) layer; the goal is to achieve the flexible accommodation of increasing capacity and diversifying path demands. In addition to the grooming of links at the TDM layer and the route and frequency slots for the elastic optical path to be established, MLP requires the selection of an appropriate operational mode, consisting of a combination of modulation formats and symbol rates supported by digital coherent transceivers. Our proposed MLP planning method defines a planning policy for each of these parameters and embeds the values calculated by combining these policies in an auxiliary graph, which allows the planning parameters to be calculated for MLP demand requirements in a single step. Simulations reveal that the choice of operational mode significantly reduces the blocking probability and demonstrate that the edge weights in the auxiliary graph allow MLP planning with characteristics tailored to MLP demand and network requirements. Furthermore, we quantitatively evaluate the impact of each planning policy on the MLP planning results.

To achieve the extreme high data rate and extreme coverage extension requirements of 6G wireless communication, new spectrum in sub-THz (100-300GHz) and non-terrestrial network (NTN) are two of the macro trends of 6G candidate technologies, respectively. However, non-linearity of power amplifiers (PA) is a critical challenge for both sub-THz and NTN. Therefore, high power efficiency (PE) or low peak to average power ratio (PAPR) waveform design becomes one of the most significant 6G research topics. Meanwhile, high spectral efficiency (SE) and low out-of-band emission (OOBE) are still important key performance indicators (KPIs) for 6G waveform design. Single-carrier waveform discrete Fourier transform spreading orthogonal frequency division multiplexing (DFT-s-OFDM) has achieved many research interests due to its high PE, and it has been supported in 5G New Radio (NR) when uplink coverage is limited. So DFT-s-OFDM can be regarded as a candidate waveform for 6G. Many enhancement schemes based on DFT-s-OFDM have been proposed, including null cyclic prefix (NCP)/unique word (UW), frequency-domain spectral shaping (FDSS), and time-domain compression and expansion (TD-CE), etc. However, there is no unified framework to be compatible with all the enhancement schemes. This paper firstly provides a general description of the 6G candidate waveforms based on DFT-s-OFDM enhancement. Secondly, the more flexible TD-CE supporting methods for unified non-orthogonal waveform (uNOW) are proposed and discussed. Thirdly, a unified waveform framework based on DFT-s-OFDM structure is proposed. By designing the pre-processing and post-processing modules before and after DFT in the unified waveform framework, the three technical methods (NCP/UW, FDSS, and TD-CE) can be integrated to improve three KPIs of DFT-s-OFDM simultaneously with high flexibility. Then the implementation complexity of the 6G candidate waveforms are analyzed and compared. Performance of different DFT-s-OFDM enhancement schemes is investigated by link level simulation, which reveals that uNOW can achieve the best PAPR performance among all the 6G candidate waveforms. When considering PA back-off, uNOW can achieve 124% throughput gain compared to traditional DFT-s-OFDM.

Improvement of the on/off ratio in organic field-effect transistors through the use of pentacene and molybdenum trioxide (MoO3) layers was attempted via the preparation of a discontinuous MoO3 layer using a mesh mask. We prepared three types of devices. Device A had a conventional top-contact structure with an n-type Si wafer and a 200-nm-thick SiO2 film onto which we deposited a 70-nm-thick pentacene film and a 30-nm-thick layer of Au top electrodes. Devices B and C had a similar structure to device A but received a continuous and a discontinuous MoO3 layer, respectively. The off current in Device B was remarkably high; in contrast, the off current in Device C was reduced and dependent on the separation of the MoO3 layer. It was deduced that the high resistance of the area without MoO3 contributed to the reduced off current.

We investigated the photochemical stability of an electro-optic (EO) polymer under laser irradiation at 1310nm to reveal photodegradation mechanisms. It was found that one-photon absorption excitation assisted with the thermal energy at the temperature is involved in the photodegradation process, in contrast to our previous studies at a wavelength of 1550nm where two-photon absorption excitation is involved in the photodegradation process. Thus, both the excitation wavelength and the thermal energy strongly affect to the degradation mechanism. In any cases, the photodegradation of EO polymers is mainly related to the generation of exited singlet oxygen.

We have developed multilayered polymer-based inverted organic light emitting diodes (iOLED) using transfer-printing and push-coating techniques. We obtained the higher efficiency and lower operation voltage with push-coated blue light emitting polymer and hole transporting polymer than the devices with spin-coated film. The β-phase obtained for blue emitting layer is attributable to the improved performance of relatively efficient bule and white iOLEDs with an external quantum efficiency (EQE) of above 2%.

Lead bromide-based perovskite organic-inorganic quantum-well films incorporated polycyclic aromatic chromophores into the organic layer (in other words, hybrid quantum-wells combined lead bromide semiconductor and organic semiconductors) were prepared by use of the spin-coating technique from the DMF solution in which PbBr2 and alkyl ammonium bromides which were linked polycyclic aromatics, pyrene, phenanthrene, and anthracene. When the pyrene-linked methyl ammonium bromide, which has a relatively small molecular cross-section with regard to the inorganic semiconductor plane, was employed, a lead bromide-based perovskite structure was successfully formed in the spin-coated films. When the phenanthrene-linked and anthracene-linked ammonium bromides, whose chromophore have large molecular cross-sections, were employed, lead bromide-based perovskite structures were not formed. However, the introduction of longer alkyl chains into the aromatics-linked ammonium bromides made it possible to form the perovskite structure.