Orthogonal frequency division multiplexing (OFDM) provides a promising modulation technique for underwater acoustic (UWA) communication systems. It is indispensable to obtain channel state information for channel estimation to handle the various channel distortions and interferences. However, the conventional channel estimation methods such as least square (LS), minimum mean square error (MMSE) and back propagation neural network (BPNN) cannot be directly applied to UWA-OFDM systems, since complicated multipath channels may cause a serious decline in performance estimation. To address the issue, two types of channel estimators based on deep neural networks (DNNs) are proposed with a novel training strategy in this paper. The proposed DNN models are trained with the received pilot symbols and the correct channel impulse responses in the training process, and then the estimated channel impulse responses are offered by the proposed DNN models in the working process. The experimental results demonstrate that the proposed methods outperform LS, BPNN algorithms and are comparable to the MMSE algorithm in respect to bit error rate and normalized mean square error. Meanwhile, there is no requirement of prior statistics information about channel autocorrelation matrix and noise variance for our proposals to estimate channels in UWA-OFDM systems, which is superior to the MMSE algorithm. Our proposed DNN models achieve better performance using 16QAM than 32QAM, 64QAM, furthermore, the specified DNN architectures help improve real-time performance by saving runtime and storage resources for online UWA communications. INDEX TERMS Deep neural networks, OFDM systems, channel estimation, underwater acoustic communication.
Non-volatile computing-in-memory macros that are based on two-dimensional arrays of memristors are of use in the development of artificial intelligence edge devices. Scaling such systems to three-dimensional arrays could provide higher parallelism, capacity and density for the necessary vector–matrix multiplication operations. However, scaling to three dimensions is challenging due to manufacturing and device variability issues. Here we report a two-kilobit non-volatile computing-in-memory macro that is based on a three-dimensional vertical resistive random-access memory fabricated using a 55 nm complementary metal–oxide–semiconductor process. Our macro can perform 3D vector–matrix multiplication operations with an energy efficiency of 8.32 tera-operations per second per watt when the input, weight and output data are 8, 9 and 22 bits, respectively, and the bit density is 58.2 bit µm–2. We show that the macro offers more accurate brain MRI edge detection and improved inference accuracy on the CIFAR-10 dataset than conventional methods.
In underwater acoustic-orthogonal frequency division multiplexing (UWA-OFDM) systems, the performance of channel estimation is significantly affected by pilot allocation in the framework of compressed sensing (CS). However, for optimizing the pilot allocation, an exhaustive search method over all possible allocations is computationally prohibitive and random search method may not ensure convergence accuracy. In this paper, the meta-heuristic algorithm of the whale optimization algorithm (WOA) is employed to address this issue. For reinforcing the capability of balancing exploration and exploitation, an enhanced variant of WOA termed EWOA is presented with four optimization strategies. After that, a joint algorithm combining CS with EWOA (CS-EWOA) is proposed for pilot allocation in UWA-OFDM systems. Through extensive simulations, the improvement of EWOA is demonstrated on the majority of benchmark functions over other well-known meta-heuristic algorithms. With regard to bit error rate (BER) and mean square error (MSE) for channel estimation, the proposed CS-EWOA algorithm outperforms the equispaced, random, genetic algorithm (GA), particle swarm optimization (PSO), and WOA-based pilot allocation methods. Moreover, it is robust with varying system subcarriers and channel models. Furthermore, the CS-EWOA exhibits superior convergence performance without increasing the computational complexity compared with the GA-, PSO-, and WOA-based methods in the iteration process of pilot allocation optimization. It can be concluded from the simulation results that the proposed CS-EWOA algorithm is competitive to optimize pilot allocation for channel estimation in UWA-OFDM systems.
Electrocardiogram (ECG) heartbeat classification plays a vital role in early diagnosis and effective treatment, which provide opportunities for earlier prevention and intervention. In an effort to continuously monitor and detect abnormalities in patients’ ECG signals on portable devices, this paper present a lightweight ECG heartbeat classification method based on a spiking neural network (SNN), a relatively shallow SNN model integrated with a channel-wise attentional module. We further explore the best-optimized architecture, which benefits from leveraging the full advantages of the SNN potential with the attention mechanism to process the classification task at low power and capture prominent features concerning the time, morphology, and multi-channel representations of the ECG signal. Results show that our model achieves overall classification accuracy of 98.26%, sensitivity of 94.75%, and F1 score of 89.09% on the MIT-BIH database, with energy consumption of 346.33 μJ per beat and runtime of 1.37 ms. Moreover, we have conducted multiple experiments to compare against current state-of-the-art methods using their assessment strategies to evaluate our model implementation on FPGA. So far, our work achieves comparable overall performance with all the literature in terms of classification accuracy, energy consumption, and real-time capability.
BackgroundNeural stimulation is an important method used to activate or inhibit action potentials of the neuronal anatomical targets found in the brain, central nerve and peripheral nerve. The neural stimulator system produces biphasic pulses that deliver balanced charge into tissue from single or multichannel electrodes. The timing and amplitude of these biphasic pulses are precisely controlled by the neural stimulator software or imbedded algorithms. Amplitude mismatch between the anodic current and cathodic current of the biphasic pulse will cause permanently damage for the neural tissues. The main goal of our circuit and layout design is to implement a 16-channel biphasic current mode programmable neural stimulator with calibration to minimize the current mismatch caused by inherent complementary metal oxide semiconductor (CMOS) manufacturing processes.MethodsThis paper presents a 16-channel constant current mode neural stimulator chip. Each channel consists of a 7-bit controllable current DAC used as sink and source current driver. To reduce the LSB quantization error and the current mismatch, an automatic calibration circuit and flow diagram is presented in this paper. There are two modes of operation of the stimulator chip—namely, stimulation mode and calibration mode. The chip also includes a digital interface used to control the stimulator parameters and calibration levels specific for each individual channel.ResultsThis stimulator Application Specific Integrated Circuit (ASIC) is designed and fabricated in a 0.18 μm High-Voltage CMOS technology that allows for ±20 V power supply. The full-scale stimulation current was designed to be at 1 mA per channel. The output current was shown to be constant throughout the timing cycles over a wide range of electrode load impedances. The calibration circuit was also designed to reduce the effect of CMOS process variation of the P-channel metal oxide semiconductor (PMOS) and N-channel metal oxide semiconductor (NMOS) devices that will result in charge delivery to have less than 0.13% error.ConclusionsA 16-channel integrated biphasic neural stimulator chip with calibration is presented in this paper. The stimulator circuit design was simulated and the chip layout was completed. The chip layout was verified using design rules check (DRC) and layout versus schematic (LVS) design check using computer aided design (CAD) software. The test results we presented show constant current stimulation with charge balance error within 0.13% least-significant-bit (LSB). This LSB error was consistent throughout a variety stimulation patterns and electrode load impedances.
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.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.