Programmable Parameters Research Articles

Deep learning with photonic neural cellular automata.

Rapid advancements in deep learning over the past decade have fueled an insatiable demand for efficient and scalable hardware. Photonics offers a promising solution by leveraging the unique properties of light. However, conventional neural network architectures, which typically require dense programmable connections, pose several practical challenges for photonic realizations. To overcome these limitations, we propose and experimentally demonstrate Photonic Neural Cellular Automata (PNCA) for photonic deep learning with sparse connectivity. PNCA harnesses the speed and interconnectivity of photonics, as well as the self-organizing nature of cellular automata through local interactions to achieve robust, reliable, and efficient processing. We utilize linear light interference and parametric nonlinear optics for all-optical computations in a time-multiplexed photonic network to experimentally perform self-organized image classification. We demonstrate binary (two-class) classification of images using as few as 3 programmable photonic parameters, achieving high experimental accuracy with the ability to also recognize out-of-distribution data. The proposed PNCA approach can be adapted to a wide range of existing photonic hardware and provides a compelling alternative to conventional photonic neural networks by maximizing the advantages of light-based computing whilst mitigating their practical challenges. Our results showcase the potential of PNCA in advancing photonic deep learning and highlights a path for next-generation photonic computers.

Test result of the new ASD2 chips for Phase-II upgrade of the ATLAS MDT chambers at HL-LHC

For the Phase-II upgrade of the ATLAS Muon Spectrometer to High-Luminosity LHC, new front-end readout electronics for the Monitored Drift Tube chambers is required, as the old one no longer meets the demands. The first stage in the Monitored Drift Tubes readout chain is the Amplifier-Shaper-Discriminator chip. For the upgrade, the new ASD2 ASIC chip in IBM 130 nm CMOS technology has been developed. For the ATLAS experiment, 80000 ASD2 chips are produced, which have to be tested before integration in order to obtain the required 50000 well-performing chips in the end. Using a prototype tester board and pre-production ASD2 chips, the overall performance and the influence on programmable parameters is investigated. Based on these results, 1775 production chips are tested to define final optimized cut values for the automated production testing of all chips in the company.

Charge-Mode Neural Stimulator With a Capacitor-Reuse Residual Charge Detector and Active Charge Balancing for Epileptic Seizure Suppression.

This study proposes a charge-mode neural stimulator for electrical stimulation systems that utilizes a capacitor-reuse technique with a residual charge detector and achieves active charge balancing simultaneously. The design is mainly used for epilepsy suppression systems to achieve real-time symptom relief during seizures. A charge-mode stimulator is adopted in consideration of the complexity of circuit design, the high voltage tolerance of transistors, and system integration requirements in the future. The residual charge detector allows users to understand the current stimulus situation, enabling them to make optimal adjustments to the stimulation parameters. On the basis of the information on actual stimulation charge, active charge balancing can effectively prevent the accumulation of mismatched charges on electrode impedance. The capacitor- and phase-reuse techniques help realize high integration of the overall stimulator circuit in consideration of the commonality of the use of a capacitor and charging/discharging phase in the stimulation circuit and charge detector. The proposed charge-mode neural stimulator is implemented in a TSMC 0.18 µm 1P6M CMOS process with a core area of 0.2127 mm2. Measurement results demonstrate the accuracy of the stimulation's functionality and the programmable stimulus parameters. The effectiveness of the proposed charge-mode neural stimulator for epileptic seizure suppression is verified through animal experiments.

Research on Tremor Suppression Strategies Under a Constant Current Peripheral Electrical Stimulation Device for Parkinson's Disease.

Tremor, a prevalent symptom in Parkinson's patients, is conventionally treated with medications and craniotomy. However, the associated side effects and high surgical costs pose challenges for some individuals. In this study, a lightweight constant current electrical stimulator was developed, which is driven by the FPGA to control the underlying logic and has multiple programmable stimulation parameters. Clinical experiments involving patients with Parkinson's-related resting tremor symptoms were conducted to assess the efficacy of peripheral electrical stimulation. Two Co-contraction Avoidance Stimulation (CAS) strategies targeting nerves and muscles were proposed to reduce tremors. Four Parkinson's disease (PD) patients were recruited to verify the effectiveness of these strategies. Kinematic data recorded by inertial sensors showed that the radial nerve and muscle intervention strategies reduced the average angular velocity amplitude of finger joints during resting tremor by 75.92% and 82.41%, respectively. Notably, under low-frequency pulse stimulation (100 Hz) focused on muscle interference, a low-intensity current of no more than 8 mA maintained a tremor suppression rate of 59.91% even 5 minutes post-stimulation. Based on the experimental results, it is concluded that the constant current electrical stimulator developed in this study can effectively suppress tremor under specific stimulation strategies. These findings have significant implications for the development of lightweight, wearable tremor suppression devices. The stimulator's adaptability, coupled with its precise control parameters, demonstrates promise for advancing non-invasive and cost-effective tremor management in Parkinson's patients.

Forward-forward training of an optical neural network.

Neural networks (NNs) have demonstrated remarkable capabilities in various tasks, but their computation-intensive nature demands faster and more energy-efficient hardware implementations. Optics-based platforms, using technologies such as silicon photonics and spatial light modulators, offer promising avenues for achieving this goal. However, training multiple programmable layers together with these physical systems poses challenges, as they are difficult to fully characterize and describe with differentiable functions, hindering the use of error backpropagation algorithm. The recently introduced forward-forward algorithm (FFA) eliminates the need for perfect characterization of the physical learning system and shows promise for efficient training with large numbers of programmable parameters. The FFA does not require backpropagating an error signal to update the weights, rather the weights are updated by only sending information in one direction. The local loss function for each set of trainable weights enables low-power analog hardware implementations without resorting to metaheuristic algorithms or reinforcement learning. In this paper, we present an experiment utilizing multimode nonlinear wave propagation in an optical fiber demonstrating the feasibility of the FFA approach using an optical system. The results show that incorporating optical transforms in multilayer NN architectures trained with the FFA can lead to performance improvements, even with a relatively small number of trainable weights. The proposed method offers a new path to the challenge of training optical NNs and provides insights into leveraging physical transformations for enhancing the NN performance.

Online few-shot learning for multi-function radars mode recognition based on backtracking contextual prototypical memory

Modern multi-function radars (MFRs) can flexibly generate multiple fine-grained working modes for different missions through programmable parameters. Precise recognition of these working modes by analyzing the parameter combinations lays foundation for situational awareness. This recognition becomes more challenging to electromagnetic reconnaissance system when prior information of radiation source is unavailable and effective labeled signal data is not sufficient in adversarial scenarios. The presence of incremental mode involved in radar data flow requires that the recognition model enables to dynamic adjust and online perceive current data. This issue incorporates online learning to few-shot learning to accomplish efficiently recognition to a novel mode with the support of small amount of data, and still retain model's ability to recognize existing modes. This paper designed a backtracking contextual prototypical memory (BCPM) network for online MFR mode recognition. The proposed method learns temporal and spatial information from data stream as a reference predicting the signal segments as existing working modes or a novel mode. The BCPM network also designs a backtracking module and an alignment regularization term to utilize the data without annotation adequately and obtain a more reliable category representation. The experimental results verified the proposed method's good performance and high robustness to non-ideal factors and distractors.

Analysis of trajectory and motion parameters of an industrial robot cooperating with a numerically controlled machine tools

The tendency to improve manufacturing processes also applies to robotic and automated processes. Optimization of the trajectory and motion parameters of an industrial robot is one of the most important aspects of effective production. Robots are often not independent mechanical units, but interact with other work machines. The effective cooperation of mechanical devices is a significant technological challenge. In this paper, the types of positioning, trajectories and motion parameters of an industrial robot cooperating with numerically controlled machine tools were analyzed. The considered example is the process of machining the riveting tool holder. The tasks of offline programming of numerically controlled machine tools and robots were combined. For this purpose, SinuTrain and RobotStudio software were used. The main simulation analyzes indicated the relationship between the programmable parameters of the robot's motion and the actual values of the speed and acceleration of the tool center point. In the paper, the influence of speed and positioning accuracy on the duration of the process were also indicated. The results of the analyzes allowed to optimize the working time of the robot while maintaining the appropriate quality of the process.

Few-Shot Recognition of Multifunction Radar Modes via Refined Prototypical Random Walk Network

Multi-functional radars (MFRs) can generate a variety of working modes for different tasks based on flexible modulation types and programmable parameters. The neural network has been widely used to recognize of these fine-grained MFR modes. However, it requires a large number of samples with expert annotation in advance, which is hardly available in practical applications. Therefore, few-shot learning (FSL) method is used to learn “general information” and transfer it into new tasks where only a small number of labeled samples are provided. In order to improve its effect, unlabeled samples are utilized to provide “manifolds information” that describes the distribution among data. This paper proposes a framework of coding refined prototypical random walk network (C-RPRWN) combining these two kinds of information. The whole framework is divided into three modules: preprocessing module, embedding module and refined prototypical random walk (RPRW) module, which are respectively used to enhance the signal expression to adapt to non-ideal situations, extract distinguishable features to compute prototypes, and utilize “manifolds information” for better classification. The experimental results and analysis show that the proposed method achieves excellent performance for MFR fine-grained modes recognition even under the condition of a small number of samples. In addition, the robustness of the proposed method is verified under the influence of different non-ideal factors.

Sub-10 μHz/√Hz Measurement Instrumentation for 140-dB DR Frequency-Modulated MEMS Sensors

The field of frequency-modulated (FM) microelectromechanical system (MEMS) sensors is emerging as an alternative to conventional amplitude-modulated (AM) architectures, especially because of new application requirements on extended full-scale and low noise, which turn into a wide dynamic range (DR). As a consequence, a need for characterization tools dedicated to this novel class of transducers is arising. This work presents the development of the key building blocks of an instrument for the characterization of FM sensors, including an integrated circuit and external board-level components. In the instrument design, the focus is set on versatility in terms of input range frequencies and on guaranteeing a DR in the order of 140 dB. The instrument core is based on a frequency-to-digital converter, implemented through a type-II phase-locked loop with a differentiator implementing a period meter. Low-quantization noise is obtained using a high-frequency (100 MHz) time reference, while versatility is guaranteed at the same time by having programmable parameters within the integrated circuit. The instrument is tested across the 27–260-kHz input range and demonstrates sub-10- <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text {Hz}/\sqrt {\text {Hz}}$ </tex-math></inline-formula> frequency noise density in the range from 10 to 100 Hz offset from the carrier and up to 100-Hz full scale. The instrument is available for testing and used by whoever is interested in its features.

Mode Recognition of Multifunction Radars for Few-Shot Learning Based on Compound Alignments

The multifunctional radars can switch among a variety of fine-grained working modes, which often have flexible modulation types and programmable parameters. In an electromagnetic reconnaissance system, the process of identifying different working modes in pulse sequences guarantees the subsequent intention analysis and assists in devising a jamming strategy. Most of the existing working mode recognition methods attempt to establish a machine learning mechanism by training a model using a large number of annotated samples. However, this is hardly applicable in the real-world scenarios where only a few samples can be intercepted in advance. As the labeled signal samples are expensive, one direction is to augment the dataset by generating either samples or signal features in an embedding space. In this article, inspired by the fact that different modalities of the same working modes are generated from a set of fixed parameters, a few-shot learning framework based on compound alignments is proposed. Three branches, which take observations of long windows, observations of short windows, and coded semantic attributes as inputs, are aligned in both the latent variable space and the reconstruction space to learn a shared embedded space. A simple soft-max classifier is trained by sampling sufficient instances in the learned space to realize the final identification process. The experimental results and analysis show that the proposed method achieves excellent performance for fine-grained working mode recognition even for a small number of observed pulses. In addition, the proposed method is robust under different nonideal conditions, such as noise contamination, incomplete sequences, and spurious pulses.

A Fully Implantable Opto-Electro Closed-Loop Neural Interface for Motor Neuron Disease Studies.

This paper presents a fully implantable closed-loop device for use in freely moving rodents to investigate new treatments for motor neuron disease. The 0.18 μm CMOS integrated circuit comprises 4 stimulators, each featuring 16 channels for optical and electrical stimulation using arbitrary current waveforms at frequencies from 1.5 Hz to 50 kHz, and a bandwidth programmable front-end for neural recording. The implant uses a Qi wireless inductive link which can deliver >100 mW power at a maximum distance of 2 cm for a freely moving rodent. A backup rechargeable battery can support 10 mA continuous stimulation currents for 2.5 hours in the absence of an inductive power link. The implant is controlled by a graphic user interface with broad programmable parameters via a Bluetooth low energy bidirectional data telemetry link. The encapsulated implant is 40 mm × 20 mm × 10 mm. Measured results are presented showing the electrical performance of the electronics and the packaging method.

How to: Basics of programming cardiac pacemakers

Die Programmierung von Schrittmachersystemen, insbesondere von Zweikammerschrittmachersystemen kann die untersuchenden Ärzte vor große Herausforderungen stellen. Eine genaue Kenntnis der zu programmierenden Parameter sowie der herstellerspezifischen Algorithmen ist essentiell. Bei der Programmierung sollte darauf geachtet werden, die Programmierung den individuellen Bedürfnissen der Patient:innen anzupassen und „Out-of-the-box“-Programmierungen zu vermeiden. Ein weiteres wichtiges Ziel der Programmierung ist es, unnötige Stimulation im rechten Ventrikel zu vermeiden und dem Patienten eine gute Belastbarkeit zu ermöglichen sowie zu vermeiden, dass er eine Stimulation wahrnimmt. Algorithmen der Hersteller können hierbei helfen, müssen jedoch verstanden und bei inadäquatem Verhalten ggf. deaktiviert werden.

Magnetoelectric Bio-Implants Powered and Programmed by a Single Transmitter for Coordinated Multisite Stimulation.

This paper presents a hardware platform including stimulating implants wirelessly powered and controlled by a shared transmitter for coordinated leadless multisite stimulation. The adopted novel single-transmitter, multiple-implant structure can flexibly deploy stimuli, improve system efficiency, easily scale stimulating channel quantity and relieve efforts in device synchronization. In the proposed system, a wireless link leveraging magnetoelectric effects is co-designed with a robust and efficient system-on-chip to enable reliable operation and individual programming of every implant. Each implant integrates a 0.8-mm2 chip, a 6-mm2 magnetoelectric film, and an energy storage capacitor within a 6.2-mm3 size. Magnetoelectric power transfer is capable of safely transmitting milliwatt power to devices placed several centimeters away from the transmitter coil, maintaining good efficiency with size constraints and tolerating 60-degree, 1.5-cm misalignment in angular and lateral movement. The SoC robustly operates with 2-V source amplitude variations that spans a 40-mm transmitter-implant distance change, realizes individual addressability through physical unclonable function IDs, and achieves 90% efficiency for 1.5-to-3.5-V stimulation with fully programmable stimulation parameters.

Stochastic Computing Emulation of Memristor Cellular Nonlinear Networks.

Cellular Nonlinear Networks (CNN) are a concept introduced in 1988 by Leon Chua and Lin Yang as a bio-inspired architecture capable of massively parallel computation. Since then, CNN have been enhanced by incorporating designs that incorporate memristors to profit from their processing and memory capabilities. In addition, Stochastic Computing (SC) can be used to optimize the quantity of required processing elements; thus it provides a lightweight approximate computing framework, quite accurate and effective, however. In this work, we propose utilization of SC in designing and implementing a memristor-based CNN. As a proof of the proposed concept, an example of application is presented. This application combines Matlab and a FPGA in order to create the CNN. The implemented CNN was then used to perform three different real-time applications on a 512 × 512 gray-scale and a 768 × 512 color image: storage of the image, edge detection, and image sharpening. It has to be pointed out that the same CNN was used for the three different tasks, with the sole change of some programmable parameters. Results show an excellent capability with significant accompanying advantages, such as the low number of needed elements further allowing for a low cost FPGA-based system implementation, something confirming the system’s capacity for real time operation.

Cytomorphic Electronic Systems: A review and perspective.

Boltzmann-exponential thermodynamic laws govern noisy molecular flux in chemical reactions as well as noisy subthreshold electron current flux in transistors. These common mathematical laws enable one to map and simulate arbitrary stochastic biochemical reaction networks in highly efficient cytomorphic systems built on subthreshold analog circuits. Such simulations can accurately model noisy, nonlinear, asynchronous, stiff, and non-modular feedback dynamics in interconnected networks in the physical circuits, automatically. The scaling in simulation time for stochastic networks with the number of reactions or molecules is constant in cytomorphic systems. In contrast, it grows rapidly in digital systems, which are not parallelizable. Therefore, cytomorphic systems enable large-scale supercomputing systems-biology simulations of arbitrary and highly computationally intensive biochemical reaction networks that can nevertheless be compiled to them via digitally programmable parameters and connectivity. We outline how cytomorphic systems can be utilized for rapid drug-cocktail formulation and discovery in future pandemics like COVID-19; can simulate networks important in cancer; and can help automate the design of synthetic biological circuits, e.g. a synthetic biological operational amplifier for robust and precise drug delivery. Thus, just as neuromorphic systems have enabled multiple applications in A.I., cytomorphic systems will enable multiple applications in biology and medicine.

Porous aligned ZnSr-doped β-TCP/silk fibroin scaffolds using ice-templating method for bone tissue engineering applications

The bone is a complex and dynamic structure subjected to constant stress and remodeling. Due to the worldwide incidence of bone disorders, tissue scaffolds and engineered bone tissues have emerged as solutions for bone grafting, which require sophisticated scaffolding architectures while keeping high mechanical performance. However, the conjugation of a bone-like scaffold architecture with efficient mechanical properties is still a critical challenge for biomedical applications. In this sense, the present study focused on the modulating the architecture of silk fibroin (SF) scaffolds crosslinked with horseradish peroxidase and mixed with zinc (Zn) and strontium (Sr)-doped β-tricalcium phosphate (ZnSr.TCP) to mimic bone structures. The ZnSr.TCP-SF hydrogels were tuned by programmable ice-templating parameters, and further freeze-dried, in order to obtain 3D scaffolds with controlled pore orientation. The results showed interconnected channels in the ZnSr.TCP-SF scaffolds that mimic the porous network of the native subchondral bone matrix. The architecture of the scaffolds was characterized by microCT, showing tunable pore size according to freezing temperatures (−196 °C: ∼80.2 ± 20.5 µm; −80 °C: ∼73.1 ± 20.5 µm; −20 °C: ∼104.7 ± 33.7 µm). The swelling ratio, weight loss, and rheological properties were also assessed, revealing efficient scaffold integrity and morphology after aqueous immersion. Thus, the ZnSr.TCP-SF scaffolds made of aligned porous structure were developed as affordable candidates for future applications in clinical osteoregeneration and in vitro bone tissue modelling.

Design and Implementation of Partial Shared Digital Channelized Receiver

A novel efficient partial sharing channelization structure with odd and even stacking is designed and implemented. There are two special designs in the proposed structure. Firstly, by the intensive channel overlap design, for non-cooperative wideband signals, the proposed structure can achieve good parameter estimation accuracy and high probability of complete interception. Secondly, based on the partial sharing design developed in this paper, the computation burden of the proposed structure can be greatly reduced compared with the traditional directly implemented structures. Experiments and numerical simulations are conducted to evaluate the proposed structure, which shows its improvements over traditional methods in terms of field programmable gate arrays (FPGA) resource consumption and parameter estimation accuracy.

Nucleic acid nanoparticles (NANPs) as molecular tools to direct desirable and avoid undesirable immunological effects.

Nucleic acid nanoparticles (NANPs) represent a highly versatile molecular platform for the targeted delivery of various therapeutics. However, despite their promise, further clinical translation of this innovative technology can be hindered by immunological off-target effects. All human cells are equipped with an arsenal of receptors that recognize molecular patterns specific to foreign nucleic acids and understanding the rules that guide this recognition offer the key rationale for the development of therapeutic NANPs with tunable immune stimulation. Numerous recent studies have provided increasing evidence that in addition to NANPs' physicochemical properties and therapeutic effects, their interactions with cells of the immune system can be regulated through multiple independently programmable architectural parameters. The results further suggest that defined immunomodulation by NANPs can either support their immunoquiescent delivery or be used for conditional stimulation of beneficial immunological responses.

Optical Purity as a Programmable Variable for Controlling Polyolefin Tacticity in Living Coordinative Chain Transfer Polymerization: Application to the Stereomodulated LCCTP of α,ω-Nonconjugated Dienes

Two-state, stereomodulated living coordinative chain transfer polymerization (LCCTP) of 1,6-heptadiene has been achieved to produce a series of cis-poly(methylene-1,3-cyclohexane) (cis-PMCH) with stereochemical microstructures that vary between two limiting forms as a function of the optical purity of the chiral C1-symmetric hafnium preinitiator 1 when activated by the borate co-initiator B1 in the presence of a fixed excess amount of ZnEt2. Optical purity as a programmable variable was demonstrated by premixing the pure homochiral (SC,SHf) and (RC,RHf) enantiomers of 1 in different ratios prior to activation. In contrast, the LCCTP of 1,5-hexadiene using a “racemic” mixture of 1 to produce cis/trans-poly(methylene-1,3-cyclopentane) (cis/trans-PMCP) was less successful due to the loss of stereocontrol for insertion under chain transfer conditions that arises with site epimerization at the transition metal center. A correlation is proposed between the diastereoselectivity and efficiency for the cyclization step in propagation and the magnitude of site epimerization due to reversible chain transfer. These results serve to validate the strategy of using the optical purity of an initiator as a programmable parameter for controlling polyolefin tacticity as well as to better establish the criteria that are required for achieving successful stereomodulated LCCTP for the production of next generation polyolefin materials.

ОПТИМИЗАЦИЯ ЦИФРОВОГО ПРОЕКТИРОВАНИЯ СЛОЖНОЙ ТЕХНОЛОГИЧЕСКОЙ СИСТЕМЫ

The article investigates the issue of optimization of a complex technical system at the stage of its design using a heuristic–phenomenological approach. The analysis of the principles of complex optimization of complex technical systems is carried out. A four–level structure for the synthesis of methodological, informational and software support for complex optimization of complex technical systems is proposed, which will allow controlling the programmable parameters of complex technical systems at each stage of their design.