Journal of Circuits, Systems and Computers

Intelligent Traffic Management is a crucial issue closely related to daily life and productivity, with traffic congestion being a complex and challenging problem faced by most cities. Traffic Signal Control (TSC) stands out as the most direct and effective method to tackle congestion. It aims to minimize travel time, enhance throughput, improve traffic safety, reduce emissions, and conserve energy by coordinating the direction and timing of vehicle movements at intersections. Traditional TSC methods mostly rely on simple rules, limited data, and expert knowledge, making them inadequate for increasingly complex traffic scenarios. In the context of TSC, an increasing number of researchers are turning to Deep Learning (DL) methods to address identification, decision-making, and optimization challenges. Although many reviews have examined the TSC problems and the application of Reinforcement Learning in this field, there remains a notable gap in comprehensive analyses of TSC utilizing a wider range of DL techniques, including Deep Reinforcement Learning, Federated Learning, and Meta-learning. This paper, building upon the basic concepts and traditional approaches of TSC, provides a detailed overview of the latest research advancements employing different DL methods for this issue. Experimental settings and evaluations are also introduced. Furthermore, to spark new interest in this research field, future works are proposed.

Traffic prediction is essential for transportation planning, resource allocation, congestion management and enhancing travel experiences. This study optimizes data preprocessing techniques to improve machine learning-based traffic prediction models. Data preprocessing is critical in preparing the data for machine learning models. This study proposes an approach that optimizes data preprocessing techniques, focusing on flow-based analysis and optimization, to enhance traffic prediction models. The proposed approach explores fixed and variable orders of data preprocessing using a genetic algorithm across five diverse datasets. Evaluation metrics such as root mean squared error (RMSE), mean absolute error (MAE) and R-squared error assess model performance. The results indicate that the genetic algorithm’s variable order achieves the best performance for the ArcGIS Hub and Frementon Bridge Cycle datasets, fixed order one preprocessing for the Traffic Prediction dataset and variable order using the genetic algorithm for the PeMS08 dataset. Fixed order 2 preprocessing yields the best performance for the XI AN Traffic dataset. These findings highlight the importance of selecting the appropriate data preprocessing flow order for each dataset, improving traffic prediction accuracy and reliability. The proposed approach advances traffic prediction methodologies, enabling more precise and reliable traffic forecasts for transportation planning and management applications.

Day-to-day mobility among the population has increased with economic growth. Smart cities are renovated with advanced technologies to admire modern life in which intelligent transportation becomes highly focused. Because the traffic signal control systems are fixed at a constant time. They split the traffic signal into predetermined intervals and function inefficiently; they result in long wait times, waste fuel and increased carbon emissions. This research study introduces a novel technique for traffic light management to reduce the uncertainties in the system. A dynamic and intelligent traffic light adaptive optimal management system (DITLAOCS) is implemented in this research. It does this by modifying the traffic signal duration in run time and using real-time traffic data as input. Furthermore, the proposed DITLAOCS executes based on three modes: fairness mode (FM), priority mode (PM) and emergent mode (EM). In fairness mode (FM), all vehicles are prioritized equally, while vehicles in different categories receive varying priority levels. Emergency vehicles, on the other hand, receive the highest priority. Furthermore, a fuzzy inference method based on traffic data is shown to choose one mode out of three (FM, PM and EM). This model uses deep reinforcement learning to switch traffic lights in three different phases (red, green and yellow). We evaluated and accurately simulated DITLAOCS on the Shaanxi city map in China using Simulation of Urban MObility (SUMO), an open-source simulator. The simulation results illustrate the efficiency of DITLAOCS when compared to other cutting-edge algorithms on several performance measures.

With the increasing popularity of renewable energy, many power electronics are connected to the grid. This integration can induce subsynchronous oscillations (SSO) due to interactions between power electronic converters or between converters and the AC grid. SSO poses a severe threat to the safe and stable operation of power systems, necessitating the development of a real-time online SSO assessment method. Focusing on the SSO caused by integrating doubly-fed induction generators (DFIGs) into the power grid via series compensation transmission lines, this study proposes an SSO assessment method based on improved transformer networks and adaptive selection classifiers. First, the IMF-IE information is extracted to describe SSO characteristics, and a scenario classification model based on P-DDQN is established. The introduction of the prioritized experience replay (PER) mechanism accelerates the learning process and enhances performance. Next, the interference level is calculated based on data energy information for scenarios that include SSO. Subsequently, an SSO parameter estimator is developed for PMU data corresponding to different interference levels. This estimator integrates multi-head attention (MHA) with an improved sparse autoencoder to balance accuracy and speed in the assessment. Finally, a simulation model is constructed to validate the proposed method. The simulation results demonstrate that the proposed method can quickly and accurately estimate multiple oscillation modes’ frequencies and damping ratios in complex field scenarios.

In wireless sensor networks, accurate node localization is essential for ensuring the precision of data collection. The DV-Hop algorithm, a popular range-free localization method, estimates distances between nodes by multiplying hop counts with average hop distances obtained through distance vector routing. However, this algorithm often experiences localization errors in randomly distributed network environments due to considerable inaccuracies in average hop distance calculations and the approximation of actual paths by straight-line paths. This paper introduces an enhanced DV-Hop localization algorithm, which constructs a mathematical model to optimize the mean square error of the average hop distance for any anchor node. This optimization corrects the average hop distance across the network, bringing it closer to the actual value, thus reducing errors and enhancing accuracy. Simulation results indicate that with 150 nodes, a 30% beacon node ratio and a 100-m communication range, the localization error of the improved FuncDV-Hop model decreased from 0.3916 to 0.1705, and the Root Mean Square Error (RMSE) decreased from 24.78m to 14.39m, thereby improving localization accuracy by 56.46%.

As railway engineering moves toward a more severe environment, the service safety of track structure infrastructure needs more detailed control, and correct cognitive mechanical behavior of infrastructure has always been the difficulty and bottleneck of refined analysis of railway engineering. In this thesis, taking the stress–sensitive nonlinear constitutive of subgrade as a breakthrough, this thesis puts forward that the constitutive relationship that is self-consistent with material behavior is embedded in the finite element calculation model of subgrade, which provides a feasible method for fine analysis of railway engineering. In the study, the bottom layer of the subgrade bed for the Ballastless track is taken as the research object. The stress-related nonlinear constitutive relationship N37A is inserted into the subgrade material through a user-defined subroutine. Then, the material’s rebound modulus evolution parameters returned from the triaxial test are used to analyze the nonlinear base mechanical behavior of the embedded material. The following conclusions are obtained: (1) The N37A nonlinear model can capture the stress sensitivity of the material, and describe the mechanical behavior of the bottom layer of subgrade bed factually. (2) The deadweight has led to the nonlinear significant increase of the resilient modulus, and the train load causes the different development of the stress–strain response at different locations. (3) The resilient modulus of the bottom layer of the subgrade bed is small at the top surface but large at the bottom surface, which is monotonically increasing. With the loading of the train, the resilient modulus curve with different depths rotates approximately clockwise at the center about 0.75m away from the top surface of the bottom layer of the subgrade bed. (4) The value of resilient modulus at the top surface of the bottom layer of the subgrade bed is as the center line of the subgrade $>$ under the rail $>$ the edge of the base, and the bottom surface is just opposite. The relative difference in resilient modulus between them is less than 3MPa.

As artificial intelligence technology continues to evolve, numerous advancements enable smoother communication and collaboration between humans and computers. Natural language processing (NLP), as a key direction in the field of computer and artificial intelligence, has been attracting much attention. It focuses on the language used by people in daily life, devoting itself to improving the ability of language generation and understanding, thus making communication between people and computers more fluent and natural, effectively breaking through the problem of human-computer interaction. With the advancement of deep learning technology, the migration of NLP models onto deep learning platforms has emerged as a key trend. Model transfer, a crucial aspect of deep learning, holds significant value in enhancing the performance and efficiency of NLP models. This paper begins by outlining the fundamentals of deep learning platforms and NLP model transfer, followed by a comprehensive examination of current research progress and challenges in this field. Subsequently, we introduce a novel NLP model transfer strategy tailored for deep learning platforms and validate its effectiveness through rigorous experiments. In conclusion, the paper highlights our noteworthy advancements, pointing toward promising future developments.

This work presents an all-digital time-mode subthreshold temperature sensor for monitoring SoC self-heating problem. The sensor incorporated a low-power differential inverter interlaced cascaded delay cell followed by a gated ring oscillator with different temperature coefficients of $-$1736.10 and $-$5880.33 ppm/^∘C, respectively. The differential inverter interlaced cascaded delay cell and XOR gate produce a temperature-dependent large delay pulse, whereas the gated ring oscillator produces temperature-dependent output code for each point of temperature within the delay pulse generated by the differential inverter interlaced cascaded delay cell and XOR. The corresponding temperature code is recorded by an asynchronous counter. At 90-nm CMOS process and 0.6/0.3V supply voltage Vdd, the sensor occupies 0.009mm² and consumes only 876nW at 27^∘C. The sensor has a sampling rate of 52.16KS/s and the resolution is noted to be 0.40^∘C for a temperature span of 0–100^∘C. Besides, the proposed design achieves a very small energy/conversion of 17 pJ, which ensures 2.72 pJ$\times$C² and 0.13 nJ%² resolution FoM and inaccuracy FoM, respectively.

This work focuses on an intelligent management framework for vocal music media resources, with use of deep neural networks and web crawlers. The web driver module is used to simulate the Google browser access mode to collect and preprocess the data of the China vocal music network webpage project, and the delay code is added to solve the problem of Internet Protocol (IP) restrictions on crawling webpages. After parsing, the content of the project description is obtained and temporarily stored in the local file through the panda’s module to realize the automatic collection of project data. Membership functions of four basic types of emotions are designed. By utilizing the analytic hierarchy process, the experience and wisdom of many music experts are gathered. The weights of activity and the evoking force in determining the music emotion membership value are discussed, trying to make the membership function closer to reality. Through the application of the music emotion analysis model, the music emotion analysis technology is combined with the recommendation algorithm, and the music recommendation algorithm combined with the emotion analysis is realized. The experimental test proves that the algorithm has higher accuracy than the traditional music recommendation algorithm.

The monitoring and early warning of financial risks have become a crucial link in maintaining market stability and safeguarding the rights and interests of investors. Traditional financial risk monitoring methods often rely on a single data source or analysis model, making it challenging to comprehensively and accurately capture risk signals. Therefore, this paper proposes a novel financial risk monitoring model based on multimodal neural networks, which innovatively integrates multiple data sources, such as vision, language and audio, and utilizes their inherent correlations to enhance the accuracy of risk identification. First, by employing the Bidirectional Long Short-Term Memory Network (BiLSTM) structure and incorporating the self-attention mechanism, the semantic information of financial texts is deeply analyzed through the calculation of dynamic weight coefficients. Additionally, Option-based Hierarchical Reinforcement Learning (OHRL) is utilized to accurately model the behavior of market participants, capturing nuanced changes in their decision-making process. By integrating these two types of information, a comprehensive BiLSTM-OHRL model is formulated to evaluate the risk status of financial markets in a more comprehensive and accurate manner. The results demonstrate that the model performs impressively in financial risk monitoring, accurately capturing the emotional and behavioral characteristics of market participants, thereby enhancing the comprehensiveness and predictive capability of the monitoring model. It provides robust technical support for the stable operation of the financial market.

Current research on lane-changing for intelligent connected vehicles (ICVs) primarily focuses on the control of a single ICV. Even studies on cooperative lane-changing among multiple ICVs often use fixed control algorithms, which are difficult to adapt to constantly changing traffic conditions. This paper proposes a method based on the deep deterministic policy gradient (DDPG) model. The DDPG algorithm has strong generalization capabilities, enabling it to handle complex traffic scenarios. Combined with the design of innovative reward functions, it successfully achieves the control of cooperative lane-changing for multiple ICVs. The novelty of this model is mainly reflected in the implementation of multiple ICVs control and cooperative strategies through the design of a speed reward function. Specifically, when two or more ICVs come within a 15-m range, a collaborative strategy is implemented. In this strategy, the original speeds of the vehicles involved are replaced with their average speed. This average speed is then used to compute the velocity reward. Moreover, an additional headway reward is introduced when two vehicles are on the same lane with a distance of less than 10m between them, further optimizing lane-changing decisions and enhancing driving efficiency. The innovative approach described in this paper has been validated through simulations in the highway-env simulation environment. The model’s effectiveness was tested under various conditions, including different road densities and numbers of ICVs. Results indicate that with an increasing number of episodes, the model consistently enhances both the fleet’s and each individual vehicle’s average speed and steps. It also ensures that the headway is maintained within a desired range of 10–20m. This paper showcases the potential applications of a new DDPG-based multi-networked vehicular cooperative lane-changing decision model across different scenarios. These findings provide important reference values for future research and development in intelligent transportation systems (ITS), particularly in optimizing vehicle cooperative behaviors and enhancing traffic efficiency. This paper emphasizes the model’s potential to significantly impact the management and operation of future vehicular networks, paving the way for more sophisticated and efficient transportation systems.

Biomedical applications like body area networks (BAN) necessitate the construction of power optimized SRAMs to enhance the batteries life at BAN nodes. In this work, we have designed a one-sided near-threshold 9TSRAM for low-power portable biomedical applications. The proposed near threshold 9T SRAM (PNT9T SRAM) employs a cross-connected Schmitt trigger (ST) inverter and normal inverter in its cell core. A separate path for reading is also employed to eliminate the reading disturbance. The write disturbance is removed in the PNT9T SRAM by removing the trail from VDD and ground. The write ability is improved with the use of a feedback-cutting approach. The leakage power dissipation of the memory is mitigated by using a tail transistor, virtual ground (VGND). To evaluate the performance, the PNT9T SRAM is compared with conventional 6T (C6T), ST11T, ST9T, TG9T, SBL9T, and SE9T SRAM cells using FinFET 18nm technology at 0.6V power supply. The PNT9T SRAM mitigates the read power, write power, and leakage power by 51.10%, 50.57%, and 78.97%. Furthermore, the read and write static noise margins were improved by 54% and 39.5%, respectively, compared to C6T SRAM.

This paper presents a novel design of second-order triple band waveguide bandpass filter using planar insert technology. A conventional WR-90 rectangular waveguide has two identical inserts inserted into its transverse plane at an optimal spacing of 8.41mm. The insert consists of a combination of a double-split square resonator and a single split square resonator with stub. These resonators are used to achieve triple-band bandpass characteristics. The structure allows independent control of three center frequencies, insertion loss, return loss and 3-dB bandwidth. The simulation analysis of the proposed filter is carried out using the CST microwave studio, and its frequency response is compared with its equivalent circuit response. To validate the numerical analysis, the WR-90 waveguide and planar insert are fabricated independently and inserted manually. The dual-pole triple band bandpass characteristic of the manufactured filter, with center frequencies of 8.19, 9.39 and 11.09GHz, is shown by the measured data. The complete design process has also been provided, along with a corresponding circuit diagram. The suggested waveguide filter’s performance is evaluated against those found in the literature already.

With the rapid advancement of Digital Twins (DTs) technologies, they have emerged as viable solutions for enhancing the control and optimization of thermal power combustion processes. This paper presents a pioneering approach that leverages DTs to revolutionize the combustion process control system within thermal power generating units. Focused on bridging the physical and digital realms, our research establishes DTs of the combustion process control system (CPCS), situated within the Networked Control System Laboratory (NCSLab). We propose a novel five-dimensional (5D) DTs model encompassing the physical, data, service, connection, and virtual models, thereby facilitating a comprehensive fusion of physical and digital information. By analyzing the demands and technical intricacies of the combustion process control system, we design and implement a virtual 3D model, enriching the visualization and analytical capabilities of the DT environment. Furthermore, through controlled experiments on the combustion process and rigorous comparison with offline simulations, we validate the efficacy and feasibility of our approach. Importantly, our research opens avenues for future exploration, particularly in harnessing industrial real-time data to drive the virtual 3D model. Through the integration of Artificial Intelligence (AI) techniques such as Deep Learning, we elevate the capabilities of our DT framework. This enhanced predictive analytics, dynamic optimization, human–machine interaction, and virtual model realism. This research contributes to the advancement of DT technology in the context of thermal power plants and holds promise for driving innovation and efficiency across diverse industrial sectors.

This paper analyzes the efficiency of different FPGA architectures consisting of Hybrid logic blocks (HLB) formed with Look-up Tables (LUTs) and Universal Logic Gates (ULGs). Pure LUT-based architectures induce area overhead and a longer delay. A ULG is designed to realize an efficient architecture compared to pure LUT-based architectures. Using this ULG and LUT in different proportions, HLBs are designed. The Field Programmable Gate Array (FPGA) architectures containing these proposed HLB structures are analyzed using various Microelectronics Centre of North Carolina (MCNC) benchmark circuits and compared with the performance of the existing designs. The result shows that the proposed architecture design can save a tremendous amount of time in overall flow and routing along with area improvements.

Content Addressable Memories (CAMs) are high-speed hardware lookup tables that are crucial in routing, packet classification, search, and other fast look-up applications. Despite having the advantage of high speed, they have limitations due to their power hungriness. This paper addresses the limitations of existing CAMs with two novel modifications in the proposed Search Line Pre-Charge-Free Low-Power-Content Addressable Memory (SLPFCAM) structure. The first novelty is their gated evaluation matching circuit and the second is a match line pre-charge controller. The proposed design reduces the energy consumption and improves the speed of the CAM by eliminating the SLprecharge phase in the CAM operation. The pre-charge controller prevents the attempt of match line pre-charge when it is not required. The design has been implemented using Cadence Virtuoso in a 90-nm CMOS technology with 1-V supply, and the post-layout simulation results show that the SLPFCAM makes significant improvements with a minimum of 83% less power consumption, a 91% less search delay, and the average energy consumption being 0.765 fJ/bit/search.

Truck platooning technology has been widely studied for its advantages in traffic efficiency and energy savings. Although recent studies have made significant progress in specific scenarios, research on merging ramp trucks into the main road truck platoons remains insufficient. This paper proposes an innovative cooperative control method for merging ramp trucks and main road truck platoons based on vehicle-to-everything (V2X) technology. The method integrates the intelligent driver model (IDM), acceleration control logic for merging trucks before merging, lane-changing decision logic, steering control methods, and on-board unit (OBU) and roadside unit (RSU) communication technologies. This ensures that autonomous merging trucks can smoothly join the main road truck platoon under various insertion positions and acceleration lane lengths. After successfully joining the platoon, the merging trucks can maintain appropriate spacing and high speed, thereby improving overall traffic efficiency. Further analysis on different insertion positions and acceleration lane lengths reveals that when using the proposed method, merging trucks have the least impact on the overall platoon when they join at the head or tail of the main road truck platoon. Although the proposed method performs better with longer acceleration lanes, these lanes are costly to construct. The proposed method can also enable successful platoon joining with shorter acceleration lanes, optimizing the use of acceleration lanes to some extent.

Aspect-level sentiment analysis is a critical component of sentiment analysis, aiming to determine the sentiment polarity associated with specific aspect words. However, existing methodologies have limitations in effectively managing aspect-level sentiment analysis. These limitations include insufficient utilization of syntactic information and an inability to precisely capture the contextual nuances surrounding aspect words. To address these issues, we propose an Aspect-Oriented Graph Attention Network (AOGAT) model. This model incorporates syntactic information to generate dynamic word vectors through the pre-trained model ALBERT and combines a graph attention network with BiGRU to capture both syntactic and semantic features. Additionally, the model introduces an aspect-focused attention mechanism to retrieve features related to aspect words and integrates the generated representations for sentiment classification. Our experiments on three datasets demonstrate that the AOGAT model outperforms traditional models.

This paper presents a novel footless single clock-phase three-stage comparator with internally generated regenerative voltage signals for low kickback noise and high speed. The preamplifier and clocked latch topology of dynamic comparators are considered in this brief. The proposed design has been compared by analyzing and optimizing three state-of-the-art comparator designs: Modified StrongARM, Miyahara’s and Three-Stage. These designs are simulated using the 40nm CMOS technology process. The area of the proposed comparator is 28.025$\mu$m². When simulated in the SPICE-based simulator HSPICE, under typical performance conditions of 0.9V supply voltage, 25^∘C, 1 GHz operational frequency, typical process corner (tt), and 1mV differential voltage, the proposed comparator produces the best post-layout results, with a minimum delay of 59.9ps from meta-stability analysis, energy per comparison of 0.175pJ/comparison and a kickback noise of 44.55 $\mu$A. The proposed design also shows a significant improvement in the measured performance parameters over the other state-of-the-art dynamic comparators that have been discussed in this brief. The obtained results show that the proposed comparator is appropriate for 12-bit Successive Approximation Resister Analog-to-Digital Converters (SAR-ADCs) in IoT applications.

This paper presents the design of an eight-channel digital receive beamformer application-specific integrated circuit (ASIC) based ultrasound (US) imaging system. The designed prototype system supports up to 64 transmit and 16 receive channels. ASIC-based 16-channel US transceiver board is designed for frontend processing, and for backend processing Intel${}^{\text{®}}$ Core™-i5 based PC is employed. The transceiver board employs two ASICs for digital receive beamforming and a field-programmable gate array (FPGA) for transmit beamforming and system control. The necessary power supplies for the transceiver board are provided by a custom designed and developed high voltage (HV) and low voltage (LV) power supply board. The transceiver board performs the frontend processing like transmit beamforming, control, analog signal conditioning and digital receive beamforming. The received beamformed data are directly transferred to a host PC via the Gigabit Ethernet interface for image formation, post-processing and display. The ASIC-based system developed can provide real-time B-mode imaging with a maximum frame rate of 102 fps with a field of view (FOV) of $90^{\circ }$. These results indicate that the ASIC-based US imaging systems can provide better processing capability and performance.