CPU Usage Research Articles

Reduced CPU Workload for Human Pose Detection with the Aid of a Low-Resolution Infrared Array Sensor on Embedded Systems

Modern embedded systems have achieved relatively high processing power. They can be used for edge computing and computer vision, where data are collected and processed locally, without the need for network communication for decision-making and data analysis purposes. Face detection, face recognition, and pose detection algorithms can be executed with acceptable performance on embedded systems and are used for home security and monitoring. However, popular machine learning frameworks, such as MediaPipe, require relatively high usage of CPU while running, even when idle with no subject in the scene. Combined with the still present false detections, this wastes CPU time, elevates the power consumption and overall system temperature, and generates unnecessary data. In this study, a low-cost low-resolution infrared thermal sensor array was used to control the execution of MediaPipe’s pose detection algorithm using single-board computers, which only runs when the thermal camera detects a possible subject in its field of view. A lightweight algorithm with several filtering layers was developed, which allowed the effective detection and isolation of a person in the thermal image. The resulting hybrid computer vision proved effective in reducing the average CPU workload, especially in environments with low activity, almost eliminating MediaPipe’s false detections, and reaching up to 30% power saving in the best-case scenario.

Monintainer: An orchestration-independent extensible container-based monitoring solution for large clusters

Container virtualization has recently gained popularity due to its low performance and resource allocation overhead. The rise of this technology can be attributed to the advancement of cloud computing and the adoption of micro-services architecture. These new approaches offer a more efficient and fine-grained system design through the benefits of containerization, such as isolation, portability, and improved performance. However, container-based systems have created new challenges in monitoring due to their automated flexibility, ephemerality, and the increasing number of containers in a system. So there is a practical need for effective monitoring and performance management tools. This paper analyses the key performance metrics for machine, container and application services, including CPU usage, memory usage, disk usage, and network usage. Furthermore, we review several widespread tools for collecting and monitoring these metrics and present the Monintainer tool. It is a solution designed to monitor entire container-based systems, from applications to their underlying infrastructure, allowing users to better understand their systems’ behavior in run-time. The tool’s results can aid container-based systems’ design, implementation and optimization.

PiGateway: Real-time granular analysis of smart home network traffic using P4

The Internet of Things (IoT) makes the home smarter by real-time sensing of the situation and transmission of data to the cloud. As IoT devices continue to expand rapidly, a critical need arises for effective device-type identification methods. Accurate identification of IoT device types is essential for various applications, including network management, security, etc. To identify the type of IoT devices, existing works have proposed various Media Access Control (MAC)-based and Machine Learning (ML) approaches. These solutions lack real-time granular analysis and In-network classification of IoT traffic. In this work, we propose PiGateway: a P4-based gateway to classify the type of IoT devices at a line rate and prioritize the device type. Initially, PiGateway develops a Decision Tree (DT) model and generates flow rules from the model. The developed model is deployed in the PiGateway using two mechanisms. One mechanism is to convert the DT model into Match-Action Tables (MATs), where the tables store the flow rules. Another approach developed an optimized model by deploying the DT model into PiGateway as a mathematical expression. Further, to achieve the In-network classification, the gateway calculates flow-level features dynamically from incoming IoT traffic flow. The feature values are given as input to the deployed DT model for classifying the flow into a specific device type. Subsequently, the flow is prioritized based on device type and Type of Service (ToS) field value. Additionally, our proposed solution is designed and tested in a real-time environment. Through implementation results, we show that the PiGateway can classify the type of IoT devices with 98.61% accuracy and an average CPU usage of 10.68%. Also, the classification time in PiGateway is 0.098ms for 8000 flows using an optimized model. Finally, the proposed system serves as a gateway in a smart home IoT network.

Lightweight Intrusion Detection Method of Vehicle CAN Bus under Computational Resource Constraints

In order to improve the security protection performance of the vehicle Controller Area Network (CAN) bus, the research builds an adaptive lightweight intrusion detection algorithm based on the limited computing and storage resources of the on-board ECU environment and the message cycle characteristics to supervise and detect the vehicle CAN bus intrusion. The results showed that the message cycle-based adaptive intrusion detection algorithm had high accuracy and recall rate, and fast computational search efficiency, with a stable detection time of less than 3 seconds. The intrusion detection capability is continuously optimized as the training time increases, and after stabilization, the resource utilization rate reaches over 95% with a throughput of 100Mb/s. The algorithm has strong protection capabilities. The average vehicle CPU usage of the algorithm is only 4.76%, which is 10.17% lower than the intrusion detection algorithm based on support vector machines. It can effectively prevent interference with the normal operation of the vehicle CAN bus. The algorithm has high detection accuracy for interrupt type attacks, and there are no false positives or missed alarms. For injection type attacks, the probability of missed alarms is less than 1%. The intrusion detection of vehicle CAN bus based on the message cycle characteristics provides technical reference for the safety and stability of the vehicle network, and has important practical value for the intelligent and networked development of the automobile industry.

ASA APPLICATION (ARFF SMART APPLICATION) AS LEARNING MEDIA TO IMPROVE THE COMPETENCE OF ARFF PERSONNEL

The ARFF unit is responsible for incidents and accidents at the airport and its surroundings. They have skills such as the use of fire extinguishers, safety equipment operation, and evacuation measures. This preparedness is enhanced through the ARFF Smart Application (ASA) to improve ARFF learning competencies at I Gusti Ngurah Rai Airport. The research used a waterfall research and development (R&D) model with observation as the data collection method. Development involves communication, planning, modeling, construction, and implementation stages. Validation was conducted with a questionnaire according to ISO 25010, measuring the functionality, usability, compatibility, and performance efficiency of the application. Functionality testing involved six testers (3 media experts and 3 material experts), with 92% of the results showing excellent quality. Usability testing involved 37 users with an excellent 95.40% rating in ease of use, learning, and satisfaction. Application compatibility was tested on various Android devices. Performance efficiency testing showed efficient performance in execution time, CPU usage, and memory. ARFF Smart Application fulfills all four aspects of the ISO 25010 standard with an assessment of "very feasible" and is suitable for use in learning by personnel in classroom activities.

Region-aware dynamic job scheduling and resource efficiency for load balancing based on adaptive chaotic sparrow search optimization and coalitional game in cloud computing environments

Uneven distribution of workloads represents a significant challenge in cloud data centers, negatively impacting the efficient utilization of resources. To address this problem and optimize the performance-to-profit (PP) ratio, we introduced the Regional Awareness dynamic Scheduling Algorithm (RASA) through a three-phase approach. In the initial phase, tasks are categorized based on critical parameters like CPU usage, memory utilization, and I/O operations, using a task classification model. The second phase involves estimating the CPU, memory, and I/O capacities of each node. Subsequently, a unique merge-and-split-based coalitional game-theoretic process is employed to group nodes into clusters. In the final phase, we have improved task placement by introducing an enhanced Sparrow Search Algorithm (ISSA) to overcome the slow convergence and local optimization issues found in the traditional SSA. We have incorporated the Tent chaotic mapping approach to enhance both global and local search efficiency during this stage. Additionally, we have fine-tuned fair load distribution using the Cauchy mutation process. We performed simulations using the CloudSim toolkit, and the results underscore the RASA algorithm’s superiority in terms of how quickly it converges its precision in task placement and its ability to distribute workloads fairly. Our proposed algorithm achieves a 9% reduction in latency overhead, a 14% decrease in processing time, a 15% reduction in workload imbalance, a 19% decrease in energy consumption, and a 26% decrease in idle periods. Furthermore, it enhances resource availability and efficiency by 22% and 27% respectively, while simultaneously boosting service throughput by 32%.

On the impact of event-driven architecture on performance: An exploratory study

Event-driven architecture has been widely adopted in the software industry in recent years. This adoption is motivated by the increase in software modularity, and mainly the desired performance of decomposed monolithic applications. Although it has become popular, the current literature lacks studies that demonstrate the impact of event-driven architecture on performance. Hence, developers and architects end up adopting event-driven architecture without empirical evidence, but considering “expert advice.” This study, therefore, reports an empirical study on the impact of the adoption of event-driven architecture on performance. For this, the performance of an application implemented with an event-based architecture was compared with the performance of the same application implemented using a monolithic architecture. The comparison was made using metrics such as CPU usage, memory, response time, throughput, and total packages sent and received. The results, supported by statistical tests, show that the monolithic architecture, compared to the event-driven architecture, consumes less computational resources and has better response times. Finally, this study reflects on the adoption of event-driven architecture, as well as points out challenges and implications that need to be considered by the scientific community in future research.

Review on Comparative Study of Flutter App and Android App

Abstract: In today lives, mobile application can be increasing vital. In flutter is a cross platform framework and other native mobile application is one particular operating system. The time of 2015 year mobile application development increasing day by day. and also, the time of COVID-19 mobile application was very popular there was huge competition between the mobile application that available in a market. flutter can support both iOS and Android platform application. The main focused on the paper highlight the comparison between flutter mobile application and other native android mobile application. In this we can comparison on flutter SDK and android SDK on the basic of compilation, memory consumption, CPU usage, app size, etc.

Cogni-Sec: A secure cognitive enabled distributed reinforcement learning model for medical cyber–physical system

The advent of the Internet of Things (IoT) has resulted in significant technical development in the healthcare sector, enabling the establishment of Medical Cyber–Physical Systems (MCPS). The increased number of MCPS generates a massive amount of privacy-sensitive data, hence it is important to enhance the security of devices and data transmission in MCPS. Earlier several research studies were undertaken in order to enhance security in healthcare, but none of them could adapt to changing behaviors of data attacks. Here the role of blockchain and Reinforcement Learning (RL) comes into play since it can adjust itself to the nature of changing attacks, thus preventing any kind of attacks. This work proposes a solution, named Cogni-Sec, which employs a decentralized cognitive blockchain and Reinforcement Learning architecture and addresses the security issue. Blockchain is incorporated in the approach for data storage and transmission to increase the degree of security in the MCPS modules. Hyperledger Fabric is applied as the blockchain base which shows transaction query results with nearly 10% increased throughput, 69% less memory consumption, and 15% lower CPU usage when compared to Ethereum. Further security risk at the block mining level within a blockchain network is reduced by introducing distributed Reinforcement Learning architecture in replacement for the miner nodes, which imitates the cognitive behavior of miners in a distributed environment. Different multi-agent learning systems have been evaluated for building the mining agent. Among these, the a3c agent in distributed learning setup yields the optimum cumulative reward with a median value of 54.5 and minimizes the maximum number of data threats.

Research on the manifestation of university network Civics teaching based on adaptive network

Abstract This paper proposes an adaptive learning model for the interrelationship between civic education and college students’ career planning. Secondly, a competency point tracking model is constructed based on a neural network algorithm to explore the relationship between student behavior and competency points and how a certain student behavior affects competency points. Then, the Civics teaching platform for the college network was designed for college Civics teaching needs. Finally, the platform performance was tested, and the impact of the platform application was analyzed. The study showed that when the number of online users reached 5000, the transaction success rate remained at 100%, the average CPU usage rate was 56%, and the response time did not exceed 10 s. The average grade of Civics students has gone up by more than 10 points.

Analysis of effective integration of information platform courses and daily teaching in universities based on adaptive network

Abstract This paper first analyzes the effective application of web platforms and information technology in college teaching and constructs a web-based teaching aid platform for college courses based on the open Web. Secondly, the A3C algorithm is used to implement adaptive coding decisions to improve the performance of the platform by adaptive network coding. Finally, the performance test analysis is conducted for the web teaching aid platform built in this paper. The results show that the average response time of the system using the communication function with 2000 concurrent users is 1.807s, and the CPU and memory usage are 1146MHz and 1486MB, respectively, after running for 25s. This shows that adaptive network coding can effectively improve the performance of the supplementary teaching platform.

Analisis dan Implementasi Honeypot Honeyd Sebagai Low Interaction Terhadap Serangan Distributed Denial Of Service (DDOS) dan Malware

Every computer device connected to a wide computer network is vulnerable to security risks. These threats encompass vulnerabilities to data, information, resources, and services within the system. These threats include intrusion, eavesdropping, theft of vital data, as well as damage to the network system. These actions are carried out by parties who are not accountable, commonly referred to as intruders or attackers. One method to prevent or anticipate these malicious actions is by utilizing the honeyd Honeypot technique. The honeyd Honeypot adopts a low-interaction approach, which involves indirect interaction with attackers. This Honeypot serves as a decoy or simulated server intentionally presented as a target for attacks. The purpose of this Honeypot is to detect and analyze ongoing attacks. In this research, the honeyd Honeypot is implemented as a simulated server resembling an authentic server. This server provides various services and opens several ports deliberately prepared as attack targets, such as Port 139, and Port 21.The results of this research unveil the existence of attacks. Signs of these attacks include a surge in network traffic, reaching up to 100 Megabits above the normal level. Another indicator is a sudden spike in CPU usage, reaching 100%. The activities of these attacks can be analyzed through the installed Wireshark application on the Honeypot server. Information obtained from this analysis encompasses details about the attacker's activities, enabling more effective preventive, anticipatory, and corrective measures. These steps encompass securing the server, network system, and existing services.

MQTT and blockchain sharding: An approach to user-controlled data access with improved security and efficiency

The rapid growth of the Internet of Things (IoT) has raised security concerns, including MQTT protocol-based applications that lack built-in security features and rely on resource-intensive Transport Layer Security (TLS) protocols. This paper presents an approach that utilizes blockchain technology to enhance the security of MQTT communication while maintaining efficiency. This approach involves using blockchain sharding, which enables higher scalability, improved performance, and reduced computational overhead compared to traditional blockchain approaches, making it well-suited for resource-constrained IoT environments. This approach leverages Ethereum blockchain's smart contract mechanism to ensure trust, accountability, and user privacy. Specifically, we introduce a shard-based consensus mechanism that enables improved security while minimizing computational overhead. We also provide a user-controlled and secured algorithm using Proof-of-Access implementation to decentralize user access control to data stored in the blockchain network. The proposed approach is analyzed for usability, including metrics such as bandwidth consumption, CPU usage, memory usage, delay, access time, storage time, and jitter, which are essential for IoT application requirements. The analysis demonstrated that the approach reduces resource consumption, and the proposed system outperforms TLS and existing blockchain approaches in these metrics, regardless of the choice of the MQTT broker. Additionally, thoroughly addressing future research directions, including issues and challenges, ensures careful consideration of potential advancements in this domain.

A Hierarchical Optimized Resource Utilization based Content Placement (HORCP) model for cloud Content Delivery Networks (CDNs)

Content Delivery Networks (CDNs) have grown in popularity as a result of the ongoing development of the Internet and its applications. The workload on streaming media service systems can be significantly decreased with the help of the cooperative edge-cloud computing architecture. In the traditional works, a different types of content placement and routing algorithms are developed for improving the content delivery of cloud systems with reduced delay and cost. But, the majority of existing algorithms facing complexities in terms of increased resource usage, ineffective delivery, and high system designing complexity. Therefore, the proposed work aims to develop a new framework, named as, Hierarchical Optimized Resource Utilization based Content Placement (HORCP) model for cloud CDNs. Here, the Chaotic Krill Herd Optimization (CKHO) method is used to optimize the resource usage for content placement. Then, a Hierarchical Probability Routing (HPR) model is employed to enable a dependable end-to-end data transmission with an optimized routing path. The performance of the proposed HORCP model is validated and compared by using several performance metrics. The obtained results are also compared with current state-of-the-art methodologies in order to show the superiority of the proposed HORCP model. By using the HORCP mechanism, the overall memory usage of the network is reduced to 80%, CPU usage is reduced to 20%, response is minimized to 2 s, and total congestion cost with respect to the network load level is reduced to 100.

Hardware-based security devices using a physical unclonable function created by the irregular grain boundaries found in perovskite calcium titanate

The IoT era necessitated the need for hardware-based security devices due to their strong resistance to hacking compared to software-based systems, which are limited by CPU usage and performance. Hardware-based systems are more unpredictable due to their unique and non-discernible external properties. This research focused on developing a security device based on lead-free metal oxide CaTiO3 perovskite material with irregular electrical characteristics, known as physical unclonable functions. These irregularities were created by increasing the number of grain boundaries on the device surface through high-temperature annealing, resulting in greater adsorbed oxygen. The outcome was a highly random and unique security key using irregular electrical characteristics with 49.53% uniformity and a 46.55% inter-Hamming distance.

Enhancing The Performance of Intrusion Detection Using CNN And Reduction Techniques

There have been several security solutions based on artificial intelligence (AI), such as intrusion detection systems (IDS),cyberattacks are increasing because big data is increasing by using the internet on all sides of life, therefore, unbalanced data poses a serious problem in intrusion detection systems. The proposed detection system that is based on deep learning Convolutional Neural Network(CNN )partitions data into training and testing., Creating the classifier model for the Principal Component Analysis (PCA )technique of reducing features, is required for the development of intelligent analytic tools that need data pre-treatments and deep learning algorithm-performance enhancement. The UNSW-NB15 data set is used According to experimental findings, We employed a number of evaluation tools to assess the proposed NIDCNN strategy relying on the UNSW-NB15 data set that takes 30% of it for testing and after processing this part of the data became used to evaluate the proposed system. Measures such as a classifier's F-Score, precision, and sensitivity (Recall) are evaluated. Classifier performs better than the other approaches at determining if the data stream is normal or malicious. which is used to assess deep learning's effectiveness, the suggested model results from a high level of accuracy. The experimental findings demonstrate the suggested system's ability to accelerate the intrusion detection process while reducing memory and CPU usage. Experimental results prove the theoretical considerations.Because the UNSW-NB15 data set contains a wide range of patterns that accurately represent contemporary real network traffic, New NIDS algorithms can therefore be assessed using it.

Evaluating NiFi and MQTT based serverless data pipelines in fog computing environments

Recent advancements in Internet of Things (IoT) technology have driven massive amounts of raw data. Due to the hardware limitations of the constraint IoT environments, we cannot fully perform all data analytical tasks in the local environment. In contrast, data transferring, storing, and processing in the cloud suffer from cloud cost, latency and security issues. In this context, we require a multi-layer data processing environment to balance all these issues. Data pipeline concepts allow data processing between source and destination on multiple fog devices. On the other hand, IoT applications are event-driven. So, building these apps in a serverless computing manner minimizes operational and billing costs. Moreover, this provides flexibility to deploy tasks anywhere, irrespective of the underlying operating system. In this paper, we investigated two serverless data pipeline approaches designed with Message Queuing Telemetry Transport (MQTT) and Apache NiFi. Our proposed approaches were tested with the image streaming data, where we performed object detection in the images. Our experimental results compared these two approaches regarding pipeline execution time, memory and CPU usage. The results conclude that the Apache NiFi-based serverless data pipeline consumes more CPU than MQTT-based serverless data pipeline, however outperforms it in terms of execution time and memory utilization.

A Comparative Study of Post-Quantum Cryptographic Algorithm Implementations for Secure and Efficient Energy Systems Monitoring

The Internet of Things (IoT) has assumed a pivotal role in the advancement of communication technology and in our daily lives. However, an IoT system such as a smart grid with poorly designed topology and weak security protocols might be vulnerable to cybercrimes. Exploits may arise from sensor data interception en route to the intended consumer within an IoT system. The increasing integration of electronic devices interconnected via the internet has galvanized the acceptance of this technology. Nonetheless, as the number of users of this technology surges, there must be an aligned concern to ensure that security measures are diligently enforced within IoT communication systems, such as in smart homes, smart cities, smart factories, smart hospitals, and smart grids. This research addresses security lacunae in the topology and configuration of IoT energy monitoring systems using post-quantum cryptographic techniques. We propose tailored implementations of the Rivest–Shamir–Adleman (RSA), N-th degree Truncated Polynomial Ring Units (NTRU), and a suite of cryptographic primitives based on Module Learning With Rounding (Saber) as post-quantum cryptographic candidate algorithms for IoT devices. These aim to secure publisher–subscriber end-to-end communication in energy system monitoring. Additionally, we offer a comparative analysis of these tailored implementations on low-resource devices, such as the Raspberry Pi, during data transmission using the Message Queuing Telemetry Transport (MQTT) protocol. Results indicate that the customized implementation of NTRU outperforms both SABER and RSA in terms of CPU and memory usage, while Light SABER emerges as the front-runner when considering encryption and decryption delays.

ILEDGER: A Lightweight Blockchain Framework with New Consensus Method for IoT Applications

Considering the growth of the IoT network, there is a demand for a decentralized solution. Incorporating the blockchain technology will eliminate the challenges faced in centralized solutions, such as i) high infrastructure, ii) maintenance cost, iii) lack of transparency, iv) privacy, and v) data tampering. Blockchain-based IoT network allows businesses to access and share the IoT data within their organization without a central authority. Data in the blockchain are stored as blocks, which should be validated and added to the chain, for this consensus mechanism plays a significant role. However, existing methods are not designed for IoT applications and lack features like i) decentralization, ii) scalability, iii) throughput, iv) faster convergence, and v) network overhead. Moreover, current blockchain frameworks failed to support resource-constrained IoT applications. In this paper, we proposed a new consensus method (WoG) and a lightweight blockchain framework (iLEDGER), mainly for resource-constrained IoT applications in a permissioned environment. The proposed work is tested in an application that tracks the assets using IoT devices (Raspberry Pi 4 and RFID). Furthermore, the proposed consensus method is analyzed against benign failures, and performance parameters such as CPU usage, memory usage, throughput, transaction execution time, and block generation time are compared with state-of-the-art methods.

Comprehensive Performance and Scalability Assessment of Front-End Frameworks: React, Angular, and Vue.js

This paper discusses performance and scalability opportunities of the three major modern front-end frameworks, namely React, Angular, and Vue.js. In this respect, the research tries to point out the strengths and limitations of the compared frameworks in handling large-scale and complex web applications by evaluating each against different load conditions or scalability scenarios. It does this by running a set of benchmark applications that test performance indicators such as load times, speed of rendering, memory usage, and CPU usage. Additionally, it investigates scalability through load testing-emulation of increased user interactions and complexity testing, observing how each of these frameworks handles increasing application complexity. Overall, the results obtained were good, especially those for React, where the use of its virtual DOM was exploited, though some optimization may be necessary when handling more complex state changes. Angular has great performance, though it can show increased load times and higher memory usage due to the richness of features. Vue.js has competitive performance, with lower memory usage and faster render times, balancing simplicity with scalability quite well. The study concludes by giving insight on how to select the correct framework based on an application's particular needs and thought on scalability for guidance on choosing the right tool for the development job at hand.