Improve Network Performance Research Articles

Automatic endoscopic navigation based on attention-based network for Nasotracheal Intubation

Nasotracheal intubation (NTI) is one of the most commonly performed procedures in anesthesia and is considered the gold standard for securing the airway of patients. Endoscope operation is critical to the success of NTI. However, this operation remains challenging as it requires the surgeon to classify anatomical landmarks and detect heading targets of the endoscope tip in a sequence of monocular images. To alleviate this problem, this study presents a learning-based navigation method that automatically classifies four different anatomical landmarks and detects the heading target of the endoscope tip from endoscopic images. First, an end-to-end multitask network is introduced that consists of a branch for anatomical landmark classification and another for heading target detection. In addition, a convolutional attention module is designed to improve network performance by combining spatial and channel attention. Second, an endoscopic dataset named intuNav is built for network training. The trained network calculates navigation information without image prior knowledge. Finally, extensive experiments on the built dataset and endoscopic videos demonstrate the high performance of our method, achieving classification (94%) and detection (79.4%) accuracies. The results also indicate that the proposed method is effective and efficient in navigation generation during NTI.

A score based link delay aware routing protocol to improve energy optimization in wireless sensor network

The Wireless Sensor Network (WSN) is the most appearing expertise that has potential applications broad ranges that include environment examining, smart spaces, medical systems, and robotic study. Energy efficiency is a consideration of vital design for WSN. The collision is occurred owing to the transmission of data from the sensor nodes and the traffic at the SINK node is high because of the excess data transmission by the sensor nodes. An essential division of the consumption of resources in a WSN is managed by the mechanism of Medium Access Control (MAC). An existing MAC protocols initiated for the utilization of WSNs single channel for data transmission. The deployment of wireless sensor nodes is concerned in the harsh environment in which the drastic change of conditions suffers from sudden changes in the status of a node and the quality of a link. With limited energy, the sensor nodes are supplied and a major concern is towards improving the lifetime of a network. To deal with these issues, a novel and simple routing mechanism Score based Link Delay Aware Routing (SBLDAR) is proposed, which utilizes the multi-channel MAC procedure comprises of Carrier Sense Multiple Access/ Collision Avoidance (CSMA/CA), to exchange the data, the activity of Time Division Multiple Access (TDMA) sequencing nodes and also Frequency Division Multiple Access FDMA) to allow collision-free exchange simultaneously. Also, in order to overcome this issue, an approach has been put out in this work that involves leveraging a variety of trust variables to identify malicious nodes and then using fuzzy based Modified Sun Flower Optimization (FMSFO) algorithm to elect an effective cluster head. The score based route selection increases the network lifetime and improves network performance. By comparing with the other protocols, the proposed SBLDAR combined with the technique of multi-MAC protocol has shown that it increases the network’s overall performance based on the simulation studies.

Quality of Service Management for Fast Data Transmission in Industrial Mobile Communications Using PSO-Based Computer Vision Technique

Mobile ad hoc networks (MANETs) have rapidly expanded in recent years, mostly as a result of the mobile devices’ obvious low cost, heterogeneity, and flexibility. When communication networks are down or inaccessible, sensors can quickly create a reliable network that can be used as a rescue data system. Industrial mobile communication has developed into a significant study area for both industry and academia in recent years. The need for data interchange between various smart devices with various latency flows is enormous. Nevertheless, there has not been as much research done in this area. The suggested work suggests a fuzzy-based improved PSO optimized in MANET to alleviate the drawbacks of the conventional routing approach. The suggested study offers a numerical modelling that can be used to carry out adaptive transmission optimization with a variety of programmable module structures and guarantee cost-effective route establishment with greater throughput, goodput, and lowest delay requirements. To find the best route, the proposed approach combines energy-optimized route construction with data-driven cluster head (CH) selection based on swarm intelligence. Particle swarm optimization- (PSO-) based clustering achieves improved delay, goodput, throughput, and path difference degree as compared to other conventional approaches, according to the extensive simulation results. The energy efficiency of a network that is decentralized is more important. The MANET device’s energy efficiency helps to extend battery life and improve network performance. This research demonstrates how the fuzzy-based improved PSO optimized in MANET helps to raise the network’s energy efficiency. As a result, network energy conservation improves network performance and battery life. Additionally, this method enhances the quality of service methodology. End-to-end delay, energy consumption, packet delivery ratio, and normalized routing overhead are measured and compared between the simulation and conventional routing protocols.

Analysis of Behavioral Characteristics of Jammers to Detect Malicious Nodes in Mobile ADHOC Networks

Wireless ADHOC Networks are used to establish a wireless connection between two computing devices without the need for a Wi-Fi access point or router. This network is decentralized and uses omnidirectional communication media, which makes it more vulnerable to certain types of attacks compared to wired networks. Jamming attacks, a subset of denial-of-service (DoS) attacks, involve malicious nodes that intentionally interfere with the network, blocking legitimate communication. To address this issue, the proposed method analyzes various characteristics of nodes, such as packets sent, received, and dropped, at each node. Using the packet delivery ratio and packet drop ratio, the method detects jamming nodes from normal nodes, improving network performance. The network is simulated in NS2 environment.

Deep Learning Based Melanoma Diagnosis Identification

Abstract Malignant melanoma is considered to be one of the deadliest types of skin cancer, and it is responsible for the death of a large number of people worldwide. However, distinguishing whether melanoma is benign or malignant has been a challenging task. Many Computer Aided Diagnosis and Detection Systems have been developed in the past for this task. This paper presents a deep learning framework based approach for melanoma diagnosis and recognition. In the proposed method, the original skin mirror image is first preprocessed and then passed to the VGG16 convolutional neural network for tumor property classification. VGG16 uses smaller convolutional kernels instead of a larger convolutional kernel to achieve a reduction in network parameters and thus improve network performance. The system is trained using segmented RGB images generated from ground truth images of the ISIC2016 dataset, and finally a softmax classifier is used for pixel-level classification of melanoma lesions. In this study, a new method to become a lesion classifier was designed to classify melanoma lesion regions into benign and malignant tumors based on the results of pixel-level classification, and experiments were conducted on two well-established public test datasets, ISIC2016 and ISIC2017, with a final accuracy of 96.1%. The results indicate that convolutional neural networks are suitable for melanoma diagnosis identification. This study is of great relevance for advanced cancer caused by malignant melanoma.

Self-gated rectified linear unit for performance improvement of deep neural networks

This technical paper proposes an activation function, self-gated rectified linear unit (SGReLU), to achieve high classification accuracy, low loss, and low computational time. Vanishing gradient problem, dying ReLU, noise vulnerability are also resolved in our proposed SGReLU function. SGReLU’s performance is evaluated on MNIST, Fashion-MNIST, and Imagenet datasets and compared with seven highly effective activation functions. We obtained that the proposed SGReLU outperformed other activation functions in most cases in VGG16, Inception v3, and ResNet50. In VGG16 and Inception v3, it achieved an accuracy of 90.87% and 95.01%, respectively, exceeding other functions with the second-fastest computing time in these networks.

TORAS: Trustworthy Load-Balanced Opportunistic Routing for Asynchronous Duty-Cycled WSNs

As opportunistic routing (OR) becomes more popular in low-duty wireless sensor networks (WSNs), there is a greater demand for reliable and consistent data transmission. Only efficient and secure network protocols can guarantee reliable data transmission. Existing cryptographic algorithms for WSNs are inadequate due to their excessive energy usage. Data transmission reliability, security, and consistency will be provided via trust-based OR, while utilizing less energy and limited resources. OR uses opportunistic criteria to obtain a list of potential candidates for each sender node. This study proposes a trustworthy load-balanced opportunistic routing for asynchronous duty-cycled WSNs that consist of three parts. First, four probability distributions with their average term in <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">OR Metric</i> are considered for balancing the load among the nodes: distance from the node to the sink, expected number of hops, node degree, and residual energy. Second, two probability distributions with their average term in <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Trust Metric</i> are considered for measuring the trust factor: direct trust and recommended trust. Finally, in order to guide more data through the prioritized nodes, the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Trusted OR Metric</i> is determined by multiplying the average probability distributions of the two metrics. When compared to benchmarks, the simulation results show that the proposed protocol improves network performance in terms of energy consumption, average end-to-end delay, average throughput, packet delivery ratio, network lifetime, and detection rate.

Adaptive Repetition Strategies in IEEE 802.11bd V2X Networks

Vehicle-to-everything (V2X) communications have very strict throughput and latency requirements, even in scenarios with high mobility. IEEE 802.11bd is being developed as a WiFi amendment to improve V2X performance, allowing up to three repetitions per packet, along with other features. Message repetitions increase time diversity and enable maximum ratio combining at the receiver, thus improving the probability of correct decoding. This paper investigates the IEEE 802.11bd packet repetition feature. First, we analyze how the increased channel load due to repetitions may in some cases result in a higher collision rate leading to lower network performance. Then, we propose two strategies for exploiting the IEEE 802.11bd repetition feature. The proposed strategies use the channel busy ratio to adapt the number of transmissions to the channel load and are validated through network-level simulations, which account for both the acquisition and decoding processes. Results show that the proposed strategies improve network performance under variable traffic conditions and enable fair access to the channel.

Optimizing Container CNI Metrics for Improved Service LevelIndicators and Objectives in Container Networking

Container networking is a critical component of cloud computing, providing the infrastructure for the deployment andscaling of containerized applications. However, ensuring high-performance and reliable container networking can bechallenging due to the complexity of container network architectures and associated metrics. This paper proposes a newapproach to optimize container CNI (Container Network Interface) metrics to improve service level indicators (SLIs)and objectives (SLOs) in container networking. The approach involves analyzing key CNI metrics, such as networklatency, bandwidth, and packet loss, and fine-tuning them to optimize network performance and availability. This studyconducted experiments on Kubernetes clusters to validate the approach using different CNI plugins and workloads. Theresults show that optimizing CNI metrics can significantly improve network performance, reduce network downtime,and enhance the overall service level of container networking. The proposed approach provides a practical and effectivesolution for managing container network performance and availability and has the potential to benefit a wide range ofcloud-based applications.

Optimal energy planning of multi-microgrids at stochastic nature of load demand and renewable energy resources using a modified Capuchin Search Algorithm

The concept of interconnected multi-microgrids (MMGs) is presented as a promising solution for the improvement in the operation, control, and economic performance of the distribution networks. The energy management of the MMGs is a strenuous and challenging task, especially with the integration of renewable energy resources (RERs) and variation in the loading due to the intermittency of these resources and the stochastic nature of the load demand. In this regard, the energy management of the MMGs is optimized with optimal inclusion of a hybrid system consisting of a photovoltaic (PV) and a wind turbine (WT)-based distributed generation (DGs) under uncertainties of the generated powers and the load variation. A modified Capuchin Search Algorithm (MCapSA) is presented and applied for the energy management of the MMGs. The MCapSA is based on enhancing the searching abilities of the standard Capuchin Search Algorithm (CapSA) using three improvement strategies including the quasi-oppositional-based learning (QOBL), the random movement-based Levy flight distribution, and the exploitation mechanism of the prairie dogs in the prairie dog optimization (PDO). The optimized function is a multi-objective function that comprises of the cost and the voltage deviation reduction along with stability enhancement. The effectiveness of the proposed technique is verified on standard benchmark functions and the obtained results. Then, the proposed method is used for energy management of IEEE 33-bus and 69-bus MMGs at uncertainties conation. The results depict that the energy management with inclusion of WTs and PVs using the proposed technique can reduce the cost and summation of the VD by 46.41% and 62.54%, and the VSI is enhanced by 15.1406% for the first MMG. Likewise, for the second MMG, the cost and summation of the VD are reduced by 44.19% and 39.70%, and the VSI is enhanced by 4.49%.

From 5G to beyond 5G: A Comprehensive Survey of Wireless Network Evolution, Challenges, and Promising Technologies

The histrionic growth of mobile subscribers, disruptive ecosystems such as IoT-based applications, and astounding channel capacity requirements to connect trillions of devices are massive challenges of the earlier mobile generations, 5G turned up the key solution. The prime objective of the 5G network is not only to maintain a 1000-fold capacity gain and 10 Giga Bits per second delivered to a single user, but it also assured quality-of-service, higher spectral efficiency, the ultra-reliable and improved battery lifetime of devices and massive machine-type communication (mMTC). The huge traffic load and high amount of resource consumption in 5G applications, augmented reality and virtual reality for magnificent virtual experience, and wireless body area networks will seriously affect the channel capacity of cellular cells and interrupt the admission and service of other users which makes compulsory new means of channel capacity and spectral efficiency enhancement techniques. In this research, we review several key emerging wireless technologies to increase channel capacity and spectral efficiency that will not only lead to improve network performance but also meets the ever-increasing user demands. We investigate various benefits and current research challenges of using these technologies. We analyze massive multi-input multi-output technology (mMIMO) an efficient technique and promising solution for the 5G and Beyond 5G (B5G) networks with several benefits and features. Moreover, this paper will be of vast help to the researchers who will involve advance investigation and also to the wireless network operator industry that is in the search for smooth development of state-of-the-art 5G and B5G networks.

A 3D Occlusion Facial Recognition Network Based on a Multi-Feature Combination Threshold

In this work, we propose a 3D occlusion facial recognition network based on a multi-feature combination threshold (MFCT-3DOFRNet). First, we design and extract the depth information of the 3D face point cloud, the elevation, and the azimuth angle of the normal vector as new 3D facially distinctive features, so as to improve the differentiation between 3D faces. Next, we propose a multi-feature combinatorial threshold that will be embedded at the input of the backbone network to implement the removal of occlusion features in each channel image. To enhance the feature extraction capability of the neural network for missing faces, we also introduce a missing face data generation method that enhances the training samples of the network. Finally, we use a Focal-ArcFace loss function to increase the inter-class decision boundaries and improve network performance during the training process. The experimental results show that the method has excellent recognition performance for unoccluded faces and also effectively improves the performance of 3D occlusion face recognition. The average Top-1 recognition rate of the proposed MFCT-3DOFRNet for the Bosphorus database is 99.52%, including 98.94% for occluded faces and 100% for unoccluded faces. For the UMB-DB dataset, the average Top-1 recognition rate is 95.08%, including 93.41% for occluded faces and 100% for unoccluded faces. These 3D face recognition experiments show that the proposed method essentially meets the requirements of high accuracy and good robustness.

Performance Improvement of CDMA Wireless Sensor Networks in Low SNR Channels Based on Raptor Codes

Low signal to noise ratio (SNR) channels require careful design of communication systems in order to operate reliably in such extreme channels. Special measures and techniques must be considered when designing communication systems for such environments to counteract the effect of high noise power. Channel coding is a powerful tool to aid against adverse channel conditions including low SNR. Coupling channel coding with a spreading technique can further strengthen the performance of communication systems. Hence, we studied the performance of a system that combines Raptor codes with direct sequence code division multiple access (DS-CDMA) in low SNR scenarios. The studied design can operate in a wide range of low SNR values. Simulation results show that the system can offer satisfactory BER performance at different code rates. Also, combining Raptor codes with DS-CDMA showed an improvement of approximately 6 dB when compared with other schemes using DS-CDMA in low SNR conditions.

Mobility operator service capacity sharing contract design to risk-pool against network disruptions

We propose a new mechanism to design risk-pooling contracts between operators to improve service resilience during disruptions. We formulate a novel two-stage stochastic multicommodity flow model to determine the cost savings of a coalition under different disruption scenarios and solve it using L-shaped method along with sample average approximation. Computational tests are conducted for network instances with up to 1024 scenarios. The proposed model is applied to a regional multi-operator network in the Randstad area of the Netherlands, for four operators, 40 origin-destination pairs, and over 1400 links where disruption data is available. Using the proposed method, we identify stable cost allocations that could yield a 66% improvement in overall network performance over not having any risk-pooling contract in place. We illustrate the sensitivity of the HTM operator's bargaining power to different network structures and disruption scenarios.

VariBan: A Variable Bandwidth channel allocation algorithm for IEEE 802.15.4e-based networks

Among several link-layer protocols, Time-Slotted Channel Hopping (TSCH), defined in the IEEE 802.15.4e standard, offer a good performance improving reliability and throughput for wireless sensor network using in the best way two parameters: time and frequency. Many works in the current literature present scheduling schemes for the protocol. In this scenario, our work is among a small group that aims to explore the improvements of the protocol by exploiting different techniques over the transport layer of the network protocol. Channels with wider bandwidth can improve network performance by increasing the throughput and improving packet delivery. Therefore, this paper proposes a scheduling algorithm to dynamically allocate channels with variable bandwidth to improve the overall throughput and packet delivery. The scheduling and channel allocation happen using the conflict mapping method to identify links in congestion and schedule them. This work proposes a procedure for allocating channels with bandwidth proportional to the demand for packets per link. We implemented the algorithm in the Contiki-NG platform and evaluated the result through simulations. Results show improved time slot allocation, packet delivery ratio, and throughput. Our experiments show that VariBan achieves a throughput of over 99% and improves Packet Delivery Ratio by over 50% in all three scenarios compared to the state-of-the-art algorithm.

Trust aware clustering based secure routing techniques in wireless sensor network

Wireless Sensor Networks are becoming increasingly popular in everyday life since they offer a variety of network structures for developing cutting-edge real-time applications. Wireless sensor devices have high energy consumption limitations, so it is necessary to handle excessive energy consumption by malicious nodes properly to improve network performance. Even though numerous studies have been conducted to increase the dependability of routing in WSNs, the existing routing strategies do not meet the required security constraints by using intelligent methods to protect the sensor nodes from malicious attack. To overcome this challenge a novel Trust Aware Clustering based Secure Routing Techniques (TAC-SRT) has been proposed to minimize the overall energy consumption, improved security to nodes and to maximize the network lifetime. The proposed method is carried out in three phases. In the first phase, the cluster head is selected by using K mean clustering. In the second phase, the trust value of each node is evaluated by using Mamdani fuzzy inference rule. In the third phase, the Tversky similarity index is used to find the normal or malicious node and establishes the shortest route. The Fully Homomorphic Elliptic Curve Cryptography technique is then used to perform secure data transmission. The effectiveness of the proposed strategy is examined using several parameters, such as the lifetime of the network, data confidentiality, active nodes, and energy consumption. The proposed technique improves the network lifetime by 23.01%, 17.4%, and 13.2% better than MOSFA, SecDL, and CAR-MOSOA respectively. Finally, the proposed method demonstrated superior performance in terms of delay, throughput, encryption time, network lifetime, and packet delivery ratio compared with existing techniques.

Unsupervised detection of contrast enhanced highlight landmarks

AbstractIn the field of landmark detection based on deep learning, most of the research utilise convolutional neural networks to represent landmarks, and rarely adopt Transformer to represent and encode landmarks. Meanwhile, many works focus on modifying the network structure to improve network performance, and there is little research on the distribution of landmarks. In this article,the authors propose an unsupervised model to extract landmarks of objects in images. First, Transformer structure is combined with the convolutional neural network structure to represent and encode the landmarks; next, positive and negative sample pairs between landmarks are constructed, so that the semantically consistent landmarks on the image are pulled closer in the feature space and the semantically inconsistent landmarks are pushed farther in the feature space; then the authors concentrate their attention on the most active points to distinguish the landmarks of an object from the background; finally, based on the new contrastive loss, the network reconstructs the image by the landmarks of the object that are continuously learnt during training. Experiments show that the proposed model achieves better performance than other unsupervised methods on the CelebA, Annotated Facial Landmarks in the Wild, 300W datasets.

A review of localization algorithms based on software defined networking approach in wireless sensor network

Wireless Sensor Network (WSN) localization is an important challenge, because A strong relationship between aggregation sense data and the location of data's origin. Software Defined Networks (SDN) technique is support localization algorithm and take into account all network's limitations and constraints. By making use of global network knowledge provided by SDN controller, the results of previous studies show that considerable improvement in network performance can be achieved. This paper explores the recently proposed localization algorithms and discusses the simulation results for each method used in Software Defined Wireless Sensor Networks (SDWSN) to find the best way to localized nodes with the highest accuracy and lowest energy consumption. This paper also present Software defined networking paradigm and WSNS challenges which solved by SDWSNs.

A trio of resiliency, reliability, and uncertainty to design and plan the downstream oil supply chain

The main goal of this research is to design and plan a downstream oil supply chain network under mixed uncertainty. To reach this goal, a bi-objective mixed-integer linear model is proposed to make decisions at tactical and strategic levels. Notably, to deal with mixed uncertainty, a hybrid uncertain programming approach including adjustable possibilistic programming, chance-constrained programming, scenario-based programming, and p-robust optimization is developed and applied. Finally, the proposed model is implemented in a real-life case from the National Iranian Oil Products Distribution Company to illustrate the applicability and efficacy of the proposed approach. The analysis of the results indicates that the benefit of the used methods leads to the improvement of network performance against operational and disruption risks. Moreover, the analysis of the results indicates that the tailored approaches improve the total costs, the resiliency of the network, and the reliability of the chain in comparison to the deterministic situation.

Federated Reinforcement Learning-Based Resource Allocation for D2D-Aided Digital Twin Edge Networks in 6G Industrial IoT

The 6G is conceived to address the expected high level of requirements (such as ultra-high data transmission rate, support for the highest moving speed and seamless connection, etc.) in the next decade and beyond. In the context of 6G, a large number of Industrial Internet of Things (IIoT) devices may access the network, and thanks to the rapid development of artificial intelligence make smart manufacturing has the opportunity to be realized. However, a large number of IoT devices, the tremendous volume of data, the heterogeneous nature of devices, and the increasing concerns of privacy challenge the efficient management and quality of services in IIoT. To address these problems, in this article, a device-to-device (D2D) communication-aided digital twin edge network is proposed, where edge computing is introduced to bring computing and storage resources near to the end devices, and digital twin is utilized to fill the gap between physical and virtual space and D2D communication is applied to assist resource limited IoT devices to achieve normal communication. Moreover, digital twin-empowered federated reinforcement learning is leveraged to provide privacy awareness and decentralized resource allocation strategy training on D2D communication links to further improve network performance. The simulation results show that the proposal achieves significant network performance compared with baseline algorithms.