Search

Abstract The triage process in emergency departments (EDs) relies on the subjective assessment of medical practitioners, making it unreliable in certain aspects. There is a need for a more accurate and objective algorithm to determine the urgency of patients. This paper explores the application of advanced data-synthesis algorithms, machine learning (ML) algorithms, and ensemble models to predict patient mortality. Patients predicted to be at risk of mortality are in a highly critical condition, signifying an urgent need for immediate medical intervention. This paper aims to determine the most effective method for predicting mortality by enhancing the F1 score while maintaining high area under the receiver operating characteristic curve (AUC) score. This study used a dataset of 7325 patients who visited the Yonsei Severance Hospital’s ED, located in Seoul, South Korea. The patients were divided into two groups: patients who deceased in the ED and patients who didn’t. Various data-synthesis techniques, such as SMOTE, ADASYN, CTGAN, TVAE, CopulaGAN, and Gaussian Copula, were deployed to generate synthetic patient data. Twenty two ML models were then utilized, including tree-based algorithms like Decision tree, AdaBoost, LightGBM, CatBoost, XGBoost, NGBoost, TabNet, which are deep neural network algorithms, and statistical algorithms such as Support Vector Machine, Logistic Regression, Random Forest, k-nearest neighbors, and Gaussian Naive Bayes, as well as Ensemble Models which use the results from the ML models. Based on 21 patient information features used in the pandemic influenza triage algorithm (PITA), the models explained previously were applied to aim for the prediction of patient mortality. In evaluating ML algorithms using an imbalanced medical dataset, conventional metrics like accuracy scores or AUC can be misleading. This paper emphasizes the importance of using the F1 score as the primary performance measure, focusing on recall and specificity in detecting patient mortality. The highest-ranked model for predicting mortality utilized the Gaussian Copula data-synthesis technique and the CatBoost classifier, achieving an AUC of 0.9731 and an F1 score of 0.7059. These findings highlight the effectiveness of machine learning algorithms and data-synthesis techniques in improving the prediction performance of mortality in EDs.

Internet of Things (IoT) networks generate massive amounts of data while supporting various applications, where the security and protection of IoT data are very important. In particular, blockchain technology supporting IoT networks is considered as the most secure, expandable, and scalable database storage solution. However, existing blockchain systems have scalability problems due to low throughput and high resource consumption, and security problems due to malicious attacks. Several studies have proposed blockchain technologies that can improve the scalability or the security level, but there have been few studies that improve both at the same time. In addition, most existing studies do not consider malicious attack scenarios in the consensus process, which deteriorates the blockchain security level. In order to solve the scalability and security problems simultaneously, this paper proposes a Dueling Double Deep-Q-network with Prioritized experience replay (D3P) based secure trust-based delegated consensus blockchain (TDCB-D3P) scheme that optimizes the blockchain performance by applying deep reinforcement learning (DRL) technology. The TDCB-D3P scheme uses a trust system with a delegated consensus algorithm to ensure the security level and reduce computing costs. In addition, DRL is used to compute the optimum blockchain parameters under the dynamic network state and maximize the transactions per second (TPS) performance and security level. The simulation results show that the TDCB-D3P scheme can provide a superior TPS and resource consumption performance. Furthermore, in blockchain networks with malicious nodes, the simulation results show that the proposed scheme significantly improves the security level when compared to existing blockchain schemes by effectively reducing the influence of malicious nodes.

The incidence and mortality rates of lung cancer are high worldwide, where non-small cell lung cancer (NSCLC) accounts for more than 85% of lung cancer cases. Recent non-small cell lung cancer research has been focused on analyzing patient prognosis after surgery and identifying mechanisms in connection with clinical cohort and ribonucleic acid (RNA) sequencing data, including single-cell ribonucleic acid (scRNA) sequencing data. This paper investigates statistical techniques and artificial intelligence (AI) based non-small cell lung cancer transcriptome data analysis methods divided into target and analysis technology groups. The methodologies of transcriptome data were schematically categorized so researchers can easily match analysis methods according to their goals. The most widely known and frequently utilized transcriptome analysis goal is to find essential biomarkers and classify carcinomas and cluster NSCLC subtypes. Transcriptome analysis methods are divided into three major categories: Statistical analysis, machine learning, and deep learning. Specific models and ensemble techniques typically used in NSCLC analysis are summarized in this paper, with the intent to lay a foundation for advanced research by converging and linking the various analysis methods available.

Applications supporting businesses, smart systems, social networks, and advanced video applications such as eXtended Reality (XR) require large amounts of data processing to be provided in real-time. Therefore, the processing speed of big data systems is more important than ever. On the other hand, protecting a big data system is not easy, as various types of nodes and clusters are supported by various wired and wireless networks. Commonly security procedures slow down the response time of big data networks, and therefore, enhanced security and performance speed techniques need to be co-designed into the system. In this paper, a trusted streaming adaptive failure-compensation (TSAF) scheme is proposed that uses a trust management scheme to identify malicious nodes in Spark big data systems, exclude them from job/task processing, and calculate the number of nodes that can satisfy the process’s object completion time. The TSAF scheme shows an improved processing performance when there are attacks on the big data system compared to other existing real-time big data processing schemes. For the case of no security attack, the results show that the processing time of TSAF is faster by about 1 ~ 2% compared to the existing big data processing schemes when the process completion object time is set to 0.5 s. Even when the ratio of malicious nodes performing security attacks on worker nodes reaches 0.5, the results show that TSAF can satisfy over 75% of the tasks within the object time, which is significantly higher compared to the existing big data processing schemes.

Abstract In this letter, a modified message passing algorithm to prevent short cycles in the factor graph is proposed and combined with a relaxation technique to improve the performance of cyclic redundancy check aided belief propagation decoding for polar codes. Numerical results show that the proposed scheme can achieve a 0.7 dB coding gain, for the code rate of 0.5 at the block error rate of 10−5 with a comparable average number of iterations over the pure belief propagation decoding scheme when the code block length is 1024.

Because of the rapid increase of mobile traffic, flexible broadband supportive unmanned aerial vehicle (UAV)‐based 6G mobile networks using free space optical (FSO) links have been recently proposed. Considering the advancements made in UAVs, big data processing, and artificial intelligence precision control technologies, the formation of an additional wireless network based on UAV aerial platforms to assist the existing fixed base stations of the mobile radio access network is considered a highly viable option in the near future. In this paper, a combined time bound optimization scheme is proposed that can adaptively satisfy the control and communication time constraints as well as the processing time constraints in FSO‐based 6G UAV aerial networks. The proposed scheme controls the relation between the number of data flows, input data rate, number of worker nodes considering the time bounds, and the errors that occur during communication and data processing. The simulation results show that the proposed scheme is very effective in satisfying the time constraints for UAV control and radio access network services, even when errors in communication and data processing may occur.

Ultra Reliable Low Latency Communication (URLLC) mode is a promising 5G technology for various real-time applications, such as, autonomous vehicles, augmented reality, and factory automation. Several papers on URLLC exist, however, most papers only focus on ways to achieve high reliability, and lack mathematical analysis of delay related to packet errors, retransmissions, and bandwidth. In this paper, a M/G/1 queuing model is applied to analyze the erroneous transmission recovery delay of URLLC multi-user services, and using this model the minimum required bandwidth is derived and applied into an adaptive control scheme. The proposed Pollaczek-Khinchine (P-K) formula based quadratic optimization (PFQO) scheme optimizes the bandwidth requirement by controlling the maximum retransmission parameter of the hybrid automatic repeated request (HARQ) mechanism in URLLC. Simulation results show the bandwidth saving effect of the proposed PFQO scheme based on various signal to interference plus noise ratios (SINRs) and packet length distributions.

Approximately 50 million people have epilepsy worldwide. Prognosis may vary among patients depending on their seizure semiology, age of onset, seizure onset location, and features of electroencephalogram (EEG). Several researchers have focused on EEG patterns and demonstrated that EEG patterns of individuals with epilepsy can be used to predict prognosis and treatment responses. However, accurate EEG analysis requires an experienced epileptologist with several years of training, who are often unavailable in small or medium sized hospitals. In this paper, a novel machine learning (ML) model that accurately distinguishes Benign Epilepsy with Centrotemporal Spikes (BECTS) from Temporal Lobe Epilepsy (TLE) is proposed. BECTS and TLE show different seizure types and age of onset, but differential diagnosis can be challenging due to the similar location and patterns of the EEG spikes. The proposed hybrid machine learning (HML) model processes the diagnosis in the order of (1) creating feature matrices using statistical indexes after signal decomposition, (2) processing feature selection using Support Vector Machine (SVM) technology, and (3) classifying the results through ensemble learning based on decision trees. Simulation was performed using real patient data of 112 BECTS and 112 TLE EEG signals, where training was performed using 80% of the data and 20% of the data was used in the performance analysis comparison with the actual labeled data based on the diagnosis of medical doctors. The performance of the hybrid classification model is compared with other representative ML algorithms, which include logistic regression, KNN, SVM, and ensemble learning based decision tree. The model proposed in this paper shows an accuracy performance exceeding 99%, which is higher than the performance obtainable from the other ML classification models. The purpose of this study is to introduce a novel EEG diagnostic system that shows maximum efficiency to support clinical real-time diagnosis that can accurately distinguish epilepsy types. Future research will focus on expanding this ML model to categorize other types of epilepsies beyond BECTS and TLE and implement the HML diagnostic blockchain database into the hospital system.

Apache Spark and Hadoop are open source frameworks for big data processing, which have been adopted by many companies. In order to implement a reliable big data system that can satisfy processing target completion times, accurate resource provisioning and job execution time estimations are needed. In this paper, time estimation and resource minimization schemes for Spark and Hadoop systems are presented. The proposed models use the probability of failure in the estimations to more accurately formulate the characteristics of real big data operations. The experimental results show that the proposed Spark adaptive failure-compensation and Hadoop adaptive failure-compensation schemes improve the accuracy of resource provisions by considering failure events, which improves the scheduling success rate of big data processing tasks.

5G systems use multiple radio access technology (multi-RAT) to provide broadband, low latency, and high reliability services. Multi-RAT is effective when multipath transmission is used over multichannels. This is why multipath based transmission is one of the key technologies that can improve the throughput and reliability of real-time video streaming over 5G networks. This paper proposes two multipath based concurrent transfer techniques; fast concurrent transfer (FCT) and reliable concurrent transfer (RCT). FCT aims to minimize the round trip time (RTT) required for frame transmission by arranging each of the packets constituting a frame to be sent over multiple channels. In an environment where there are multiple channels, RCT performs redundant transmission of packets, so that even if a packet is lost in one path, the lost packet can be received safely over another path. FCT aims at securing the reliability of data transmission. In addition, a multipath based adaptive concurrent transmission (MACT) scheme is proposed which controls the ratio of FCT and RCT adaptively considering the state of the receiver buffer, inflow throughput, and decoding rate to enhance the reliability and throughput performance. The simulation results show that the proposed MACT scheme results in a better performance compared to the conventional transmission schemes in various network channel conditions.

Blockchains guarantee data integrity through consensus of distributed ledgers based on multiple validation nodes called miners. For this reason, any blockchain system can be critically disabled by a malicious attack from a majority of the nodes (e.g., 51% attack). These attacks are more likely to succeed as the number of nodes required for consensus is smaller. Recently, as blockchains are becoming too large (making them difficult to store, send, receive, and manage), sharding is being considered as a technology to help improve the transaction throughput and scalability of blockchains. Sharding distributes block validators to disjoint sets to process transactions in parallel. Therefore, the number of validators of each shard group is smaller, which makes shard-based blockchains more vulnerable to 51% attacks than blockchains that do not use sharding. To solve this problem, this paper proposes a trust-based shard distribution (TBSD) scheme that assigns potential malicious nodes in the network to different shards, preventing malicious nodes from gaining a dominating influence on the consensus of a single shard. TBSD uses a trust-based shard distribution scheme to prevent malicious miners from gathering in on one shard by integration of a trust management system and genetic algorithm (GA). First, the trust of all nodes is computed based on the previous consensus result. Then, a GA is used to compute the shard distribution set to prevent collusion of malicious miners. The performance evaluation shows that the proposed TBSD scheme results in a shard distribution with a higher level of fairness than existing schemes, which provides an improved level of protection against malicious attacks.

Abstract A trust management mechanism for mobile ad hoc networks (MANETs) is proposed to cope with security issues that MANETs face due to time constraints as well as resource constraints in bandwidth, computational power, battery life, and unique wireless characteristics. The trust-based reputation scheme GlobalTrust is a reliable trust management mechanism. In this paper, a clustering algorithm is applied to the GlobalTrust scheme (named Cluster-based GlobalTrust (CGTrust)) to find the optimal group size to minimize the configuration time, which consists of trust information computational time and complexity, while having to satisfy the trust reliability requirements. The optimal number of clusters is derived from the minimizing point of the computation complexity function. Simulation results show that the computational time and complexity of CGTrust are controllable and can be used effectively in time critical network operations that require trust analysis.

5th generation mobile networks, 5G, are the proposed next‐generation communication network standards. In addition to providing more than 1 Gb per second faster speeds, 5G will be a global game changer from technological, economic, societal, and environmental perspectives by integrating multiple networks in diverse sectors for various up to date applications such as the Internet of Things (IoT), device‐to‐device direct communication (D2D), vehicular communications (V2X), and disaster resilient communication. Currently, millimeter‐wave (mmWave) radio spectrum between 30 GHz and 300 GHz is critical for 5G rollout. Regulatory bodies around the world are now working towards opening up new spectrum bands from 6 GHz–100 GHz, and new technologies to overcome the challenges of these mmWave bands have been developed. It is envisioned that advanced 5G network infrastructure includes ultra‐broadband access, high‐speed backhaul and relay, softwarized flexible evolved packet core solutions for efficient system management, and satellite communications as an inherent component of 5G systems.

In this article, an emergency message reception power quantization–based time-slot broadcast scheme is proposed for vehicle-to-vehicle multihop communications. The power quantization–based time-slot broadcast scheme derives the optimal reception power quantization size such that time-slot assignments can be made to minimize the average time delay to support quick multihop emergency message broadcasting. The mathematical and simulation performance analysis demonstrates that the proposed power quantization–based time-slot broadcast protocol can reduce the average time delay when compared to the infrastructure-less framework, binary-partition-assisted broadcast, and the trinary-partitioned black-burst-based broadcast schemes.

A cognitive radio based hybrid data-type clustering (CR-HDC) algorithm is proposed to maximize network energy efficiency of cognitive radio (CR) sensor networks (CRSNs). By analyzing the overall energy consumption of CRSNs under various conditions, the optimal transmission range of a sensor node can be obtained for both when spectrum handoff (SHO) is applied and when it is not. Simulation results show that CR-HDC achieves performance enhancements in terms of network lifetime and the number of packets received at the base station (BS) compared to when applying the centralized low energy adaptive clustering hierarchy (LEACH-C) or hybrid data-type clustering (HDC) to CRSN environments.