Advances on Broad-Band and Wireless Computing, Communication and Applications

The emergency disaster response headquarters (HQ) is responsible for incrementally collecting the disaster information from the whole area of a region. Our previous work studied a routing problem for patrolling vehicles to monitor such information and bring it to HQ multiple times on the way. The optimal routes for vehicles can be found by a systematic search to minimize the average delay time for the information collection. Further, we have also studied an area-segmentation approach in which each vehicle collects the information only in one of sub-areas. However, in disaster situations, some parts of the vehicles’ routes are likely to be damaged, incurring a considerable delay of information collection. In this paper, we investigate the resiliency of area-segmentations with different topologies, damaged locations, and delays incurred by damaged links. Our simulation results show that the area-segmentation approach is more robust than the whole-area approach in some cases by mitigating the risk of a large delay in information collection.

Random walks, including the Nomadic Lévy Walk, play a crucial role in various computer science domains such as networking, distributed systems, and optimization. The Nomadic Lévy Walk, a variant of the Homesick Lévy Walk, holds promise as a potential candidate for message ferry routing schemes. This approach combines the homing behavior characteristic of the Homesick Lévy Walk with the base relocation strategy employed in Lévy walks. In Delay-Tolerant Network (DTN) routing, encounter probability is a critical factor. In this paper, we present simulation results that highlight the impact of base relocation in the Nomadic Lévy Walk on encounter probability. Additionally, we conduct thorough comparisons of encounter probabilities using unit-disk graphs with varying diameters to further investigate the efficacy of the Nomadic Lévy Walk approach.

The performance of successive interference cancellation (SIC) employing Sequential Power Allocation (SePA) on the NOMA system downlink is discussed in this study. The simulation scenario is based on the variation of the SePA constant value and the number of users ranging from users who are far from the BS to users who are close to the BS. Bit error rate (BER) performance will be investigated using SePA constants and different number of users. Users who are farthest from the BS can directly decode their own signals, while users who are the second farthest to the closest to the BS must perform SIC first. In addition, the performance of the SePA algorithm is compared to Static Power Allocation (SPA) using QPSK modulation and Rayleigh fading channels. The simulation results show that the greater the SePA constant used, the greater the power transmit required to reach BER 10−3 will be even greater.

The FC (Fog Computing) model is discussed to implement the IoT (Internet of Things). Here, it is significant to protect fog nodes and devices from malicious accesses since the both are only supported with poor computation resources. The CBAC (Capability-Based Access Control) model is useful because each node can just check a capability token carried with an access request. Sensor data is obtained from a device and stored in another device by an authorized subjects who are granted tokens. However, even unauthorized subjects get the data from another device, i.e. illegal information flow. In the FC model, nodes receive sensor data from devices and send processed data to another nodes. In addition to devices, we have to prevent illegal information flow among nodes. Since sensor data for devices is processed and new data is generated by nodes, it is critical to discuss how much sensor data is included in data sent by nodes. We discuss a concept of component degree of each data which nodes exchange. By taking advantage of the concept, we newly propose an RR (Request Rejection) protocol for preventing illegal information flow among nodes. In the evaluation, we show the ratio of requests rejected to all the requests of nodes in the RR protocol.

The rapid development of technology is accompanied by a large number of demands and users. One of the solutions is to create reliable infrastructure that can support application performance. Kubernetes is a platform which offers one of the solutions by providing container application orchestration and enabling high availability and scalability through various automatic scaling mechanisms such as Horizontal Pod Autoscaler (HPA), which dynamically scales the number of pod resources without restarting the entire system. By default, Kubernetes only monitors built-in resource metrics such as CPU and memory usage for each host machine and pod. To use Custom Metrics, external software such as Prometheus can be employed to monitor metrics as needed. On the other hand, Kubernetes Event-driven Autoscaling (KEDA) software offers a solution to simplify and facilitate autoscaling by efficiently implementing event-based automatic scaling and achieving the scale-to-zero capability. In this paper, we investigate the performance of autoscaling through various experiments to understand the behavior of each custom metrics autoscaling solution.

Indonesia is a country that has diverse natural resources, cultures, and languages. One of the cultural diversities in Indonesia is Batik, which is an Indonesian cultural heritage consisting of cloth drawn by hand using traditional techniques. To assist the public in recognizing various batik motifs, a classification method was developed to identify the types of batik through input images. The classification method uses Convolutional Neural Network (CNN) based on MobileNet V1 and ResNet-152 V2 architecture. This research uses a dataset consisting of 3300 batik images from six different batik motifs, namely ceplok, parang, nitik, megamendung, kawung, and tambal. The optimal classification model was obtained using ResNet-152 V2 architecture with shear pre-processing method and RMSprop optimizer with test accuracy value of 89.67% and validation loss of 0.44.

In this paper, the authors introduce web viewers for educational VR contents. Especially, they propose new web viewer for educational VR contents of 3D scene models. Although there have been many types of VR goggles and a lot of entertainment VR contents, there have been a few educational VR contents. To enhance educational efficiency using VR technology, such educational VR contents should be created more and more. On the other hand, many types of scanning devices have been researched and developed, e.g., Lidar scanners, 360VR cameras and so on. Using these scanning devices, 3D data like Point Cloud Data (PCD), 3D model data and 360VR images/videos, can be obtained easily. To use these 3D data for educational VR contents, the authors have already developed viewer applications as web services. These web viewers have functionality to upload 3D data from users' web-browser to the web-server on that the web viewers work. However, 3D scene models are not supported. Standard videos are still important media for the on-demand online educations. Therefore, the authors propose new web viewer for 3D scene models supporting standard videos and VR goggles in this paper.

Some users (“Help recipients”) delegate necessary security actions to their family, friends, or others close to them. It is important to be able to take appropriate defensive actions against security threats by themselves when help is not available from neighbors (“Helpers”). In this paper, we interviewed 9 users who used to be Help recipients, but who have now started to take security actions by themselves. We investigated the reason why Help recipients delegated their security actions to Helpers and the human factors that have an impact when one takes security actions by oneself. As a result, Help recipients take their own security actions when they try new hobbies or feel a sense of ownership. Based on these findings, we classify Help recipients into four groups and propose an optimized system that shows security action lists according to user situation. These findings are useful when providing appropriate intervention for Help recipients.

It is important to provide personalized interventions that focus on the different security awareness of each user. We focus on updating operating system (OS) and seven major types of applications (Communication, Finance, Lifestyle, Games, Utility, Health, Entertainment) for smartphone users and aim to provide approaches that lead to updating of OS and seven types of applications (hereinafter, this is called “software”). We consider that user awareness of updating software may differ, and this relates to user understanding of the updating procedure. In this paper, we propose intervention methods according to users’ knowledge about update procedures. We conduct an online survey to evaluate effective approaches such as dialog message and type of incentive that increases the intention of smartphone users to update. We found that effective phrases of dialog messages differ according to the users’ understanding of the update procedures and that reward points are the best incentive for many users.

The number of insider threats and the expense of handling them increase every year, rendering it imperative to adopt measures against insider threats. In particular, insiders must adopt a psychological approach. Psychological approaches can be classified into three categories: deceiving insiders, demoralizing insiders, and luring insiders. The deception simultaneously accomplishes these three approaches. The disruption of enterprise systems by deception mechanisms causes a significant decrease in usability for users. The application of deception mechanisms requires careful consideration. Therefore, when designing deception as a countermeasure against insider threats, usability and security must be balanced from the viewpoint of cost-effectiveness. For evaluating cost-effectiveness of security measures, a method that models the relationship between “assets”, “threats”, and “countermeasures” as well as formulates the countermeasure selection problem as a discrete optimization problem has been proposed. However, these methods assume an external intruder, and to the best of our knowledge, no existing research explicitly covers the selection of countermeasures against insiders (malicious insiders). This paper proposes a scheme to quantitatively evaluate the effectiveness of deception against insider threats and determine the optimal deception mechanism. In the existing method, the relationship between “assets”, “threats”, and “countermeasures” is formulated as a discrete optimization problem, but the proposed method explicitly includes insiders as “threats” and deception as “countermeasures”. In addition, when evaluating the effectiveness of insider-threat countermeasures in the model, the usability of users must be considered. Therefore, by adding “operation” to “assets”, “threats”, and “countermeasures”, the proposed method incorporates the impact of selected countermeasures on the “usability” of business into the formulation of existing methods. Specifically, the existing method is sublimated into a security measure selection method that includes insider threat countermeasures (deceptions) by formulating the objective function of the existing method to be maximized with “usability” as a constraint.

In browser fingerprinting, device identification becomes increasingly difficult as the number of accesses increases because the number of combinations to be predicted increases exponentially. This paper proposes a method that employs a clustering algorithm to minimize the number of prediction pairs. The proposed method also evaluates whether two access logs were sourced from the same device and groups the results using the union-find algorithm for each device. The proposed method was applied to the access logs of an actual website, and experiments were conducted to investigate the impact of the pair reduction process. To investigate the impact of the reduction, the proposed method and the proposed method excluding the reduction were applied to a dataset of about 100,000 access logs. As a result, the processing time of the proposed method was reduced to about 1% compared to the proposed method without the reduction, and the adjusted Rand index improved by about 0.30. The experimental results demonstrate the feasibility of rapidly grouping devices, even when the target data contains unlearned variables or patterns.

In Wireless Sensor Network applications involving the monitoring of valuable objects, it is crucial to keep the locations of these objects private in order to protect them from potential adversaries. A number of schemes exist to address this problem, of which only a few of them focus on networks where multiple source nodes can report simultaneously. While those schemes provide a reasonable privacy level, they often suffer from inefficiencies in terms of packet overhead. In this paper, we present a scheme called Multi-Source Location Privacy (MSLP). In MSLP, a node uses two-hop neighbourhood information and broadcast nature of wireless communication to route packets to the sink. We perform simulation experiments of MSLP and compare it with other existing schemes and show that MSLP not only provides better location privacy but also offers better performance in terms of energy and delivery delay by reducing packet overhead.

Seafood traceability is essential in order to ensure the required level of quality control and management. However, over recent years different food scandals have damaged customer trust. Seafood is particularly affected by different types of frauds that take place at global scale. Often complex and unclear supply chains lead to misleading labels. These can falsely allow the selling of low value species as high valuable ones. Labels can also contain wrong location information. For instance, false location can be used in order to sell farmed fish as wild one. The problem of label misleading appears to be widely widespread in restaurants. A study over 23 different countries showed that restaurant mislabelling can be as high as 40%. While in general many traceability solutions have failed to meet the needs of food chain stakeholders, blockchain technology seems to be a promising solution. The decentralised and self-regulating blockchain nature can enable secure traceability in complex supply chains without the need of a centralised trusted party. Traceability data can be stored inside the blockchain which ensures high integrity, reliability and immutability. In this paper we describe a novel fish traceability system (CERTFish) that integrates in a novel way secure digital tags, blockchain technology, IoT tamper proof devices, location and time information. CERTFish ensures wild fish origin authentication and certifies the fish from the catch, throughout its conservation and transportation till the final customers at the restaurant. CERTFish has been validated in a real case study scenarios in order to certify the origin of a special type of anchovies from a specific Mediterranean area.

This paper introduces a novel system for recognition error detection in Vehicular Ad Hoc Networks (VANETs) using Fuzzy Logic (FL). The proposed system leverages a comprehensive set of input parameters, including Internal and External Distraction, Driver’s Inattention, and Inadequate Surveillance, to effectively evaluate and mitigate potential errors in vehicle recognition and response. In order to recognize the critical role of driver behavior and external factors in shaping road safety, we employ FL to model the intricate relationships and uncertainties inherent in such contexts. By incorporating linguistic variables and a rule-based inference mechanism, the system transforms the multidimensional input parameters into actionable insights regarding the likelihood of recognition errors. The distinctive contribution of this research lies in its holistic consideration of both driver-related and external variables, encompassing a wide spectrum of influences on recognition accuracy. Through simulatio validation, our proposed system demonstrates its efficacy in capturing subtle variations in driver attention and environmental conditions. Ultimately, the FL-based recognition error system holds significant promise in advancing the capabilities of VANETs, paving the way for more adaptive and responsive vehicular communication systems that prioritize safety in dynamic road environments.

Wireless Mesh Networks (WMNs) have become popular and they have many advantages. However, they have inherent wireless communication issues. To solve the problem, optimization of locations of mesh routers is a good approach. But optimizing the locations of mesh routers is an NP-hard problem. In this research work, we present an intelligent simulation system using Cuckoo Search (CS) algorithm called WMN-CS. We tune the hyperparameters and carry out computer simulations. The simulation results show that a good performance is achieved when the scale parameter ( $$\gamma $$ γ ) values range from 0.07 to 0.09 and the host bird recognition rate ( $$p_a$$ p a ) values from 0.900 to 0.925.

With the increasing complexity of the spatial topology of cellular communication networks and dynamic temporal characteristics of mobile network services, accurate prediction of cellular services is challenging. This article proposes the use of Dynamic Spatiotemporal Graph Convolutional Network (DSGCN) for mobile traffic prediction. This network modelling the dynamic characteristics of nodes in the cellular network traffic graph, and that capturing the dynamic spatiotemporal characteristics of border nodes by converting the cellular network traffic graph into hypergraph. In addition, for all nodes on different timestamps, complex spatiotemporal correlations are gathered through dynamic graph convolutional networks. By performing collaborative convolution on traffic flow maps and their hypergraphs, mobile traffic prediction is enhanced. Compared with conventional methods, the results of experiments show that this model has better predictive performance and higher training efficiency.

Named Data Networking (NDN) utilizes caching strategies to manage data caching on network nodes, making it suitable for high-traffic future networks. In-network caching brings desired data packets closer to users topologically, enhancing network performance and satisfying users. However, a critical challenge in NDN network design is determining the Content Store (CS) size allocation for each router. Previous studies overlooked router position, assuming uniform CS sizes. To address this, we proposed a caching strategy with the Social Network Analysis (SNA) algorithm was developed to calculate the influence of an NDN router on the topology. The SNA algorithm includes degree centrality, betweenness centrality, and consolidated centrality. Validated using the ICARUS caching simulator, the study’s results propose an efficient caching strategy for allocating cache storage on all NDN routers. The SNA-based approach outperforms alternatives in cache hit rate, latency, link load, and path stretch, contributing to effective and efficient network planning.

Reports have emerged of the vast amounts of data being transmitted across the Internet by Internet of Things (IoT) devices. Amidst this, the number of connected IoT devices continues to grow at a very rapid pace but over the last five years peak internet utilization has remained just under 50%. While applications have been processing increasingly massive amounts of data, users have been demanding faster or more acceptable response times and others have touted the economic value of having faster response times.In this paper we proffer arguments for adding more fungible paths to IoT and other smart devices to obtain theoretical speedup in network data transmission, and we present equations to compute the number of fungible paths needed to complete transmission of a set amount of data in a specific required minimum time frame, and the resulting speedup achieved. Our proposed model can assist IoT and smart device manufacturers, network administrators, and networked applications developers in deciding how many physical and logical layer one communication units to build into the devices they manufacture, and how many of these communication paths should be utilized during a communication session to achieve the desired data transfer/response times.

In recent years, online platforms have become more popular due to the spread of the Internet and its applications. The Internet provides a variety of human-created facts and data sources. But it consists of a vast array of disjointed structured and unstructured data that are difficult to collect by physical means and problematic to use in mechanical processes. Recently, modules and applications have been developed with various systems to collect data and transform it into organized information. However, those systems have complex menus and difficult for individual users to operate. Collecting multilingual hotels reviews data requires storing each language in a separate database to facilitate data analysis processes base on each natural language analytical method. Therefore, we consider that we need a simple tool with single button click system to collect multilingual reviews and store it in separate data base. In this research project We are developing a tool to collect hotels or attractions reviews data called MULARS (Multi-Languages Reviews Scrapers) for collecting large-scale of multi-language reviews. This paper is started by describing the web scraping process using different kinds of web scraping techniques. Then, the proposed system will be explained, and the results of data collection performances are presented. Finally, the paper discusses the advantages and future possibilities of the system applications on the different areas to exploit these data, such as development of business intelligence data, and development of new applications for data analysis.

Regression testing takes an important task in DevOps, agile software development, and software maintenance workflow. To ensure requirements quality and speed up product delivery and deployment, regression testing must overcome four challenges: identifying affected test cases correctly, reducing duplicate test cases, finding defects as early as possible, and testing with automated tools. Test case management is an important tool to assist regression testing to overcome the challenges. Applying high-quality test case management mechanisms (TCMM) is a critical way to speed regression testing tasks and assure quality. In this paper, the quality factors and characteristics of TCMM are discussed and analyzed. Quantifying the quality characteristics can identify and improve the defects and insufficiencies of TCMM. Based on quantifying characteristics, the TCMM measurement model is proposed to identify the defects and insufficiencies of TCMM, and then effectively improve the quality of TCMM. Using high-quality TCMM for regression testing makes DevOps, agile software development, and software maintenance workflow to speed up product deployment and delivery, and the required quality.

Internet Engineering Task Force (IETF) have developed many Internet Standards, described by RFCs (Request for Comments). As time goes by, some RFCs were obsoleted, some were updated, and some were developed for new concepts, functions, or options. Today, there are more than 9000 RFCs. It is not easy to figure out how many RFCs are related to a specific Internet standard, such as Domain Name Systems (DNS) or Border Gateway Protocol (BGP), and to understand the roles these RFCs play in the standard. In this study, given an Internet Standard name, we aim to generate an IETF standard map of RFCs. An experimental system is built. RFC maps of several Internet standards are generated and discussed.

Cloud computing platforms are clusters of high-end computers in Internet-based data centers that provide elastic software and hardware products at scale. In this paper, we propose a formula for predicting the number of resources needed for executing a job in a cloud render cluster. The users of the cloud computing platforms can use this formula to better balance production costs and deadlines, thus reducing costs and improving performance. According to the result of our experiments, our approach can accurately predict the system resource consumption. Thus, it can be used for setting parameters that maintain system performance.

Currently, the main applications of 5G networks are divided into three types, including: Enhanced Mobile Broadband (eMBB), Ultra-reliable and Low Latency Communications (uRLLC) and Massive Machine Type Communications, mMTC). This study modifies the base station SRA algorithm, by increasing priority and weight scheduling for packets. Among these packets, uRLLC’s are the highest, and of course, are transmitted first. The priorities of the remaining two types of packets are a little lower. We divide a base station’s resource blocks into multiple virtual networks according to time to practice network slicing with which to deliver packets, aiming to achieve effective network resource allocation.

In this paper, we present WMN-PSOHCDGA simulation system and implement two router replacement methods: RIWM and FC-RDVM for Wireless Mesh Networks (WMNs). We consider Stadium distribution of mesh clients and UNDX-m crossover method, and carry out a comparison study for small scale and middle scale WMNs. The simulation results indicate that for small scale WMNs, the SGC for RIWM and FC-RDVM is 100%, thus all mesh routers are connected. Considering NCMC, for RIWM two mesh clients are not covered, while for FC-RDVM are not covered four mesh clients. The load balancing of RIWM is better than FC-RDWM. Also for middle scale WMNs for both methods all mesh routers are connected. So, the SGC is maximized. The FC-RDVM can cover all mesh clients. However, for RIWM five clients are not covered. The load balancing for both methods is almost the same, but RIWM has slightly better load balancing than FC-RDVM.

Recently, DTN (Delay/Disruption Tolerant Networking) is getting much attentions as a communication architecture at a disaster affected area. Immediately after a disaster, countermeasure office must collect fresh disaster information for reducing disaster damages. This paper discusses how to improve freshness of disaster information. Then, we propose new method considering AoI (Age of Information). Mainly, our method employs two mechanisms: relay buffer management based on message originator, and changing order of message transmissions according to the message priority. By the results of simulation, we confirm that our proposal effectively shorten the AoI by comparison with the traditional DTN relay methods. In addition, this paper also discusses the situations that the network contains multiple senders having same priority. In order to cope with the situations, this paper also proposes to decide the transmission order according to a message creation time.

The authors focus on the message ferrying method, which is one of the route selection methods on DTN (Delay/Disruption-Tolerant Networking). In the message ferrying method, mobile nodes called ferry nodes transfer data among clusters in which there are many nodes including sink nodes. In the previous research of the message ferrying method, messages therefore are aggregated in nodes close to routes of ferry nodes. However, in the research, the method for data transferring to nodes in the clusters was not discussed. In this paper, the authors proposed a method that used records of the data transferred paths when data are aggregated to decide a route.

This paper will discuss polarization mismatch, the wave’s polarization transmitted by the transmitting antenna changes within the radio channel. When the radio waves reach the receiving antenna, the polarization does not match the antenna’s polarization. It will reduce the received power level and the system’s channel capacity. This study examined the effect of polarization loss on SISO, SIMO, and MISO channels. This aims to determine how much influence it has on system performance, especially receiving power and channel capacity. Based on the results of measurements and simulations, it was found that the mismatch between the polarization of the incident wave and the polarization of the receiving antenna causes a decrease in the system’s receiving power and channel capacity. When a polarization loss occurs, the received power decreases by approximately 25.14 dBm for linear polarization and 12.69 dBm for circular polarization, and it also reduces channel capacity.

Recently, the human-to-human and human-to-things connections are becoming significantly more complicated and less trustworthy in decision-making for diverse scenarios. As a result, the trust computing is getting interest from many study domains. The Logical Trust is one of the trust concepts. In this paper, we consider three parameters (Belief (Be), Experience (Ep), and Rationality (Ra)) for the implementation of a fuzzy-based system to evaluate the LT. We evaluate by simulation the proposed system. According to the simulation findings, the LT parameter increases when Be, Ep, and Ra are increasing. All LT values are greater than 0.5 when Ep values range from 0.5 to 0.9, Be is 0.9 and for any value of Ra. In this case, the persons or things are trustworthy.

The growth of electronic control units (ECUs) has led to modern automobiles being more convenient and technologically advanced. In order to communicate amongst these control units, the widely used Controller Area Network (CAN) communication protocol is used. CAN buses do not, however, come with built-in security features while being reliable and cost-effective. Because of this weakness, they are vulnerable to several attacks from prospective enemies. The installation of an Intrusion Detection System (IDS) is, nonetheless, a practical answer to this issue. A CAN bus system’s security can be considerably improved by the use of an IDS. We suggest a Long Short-Term Memory (LSTM) based intrusion detection system that is capable of effectively addressing the detection and security needs of CAN bus systems. We think that using Long Short-Term Memory (LSTM) models has a number of benefits, including highly accurate classification and effective anomaly detection. Our research findings have shown that our Long Short-Term Memory (LSTM) based IDS successfully identifies assaults on CAN systems with an accuracy rate of 99% and little loss. By providing this solution, we want to aid in the creation of reliable intrusion detection systems that can successfully protect CAN bus systems.

Wireless Visual Sensor Networks (WVSNs) have a mesh mechanism, which enables a wide area imaging and image collection is done through multi-hop communication between multiple sensor nodes equipped with visual sensors. Therefore, they can be used for monitoring infrastructure facilities and rivers. However, the placement of sensor nodes has a significant impact on the connectivity and transmission loss of wireless communication in WVSNs and the imaging range of visual sensors is limited. Thus, it is necessary to find the optimum sensor node arrangement and imaging direction of the visual sensors in order to cover all events within the monitoring area. In this paper, we propose an intelligent system for optimizing sensor node placement in WVSNs. The proposed system integrates two methods for sensor node placement by considering the number of events within the imaging range of the visual sensor and the imaging direction of the visual sensor. To evaluate the proposed method, computer simulations are performed for two scenarios with uniform and normal distribution of events. The simulatione results show that the proposed method maximizes the Size of Giant Compponent (SGC) and all sensor nodes are connected. From the visualization results, we found that in CCM-based SA, the nodes are spread over a wider area and all events are covered within the imaging range of the visual sensor.

This study innovatively designs a music therapy music selection recommendation system based on Case-based Spiral Diagnosis Reasoning with Consortium Blockchain Knowledge Base (CBSDR-CBKB) model. The purpose of this study is that hypertensive patients could effectively improve their physical and mental health by normalizing blood pressure through music therapy provided by CBSDR-CBKB system. In view of the lack of a music therapy recommendation system, this paper adopts the combination of case-based reasoning with spiral diagnosis procedure to recommend the music selection for the music therapist who provides to patients the therapist to listen to different types of music to make an integrated music selection recommendation application system. It not only makes it easy for hypertensive patients and their families to find music styles and selection principles suitable for music therapy, but also retains successful cases and failed cases and can screen cases more efficiently. Furthermore, this study is first design with a music therapy selection recommendation system. In CBSDR-CBKB. The steps of this study are to input the characteristic attributes of hypertensive blood pressure subjects, use the spiral theoretical model based on case reasoning to search for initial similar cases, and then perform periodic music therapy and recording after modification according to the case attributes. Next, the results output a music therapy record of the case according to the assessment of the Taiwan Hypertension Society and Taiwan Society of Cardiology, which includes a set of relative or modified music therapy strategies and a set of corresponding treatment effects, providing a music therapy recommendation. Each record will then be stored in Consortium Blockchain Platform. The system enables patients with hypertension caused by autonomic nervous system disorders to find suitable music styles to assist in the treatment or control of hypertension.

In this paper, we present the performance of multiple Unmanned Aerial Vehicles (UAVs) and Vehicle-to-Vehicle (V2V) communication in an urban environment. We employ the Epidemic protocol as the communication method to evaluate the dissemination of bundle messages by multiple UAVs and regular vehicles. We consider the urban grid and Tenjin area in Fukuoka City, where UAVs are assumed to move in a vertical direction randomly. From the simulation results, we found that the presence or absence of a UAV had a significant impact on the message delivery probability, especially when the number of vehicles was less than 90. Also, the use of UAVs improves message reachability and storage consumption on complex roads.

Recently, many kinds of robots are developed and they collaborate with various things to help humans. Over the past few decades, industrial robots have been used in factory operations. Also, these robots such as vacuum cleaner robot, security robot, therapy robot and wheelchair robot can support humans for many operations. Thus, using robots and related technologies are very meaningful to enrich our lives. In the context of welfare, in many countries in the world there are millions of people with disabilities who rely on wheelchairs. In this paper, we present an omnidirectional robot and introduce two applications: Moving Omnidirectional Access Point Robot (MOAP) and Wheelchair for tennis. These applications required a precise control of omnidirectional robot. Therefore, we measured the accuracy of moving omnidirectional robot using supersonic signals. The results shows that the omnidirectional robot was controlled correctly, because the error rate in x direction was within 1.00 [%]. However, when the robot moved in y direction, the error rate was 2.9 [%].

For this study, we specifically examine object detection methods for finding victims at a facility who are in a lying posture because of injury or for other reasons. Most large-scale human training data available on the internet are based on the standing postures of people. Lying postures are rare. Furthermore, obtaining large numbers of images of people lying down in disaster situations and creating training data are extremely difficult. Therefore, it is necessary to learn object detection models that handle diverse postures. For this study, we assess data augmentation. Because the image of a person in a reclined posture is regarded as one in which the head and body are aligned side by side, data extension of 90° rotation to the existing image of a person in a standing posture can be used to enhance the image of a person in a reclining posture. For this study, the object detection algorithm SSD is augmented with rotational data augmentation to enhance the detection of lying human figures. However, it is unknown whether the rotation data enhancement is effective for detecting lying persons or not. For an evaluation experiment, we changed the rotation rate arbitrarily and compared the models learned with each rate. The data expansion of the rotation was evaluated using the detection accuracy for normal and recumbent human figures. We confirmed that the detection accuracy of the lying person improved at a rotation of 25%.

In this study, we delve into the challenge of compressing and decompressing digital-ink data obtained from handwritten inputs on devices such as pen tablets, etc. We present a new approach using sparse Gaussian process regression in order to compress digital-ink data into some kernel matrix consisting of a sequence of pseudo-inputs. Moreover, to enhance the accuracy of compression and decompression, we introduce a dynamic programming (DP) approach for the optimal selection of dominant pseudo-inputs. The effectiveness of our method is substantiated through a series of experimental studies.