Unstable Network Environment Research Articles

A Mobile AD Hoc Network Relay Method based on OLSR Protocol

Aiming at the reliability and performance problems of data transmission in the complex mobile AD hoc network environment, a relay method is designed and implemented in the unstable network environment with complex dynamics such as mobile, multi-hop and no center according to practical needs. By introducing a prior link-state routing protocol, OLSR is introduced. Using range value judgment, sliding window statistics and improved multi-point relay selection technology, it can effectively reduce the redundancy of relay node set, reduce the forwarding frequency of service data in the network, save the network overhead, and ensure the data transmission delay and quality in the large scale and dense mobile AD hoc network.

A DNN partitioning framework with controlled lossy mechanisms for edge-cloud collaborative intelligence

Collaborative intelligence (CI) is an approach used to accelerate deep neural network computations on IoT edge devices by offloading certain DNN computations to high-performance cloud servers. This offloading process involves transferring a large volume of feature data between the IoT edge device and the cloud servers, resulting in significant transfer time overhead that contributes to the overall end-to-end latency. Despite its importance, CI research has largely overlooked the impact of an unstable network environment on transfer time, leading to suboptimal latency in real-world scenarios. In light of this, we propose CO-PILOT, a novel partition framework that aims to minimize end-to-end latency while preserving the original model accuracy. CO-PILOT achieves this by finely controlling various lossy mechanisms, such as compression level and transport protocol, and strategically selecting error-robust layers as partition points to enhance resilience against network conditions and lossy mechanisms. The results demonstrate that CO-PILOT can reduce end-to-end latency by up to 85% with less than 1% accuracy degradation, all without requiring model retraining.

Online Classification of Network Traffic Based on Granular Computing

At Presently, it is still a great challenge to achieve online classification of traffic flows due to the highly varying network environments, e.g., unpredictable new traffic classes, network noise, and congestion. Traditional classification methods work well in stable network environments, but may not exhibit their performance in dynamic environments. To address online classification issues, a granular computing-based classification model (GCCM) is developed, where the spatial and temporal flow granules are defined to make GCCM robust against variations and less sensitive to noise, and the correlations among flow granules are explored to establish the granular relation matrix (GRM). The inherent burst features between packets indicated by GRM prompt GCCM to achieve fine classification in unstable network environments. GCCM analyzes the burst features of packets without inspecting the payload information, and thus can be used to classify encrypted traffic as well as unencrypted traffic at a fast speed. In addition, the GCCM model, depending on difference measurement <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$D(\cdot)$</tex-math> </inline-formula> , is a threshold-based classification, and therefore can be used to distinguish between time-varying classes. The validity of GCCM for online traffic classification is examined through theoretical results. The experimental evaluation of classification for fine and varied classes under dynamic network environments with noise and congestion also demonstrates its superiority in terms of classification accuracy and real-time performance with the state-of-the-art.

Peer-to-Peer Federated Learning for COVID-19 Detection Using Transformers

The simultaneous advances in deep learning and the Internet of Things (IoT) have benefited distributed deep learning paradigms. Federated learning is one of the most promising frameworks, where a server works with local learners to train a global model. The intrinsic heterogeneity of IoT devices, or non-independent and identically distributed (Non-I.I.D.) data, combined with the unstable communication network environment, causes a bottleneck that slows convergence and degrades learning efficiency. Additionally, the majority of weight averaging-based model aggregation approaches raise questions about learning fairness. In this paper, we propose a peer-to-peer federated learning (P2PFL) framework based on Vision Transformers (ViT) models to help solve some of the above issues and classify COVID-19 vs. normal cases on Chest-X-Ray (CXR) images. Particularly, clients jointly iterate and aggregate the models in order to build a robust model. The experimental results demonstrate that the proposed approach is capable of significantly improving the performance of the model with an Area Under Curve (AUC) of 0.92 and 0.99 for hospital-1 and hospital-2, respectively.

Performance and Security Analysis of Distributed Ledger Under the Internet of Things Environments With Network Instability

Applying blockchain technology to the Internet of Things (IoT) environment can significantly enhance the privacy protection of information and enable trusted sharing of data. However, the classic Proof-based blockchain systems have problems, such as unsatisfactory system throughput, low-consensus efficiency, and excessive computing resource consumption, making it difficult to directly apply them to the IoT environment. The recently proposed distributed ledger technology (DLT) represented by Tangle is designed based on the logical structure of the directed acyclic graph (DAG), which significantly improves the transaction confirmation delay and system throughput, making it more suitable for IoT scenarios. This article focuses on the performance and security of Tangle DLT in IoT applications with network instability. Such an unstable network environment can be caused by the predefined node working mode or signal interference, which leads to the status switching between online (work-mode) and offline (sleep-mode) of nodes or unstable data transmission. First, a Markov-chain-based model is established to analyze the time delay of block propagation in the underlying network under different status switching rates, different generation rates of blocks, different nodes’ average uplink bandwidths, and different number of neighbor nodes. Next, the impact of block propagation delay on the consensus efficiency and system throughput is analyzed. Furthermore, we also analyze the influence of block propagation delay on the attacker’s decision to conduct a double-spending attack under different network load, and deduce the suitable moment for the attacker to broadcast the parasite chain to maximize the probability of successful attack. Finally, through detailed numerical simulations, the experimental results comprehensively reveal the performance and security of the Tangle system in the unstable network environment. This article provides a valuable reference for the deployment and performance improvement of Tangle DLT in practical scenarios.

EC-GCN: A encrypted traffic classification framework based on multi-scale graph convolution networks

The sharp increase in encrypted traffic brings a huge challenge to traditional traffic classification methods. Combining deep learning with time series analysis techniques is a recent trend in solving this problem. Most of these approaches only capture the temporal correlation within a flow. The accuracy and robustness are unsatisfactory, especially in an unstable network environment with high packet loss and reordering. How to learn a representation with a strong generalization ability for each encrypted traffic flow remains a key challenge. Our detailed analysis indicates that there is a graph with particular local structures corresponding to each type of encrypted traffic flow. Inspired by this observation, we propose a novel deep learning framework called EC-GCN to classify encrypted traffic flows based on multi-scale graph convolutional neural networks. We first provide a novel lightweight layer that only relies on the metadata and encodes each encrypted traffic flow into graph representations. So that our framework can be independent of different encryption protocols. Then we design a novel graph pooling and structure learning layer to dynamically extract the multi-graph representations and improve the capabilities to adapt to complex network environments. EC-GCN is an end-to-end classification model that learns representative spatial–temporal traffic features hidden in a traffic time series and then classifies them in a unified framework. Our comprehensive experiments on three real-world datasets indicate that EC-GCN can achieve up to 5%–20% accuracy improvement and outperforms state-of-the-art methods.

Enforcing Multilevel Security Policies in Unstable Networks

Multilevel security (MLS) systems control access to data by formalizing permissible and impermissible information flows between data sources and destinations (e.g., database servers and clients) fixed with distinct security labels. In computer networks, MLS systems have been used to prevent unauthorized data disclosure in shared-infrastructure settings where network hosts and devices may fall within different trust domains (e.g., in multi-tenant cloud networks or wireless mesh networks). However, current MLS systems assume static network behavior—thus preventing the network from being practically usable in the presence of dynamic network events that frequent unstable network environments, including sudden changes in traffic patterns, link failures, and topology changes as a result of device movement or intermittent device connectivity. In this paper, we introduce <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MLS-Enforcer</monospace> , a software-defined networking (SDN) controller application that can efficiently deploy network-level MLS policies while retaining the ability to securely relabel network nodes under changing topology state and network traffic demands. We model network adaptivity as an integer linear programming problem that reflects a given security policy. We then introduce heuristic relabeling algorithms that achieve near-optimal performance and are more tractable and efficient for larger networks. We validate <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MLS-Enforcer</monospace> on several network topologies and traffic loads, demonstrating that it can relabel the network to route 90%+ of flows under normal conditions and quickly converge (on the order of seconds for the heuristic algorithms) under changing needs—from small network structure changes to catastrophic failures. This shows that formally secured networks can feasibly be deployed in diverse, changing, and unpredictable environments.

Robust Collaborative Filtering Recommendation With User-Item-Trust Records

The ever-increasing popularity of recommendation systems allows users to find appropriate services without excessive effort. However, due to the unstable and complex network environment, the historical behavior data of users are quite sparse in most cases. The inherent drawbacks render preference prediction infeasible for cold-start users and have become a crucial issue to be resolved in recommendation systems. To deal with the problems, we first present a Trust-based Collaborative Filtering (TbCF) algorithm to perform basic rating prediction in a manner consistent with the existing CF methods. Then, we propose the Hybrid Collaborative Filtering Recommendation approach with User-Item-Trust Records ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\text {UIT}_{\text {hybrid}}$ </tex-math></inline-formula> ), a novel approach that incorporates user trust into the existing CF-based methods in a harmonious way to supplement rating information. <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\text {UIT}_{\text {hybrid}}$ </tex-math></inline-formula> employs multiple perspectives to extract proper services and achieves a good tradeoff between the robustness, accuracy, and diversity of the recommendation. We conduct extensive real-world experiments on the Epinions data set to demonstrate the feasibility and efficiency of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\text {UIT}_{\text {hybrid}}$ </tex-math></inline-formula> .

A Communication-Efficient Federated Learning Scheme for IoT-Based Traffic Forecasting

Federated learning (FL) is widely adopted in traffic forecasting tasks involving large-scale IoT-enabled sensor data since its decentralization nature enables data providers’ privacy to be preserved. When employing <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">state-of-the-art</i> deep learning-based traffic predictors in FL systems, the existing FL frameworks confront overlarge communication overhead when transmitting these models’ parameter updates since the modeling depth and breadth renders them incorporating an enormous number of parameters. In this article, we propose a practical FL scheme, namely, Clustering-based hierarchical and Two-step-optimized FL (CTFed), to tackle this issue. The proposed scheme follows a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">divide et impera</i> strategy that clusters the clients into multiple groups based on the similarity between their local models’ parameters. We integrate the particle swarm optimization algorithm and devises a two-step approach for local model optimization. This scheme enables only one but representative local model update from each cluster to be uploaded to the central server, thus reduces the communication overhead of the model updates transmission in FL. CTFed is orthogonal to the gradient compression- or sparsification-based approaches so that they can orchestrate to optimize the communication overhead. Extensive case studies on three real-world data sets and three <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">state-of-the-art</i> models demonstrate the outstanding training efficiency, accurate prediction performance, and robustness to unstable network environments of the proposed scheme.

COVID-19로 인해 전환된 화상상담의 내담자 경험에 관한 합의적 질적연구

Objectives The purpose of this study is to explore the clients’ experience of a video conferencing counseling introduced in accordance with the prolonged COVID-19 outbreak. To identify clients’ experience of a video conferencing counseling and provide basic data for the effective approach of video conferencing counseling in the future. Methods For this purpose, individual interviews were conducted from February 2022 to March 2022 for 7 clients who switched from face-to-face counseling to video conferencing counseling due to COVID-19, and 7 pieces of data were collected. Collected data was analyzed using the Consensual Qualitative Research(CQR). Results As a result of the analysis, a total of 4 areas and 12 categories were derived. The four areas derived are ‘experience of transition to video conferencing counseling’, ‘digital environment of video conferencing counseling’, ‘video conferencing counseling space environment’, and ‘suggestions for activating video conferencing counseling in the future’. In the ‘experience of transition to video conferencing counseling’, there are three categories of ‘attitude before the transition to video conferencing counseling’, ‘good points compared to face-to-face counseling’, and ‘disadvantages compared to face-to-face counseling’. In the four categories of ‘comfort due to unfamiliar digital environment’, ‘inconvenience due to unstable network environment’, ‘consultation due to unfamiliar digital environment’, ‘visual counseling space environment’ In the two categories of ‘good points’ and ‘disadvantages of counseling in a private space’, ‘suggestions for activating video conferencing counseling in the future’ Three categories of ‘consultation progress’ were derived. Conclusions Based on the study results, the implications and limitations of counseling were presented. This study compared the experiences of face-to-face and video conferencing counseling through the client's video conferencing counseling conversion experience, and provided implications for an effective approach to video conferencing counseling and understanding the client.

FedPSO: Federated Learning Using Particle Swarm Optimization to Reduce Communication Costs.

Federated learning is a learning method that collects only learned models on a server to ensure data privacy. This method does not collect data on the server but instead proceeds with data directly from distributed clients. Because federated learning clients often have limited communication bandwidth, communication between servers and clients should be optimized to improve performance. Federated learning clients often use Wi-Fi and have to communicate in unstable network environments. However, as existing federated learning aggregation algorithms transmit and receive a large amount of weights, accuracy is significantly reduced in unstable network environments. In this study, we propose the algorithm using particle swarm optimization algorithm instead of FedAvg, which updates the global model by collecting weights of learned models that were mainly used in federated learning. The algorithm is named as federated particle swarm optimization (FedPSO), and we increase its robustness in unstable network environments by transmitting score values rather than large weights. Thus, we propose a FedPSO, a global model update algorithm with improved network communication performance, by changing the form of the data that clients transmit to servers. This study showed that applying FedPSO significantly reduced the amount of data used in network communication and improved the accuracy of the global model by an average of 9.47%. Moreover, it showed an improvement in loss of accuracy by approximately 4% in experiments on an unstable network.

Nationwide Emergency Survey Results on the Response to COVID-19 Pandemic and Its Effect on Pharmacological Education Conducted in the Department of Pharmacology in Universities and Colleges Jointly by the Physiology Society of Japan and the Japanese Pharmacological Society.

COVID-19 pandemic from spring, 2020 changed pharmacological education. To reveal its actual status, the Physiology Society of Japan and the Japanese Pharmacological Society jointly conducted nationwide emergency survey on the response to COVID-19 pandemic and the change of educational environment in universities and colleges from August 26 to September 15, 2020. Here, we demonstrated the results from the Department of Pharmacology in Medical, Dental and Veterinary School. Response rate was high around 90%. How to do the lecture and the practice were changed in 80% and 70%, respectively. Instead of face-to-face teaching, online methods were widely utilized. Live and on-demand streaming used around 30% each and mixed streaming around 40% in the lecture, while live and mixed streaming used around 25% each and on-demand streaming around 45% in the practice. Online methods were also adopted in attendance confirmation and conduct of examination. There are several problems to do online methods such as "unstable network environment", "lack of the reality for students" and "difficulty in the check of their understanding". On the other hand, there are unexpected benefits in online methods such as "anytime and repeatable learning", "an increase in questions from students" and "concentration of learning". 60% of responders insist they will utilize the changed methods in the future. In this part, we would discuss on "new normal" pharmacological education from this survey results.

A Tree Based Broadcast Scheme for (m, k)-firm Real-Time Stream in Wireless Sensor Networks.

Recently, various unicast routing protocols have been proposed to deliver measured data from the sensor node to the sink node within the predetermined deadline in wireless sensor networks. In parallel with their approaches, some applications demand the specific service, which is based on broadcast to all nodes within the deadline, the feasible real-time traffic model and improvements in energy efficiency. However, current protocols based on either flooding or one-to-one unicast cannot meet the above requirements entirely. Moreover, as far as the authors know, there is no study for the real-time broadcast protocol to support the application-specific traffic model in WSN yet. Based on the above analysis, in this paper, we propose a new (m, k)-firm-based Real-time Broadcast Protocol (FRBP) by constructing a broadcast tree to satisfy the (m, k)-firm, which is applicable to the real-time model in resource-constrained WSNs. The broadcast tree in FRBP is constructed by the distance-based priority scheme, whereas energy efficiency is improved by selecting as few as nodes on a tree possible. To overcome the unstable network environment, the recovery scheme invokes rapid partial tree reconstruction in order to designate another node as the parent on a tree according to the measured (m, k)-firm real-time condition and local states monitoring. Finally, simulation results are given to demonstrate the superiority of FRBP compared to the existing schemes in terms of average deadline missing ratio, average throughput and energy consumption.

Smart in-car camera system using mobile cloud computing framework for deep learning

Deep learning is becoming a popular technology in various applications, such as image recognition, gaming, information retrieval, for intelligent data processing. However, huge amount of data and complex computations prevent deep learning from being practical on mobile devices. In this paper, we designed a smart in-car camera system that utilizes mobile cloud computing framework for deep learning. The smart in-car camera can detect objects in recorded videos during driving, and can decide which part of videos needs to be stored in cloud platforms to save local storage space. The system puts the training process and model repository in cloud platforms, and the recognition process and data gathering in mobile devices. The mobile side is implemented in NVIDIA Jetson TK1, and the communication is carried out via Git protocol to ensure the success of data transmission in unstable network environments. Experimental results show that detection rate can achieve up to four frame-per-second with Faster R-CNN, and the system can work well even when the network connection is unstable. We also compared the performance of system with and without GPU, and found that GPU still plays a critical role in the recognition side for deep learning.

Joint study on VMs deployment, assignment and migration in geographically distributed data centers

Enterprises build private clouds to provide IT resources for geographically distributed subsidiaries or product divisions. Public cloud providers like Amazon lease their platforms to enterprise users, thus, enterprises can also rent a number of virtual machines (VMs) from their data centers in the service provider networks. Unfortunately, the network cannot always guarantee stable connectivity for their clients to access the VMs or low-latency transfer among data centers. Usually, both latency and bandwidth are in unstable network environment. Being affected by background traffics, the network status can be volatile. To reduce the latency uncertainty of client accesses, enterprises should consider the network status when they deploy data centers or rent virtual data centers from cloud providers. In this paper, we first develop a data center deployment and assignment scheme for an enterprise to meet its users' requirements under uncertain network status. To accommodate to the changes of the network status and users' demands, a VMs migration-based redeployment scheme is adopted. These two schemes work in a joint way, and lay out a framework to help enterprises make better use of private or public clouds.

Qos Constrained Communication using Voice Codecs over Multicast Ad Hoc Networks

QoS support for group communication over MANETs using real time applications is essential to fulfill the requirements of users. It is very challenging for a multicast ad hoc routing protocol to perform under the QoS constraints due to the unstable network environment, unpredictable traffic loads, mobility, scalability and limited resources etc. Researchers have developed some standards to support real time audio/video traffic over network such as Voice over IP (VoIP), H.323, H.264 etc. Multicast ad hoc routing protocols are designed to work with the traditional traffic streams, so there is need to analyze their performance using real time traffic in order to extend their capabilities. In this paper, we will explore the efficiency of PUMA multicast routing protocol using various voice codecs under the QoS constraints.

Playout Scheduling Method Based on Adaptive Jitter Estimation for Enhancing VoIP Speech Quality

지터라고 불리는 패킷 도착 지연의 편차는 모바일 단말기에서 Voice over Internet Protocol (VoIP)음성의 품질을 저하시키는 주요한 원인 중 하나이다. 이 문제를 해결하기 위해서 본 논문에서는 VoIP 음질향상을 위한 적응적 지터추정 기반의 플레이아웃 스케줄링 방법을 제안한다. 제안된 알고리즘은 예측된 네트워크 지연의 변화에 따라 각각의 패킷을 신장 혹은 압축하여 전송 지터의 영향에 대응한다. 이와 함께, 네트워크 지터 추정은 신속한 지연 스파이크 검출과 네트워크 상태 변화에 대한 반응을 포함한다. 실험결과는 제안된 알고리즘이 불안정한 네트워크 환경에서 고음질을 제공하는 것을 보여 주었다. Packet arrival-delay variation, so-called 'jitter' is one of the main factors that degrade the quality of voice in mobile devices at the Voice over Internet Protocol (VoIP). To resolve this issue, a playout scheduling based on adaptive jitter estimation for enhancing VoIP speech quality is proposed. The proposed algorithm copes with the effect of transmission jitter by expanding or compressing each packet according to the predicted network delay and variations. Additionally, the active network jitter estimation incorporates rapid detection of delay spikes and reacts to changes in network conditions. The experimental results have shown that the proposed algorithm delivers high voice quality in unstable network environment.

A Service Oriented Tele-health Promotion Information System with Mobile Application

Obesity has become an important health issue in Taiwan due to its increasing prevalence among population and the patients are getting younger. In this research, we build a health care information system, named the Pauian Health, for people who were on diet and to control their weight, or others to keep in shape. The Pauian Health is designed based on model–view–controller (MVC) pattern with a proposed architecture framework and has been successfully implemented in Pauian Archiland Company. The architecture centralizes the functions of accessing databases, user identification, and the validation of inherited clinical knowledge in a single component. Under this architecture, the system ensures its scalability, extensibility, data consistency, and the ability of cross-platform. This architecture has also been implemented in different devices and platforms, including iPhone, iPad and website. The architecture also supports unstable network environment, which uses Token as session identifier and user validation. The proposed framework has the potential in duplicating the service for different targeted users or deployed as a cloud service, which is fully capable in extending the service in the future.

A Robust P2P Information Sharing System and its Application to Communication Support in Natural Disasters

To provide the stable and continuous network services in cases of large-scale natural disasters, computers must use extremely limited network and computational resources effectively without imposing additional administrative burdens. The authors propose a P2P Information Sharing System for affected areas based on our proposed structured P2P network called the Well-distribution Algorithm for an Overlay Network (WAON). By applying the WAON framework, the system configures the P2P network autonomously using the remaining nodes, and achieves load balancing dynamically without additional network maintenance costs. Therefore, the system can perform well in an unstable network environment such as that during a disaster. The authors designed and implemented the system and evaluated its overall system behavior and performance in simulations assuming the real scenario of the Great East Japan Earthquake. Results show that the authors' system can distribute safety confirmation information of victims efficiently among the remaining nodes.

Network coding and competitive approach for gradient based routing in wireless sensor networks

Energy efficiency is a key design criterion for routing protocols in wireless sensor networks since sensor nodes are strongly constrained in terms of energy supply. Gradient-Based Routing (GBR) is a well known energy efficient routing protocol that is used in WSNs. However, there exist shortcomings in the GBR scheme such as: (1) sinks make use of flooding to broadcast interest messages which leads to a lot of duplication packets which are transmitted. This leads to the waste of a lot of energy in the network and (2) nodes deliver messages in a point to point manner. As a result, the potential of data retransmissions in the network is high due to the unstable network environment in WSNs. In this study, network coding and a competitive approach are proposed to solve the above two problems. Firstly, an energy efficient broadcast algorithm using network coding for GBR (GBR-NC) is proposed. This algorithm aims to reduce network traffic, and furthermore, reduce the energy consumption and prolong the lifetime of the network. Secondly, two competing algorithms (GBR-C and auto-adaptable GBR-C) are proposed for GBR. The basic idea of the proposed competing algorithms is to reduce the retransmission attempts and save the energy by considering two forward candidates. Simulation results show that the proposed schemes give better results when compared to the traditional GBR in terms of energy efficiency.