As the user count for video-related services continues to grow, ensuring high-quality service (QoS) for them will become even more crucial in the future. Many studies have been conducted to enhance the quality of on-demand video streaming using adaptive bitrate (ABR) algorithms and artificial intelligence (AI). This study addresses a more complex challenge than that of on-demand video streaming: enhancing service quality in multi-party, full-duplex communication scenarios, such as video conferences. We propose a deep reinforcement learning (DRL)-based video bitrate allocation framework for a media server in the video conferencing system. Our framework aims to increase overall QoS by applying an appropriate bitrate for each connection in a video conferencing call, considering the network conditions for users. We train the DRL model to maximize the aggregate QoS of users in a meeting by constructing a feedback loop between a media server and a DRL server. Our experimental results demonstrate that our framework can adaptively control the video bitrate according to changes in network conditions. As a result, it achieves higher video bitrates in the user application (approximately, 5% under stable network conditions and 35% over the highly dynamic network conditions) compared to the existing rule-based bandwidth allocation.

Many studies are applying reinforcement learning to real-world problems. However, this is a difficult problem, and its application in the real world requires solving many challenges. Therefore, in order to solve this problem well, it is necessary to understand the process of data collection in the real world and the data collected through this process. Video conferencing is also an example of the application of reinforcement learning in the real world, so it is necessary to understand the video conferencing system used and the data collected through it. To this end, this paper presents a process to collect network status information data from a video conferencing system and introduces data and model analysis methods for effective reinforcement learning environment settings and problem definition. In addition, among various problems in video conferencing, the video quality selection problem is set as the target problem, and we trained the model to perform the introduced feature importance analysis. We also included the performance evaluation results and correlation with data analysis results.

Non-Fungible Tokens (NFT) is a unique digital token based on blockchain technology, which is used as a means to prove ownership of assets in various areas. With the huge popularity of NFT, the launch of new NFTs attracted many people to mint NFTs. However, because Ethereum has low processing speed and then cannot accommodate the explosive demands, it causes Crypto Gas War, which increases the overall transaction fee and wastes unnecessary gas. In this work, we propose a Raffle-based N FT launch to solve t he C rypto Gas War. We demonstrated that our proposed NFT launch solves this problem but other existing solutions do not solve the problem. Moreover, our NFT smart contract is gas-efficient. Our proposed method reduces the gas usage by 15.5% or more compared to general method, and the efficiency is improved a ccording to increasing the total number of NFTs.

Many studies have applied machine learning to bitrate control to increase Quality of Experience (QoE) of video streaming services in highly dynamic networks. However, their solutions mainly focused on HTTP adaptive streaming with one-to-one connections. This paper studies video conferencing applications where multi-party, full-duplex communication happens among participants. In particular, we propose Muno, a Deep Reinforcement Learning (DRL)-based bandwidth prediction framework for multi-party video conferencing systems. Muno learns and predicts an appropriate bitrate for each connection in a multi-party conferencing call. We trained Muno to maximize the QoE of individual connections by constructing a feedback loop between a media server and DRL servers. Our experimental results show that Muno achieves a higher video streaming rate and lower delay compared to state-of-the-art rulebased algorithms.

Several years after NFTs first appeared, people began to use NFT to prove ownership of assets in diverse domains. Then the values of NFTs rapidly increased and accordingly the interests in NFTs have grown. Because Ethereum has the most active transactions and has a lot of users, many projects use Ethereum to deploy their NFT smart contracts. However, Ethereum has a chronic disadvantage of low scalability. During the NFT drop period for famous and popular NFTs, users are crowding the event, resulting in a large number of transactions to get NFTs. The low scalability leads to a fierce competition called the Crypto Gas War. In this work, we choose three famous NFT drop events to collect on-chain data of Ethereum over the drop period in order to analyze the impact of the Crypto Gas War on the Ethereum network in detail. In addition, we analyze the collected data to uncover critical and hidden problems, and present insights to solve them.

While many studies have been conducted to apply reinforcement learning (RL) to real world problems beyond games such as Atari, video conferencing is also one of real world applications. In video conferencing, reinforcement learning is used to control the bitrate to improve the user's quality of experience (QoE). However, real world problems such as video conferencing have different characteristics compared to electronic games. Usually the rewards in real world problems are not clear or abstract, and this makes it difficult to design the RL model and training process of the model to maximize the cumulative reward. Therefore, in this paper, we present the method for stabilizing the training of the models that apply reinforcement learning to video conferencing. In addition, we established a simulation environment that can train deep RL models in 1-to-1 video conferencing. An evaluation is performed to analyze the difference between the baseline model and the models generated using the stabilization method in the simulation environment.

With the advancement of artificial intelligence(AI) technology, AI is being used in various industries such as factory automation and autonomous driving. Video conferencing systems have also added functions that use AI to overcome the limitations of existing algorithms, for example, super resolution and virtual background functions using image segmentation. However, web-based video conferencing limits the application of these features due to a limited web browser environment. In this paper, we introduce several approaches to apply deep learning in a web browser environment to provide the features that use deep learning models, and introduce image segmentation models used for virtual background functions in each method and evaluate their performance. Finally, we discuss areas that need to be considered to apply deep learning models to web-based video conferencing.

Supply chain is an entire network of producing and delivering a specific product to a final consumer. Stainless steel is a specific product being delivered on a supply chain. It is not easy to manage and monitor the entire supply chain because a supply chain has high complexity, including various organizations and activities. Due to this difficulty, several issues occur in the process of supplying stainless steel, such as forgery and alteration. Blockchain is a decentralized and distributed ledger technology that specializes in transparency and immutability. Many companies try to introduce this blockchain technology into supply chain management to conveniently monitor their supply chain. Accordingly, the blockchain technology can make steel companies be able to investigate and protect high-quality products from counterfeited low quality products. This paper proposes a design of a blockchain-based stainless steel tracking system to thoroughly track the entire process involved in supplies from stainless steel mills to the final customers. This proposed design is based on the hyperledger fabric which is one of the most popular private blockchain platforms.

Network Function Virtualization (NFV) is one of the important aspects of modern network architecture. NFV decouples Network Functions (NFs) from hardware, therefore produces Virtual Network Functions (VNFs) that can run on standard, commodity servers, which in turn mostly run Linux kernel. In this paper, we propose a general architecture for building hybrid kernel-user space VNFs which leverages extended Berkeley Packet Filter (eBPF). eBPF is a framework in Linux kernel that enables network programmability inside kernel for optimal performance. However, the programmability of eBPF is limited due to safety and security of the kernel. Our proposed architecture applies hybrid approach: leave the simple work inside the kernel with eBPF and let complex work be processed in the user space. This architecture allows building complex VNFs to have both speed and flexibility. To demonstrate, we use the proposed architecture to build two VNFs: Dynamic Load Balancer and Deep Packet Inspection with Dynamic Sniffing. The evaluation results show that both VNFs significantly outperform the widely used solutions.

Software-Defined Networking (SDN) is one of the emerging network technologies that aims to operate and manage networks in more flexible and efficient manner. Among various features from SDN, Network Virtualization (NV) is one of the most promising network technologies that provides the ability to provision multiple virtual networks on top of underlying physical networks in a way to improve network utilization and provide flexibility. Since virtual networks are dependent on underlying physical network, failures in the physical network in turn can affect virtual networks. To provide a transparent failover solution for virtual networks, a Network Hypervisor (NH)-based approach gains more and more popularity, as this approach does not expose too much details of the physical network to tenant controllers. NH-based approach can be further categorized into two folds - restoration and protection. Restoration has a drawback that failover time increases proportionally with the number of switches along the path, while protection has a weakness that it cannot deal with dynamic changes of network states. To address the drawbacks of these approaches, we propose a dynamic failover method by properly combining the two approaches that aims to preserve the dynamicity while reducing the failover time. To show the feasibility, we have designed and implemented the proposed failover method by enhancing existing open source network hypervisor, OpenVirteX (OVX), and evaluated its performance in an emulated network environment. Our outcomes show the improvement of performance on delay compared to the exiting method.