Recent research in electronic design automation (EDA) tools has focused on utilizing artificial intelligence (AI) for sizing analog circuit designs. Still, there has been a lack of focus on optimizing complex analog circuits. To optimize complex analog circuits within a few circuit simulations, we propose a partition-and-conquest-based optimizer (PC-Opt). PC-Opt assigns distinct actor-critic roles within a multi-agent system, facilitating the partitioning of complex analog circuits and conquering their optimization challenges. Partial differential training is developed for the proper prediction of each actor, which merges each other and then predicts the optimized entire circuit. To generate a compact and non-biased dataset for network training, a concentrated sampling method is devised. Experimental results on three circuits demonstrate the effectiveness of PC-Opt.

Vision Transformer ($V$iT) has gained prominence for its performance in various vision tasks but comes with considerable computational and memory demands, posing a challenge when deploying it on resource-constrained edge devices. To address this limitation, various token pruning methods have been proposed to reduce the computation. However, the majority of token pruning techniques do not account for practical use in actual embedded devices, which demand a significant reduction in computational load. In this paper, we introduce ViT-ToGo, a$V$iT accelerator with grouped token pruning. This enables the parallel execution of the$V$iT models and the token pruning process. We implement grouped token pruning with a head-wise importance estimator which simplifies the process need for token pruning, including sorting and reordering. Our proposed method achieves up to 66 % reduction in the number of tokens, resulting in up to 36% reduction in GFLOPs, with only a minimal accuracy drop of around 1 %. Furthermore, the hardware implementation incurs a marginal resource overhead of 1.13% in average.

With significant advancements in various design methodologies, modern integrated circuits have experienced noteworthy improvements in power, performance, and area. Among various methodologies, design space optimization (DSO), which automatically explores electronic design automation (EDA) tool parameters for a given design, has been extensively studied in recent years. In this study, we propose an approach to fine-tuning an effective FPGA design space to suit a specific design. By utilizing our ML-based prediction and explainable artificial intelligence (XAI) approach, we quantify parameter contribution scores, which reveal the correlation between each parameter and the final timing results. Using the valuable insights from the parameter contribution scores, we can refine the design space only with effective parameters for subsequent timing optimization. During the optimization, our framework improved the maximum operating frequency by 26% on average in six test designs. To accomplish this, our framework required even 47% fewer FPGA compilations than the baseline, demonstrating its superior capacity for achieving fast convergence.

In recent years, post-place optimization has emerged as a critical stage in physical design, aiming to improve power, performance, and area (PPA). Among various optimization techniques, buffer insertion, gate sizing, and Vth assignment have become leading optimization techniques for decades. However, these techniques are traditionally applied sequentially, leading to a critical suboptimality problem. Each technique obtains its own iterations performed step by step, preventing the best-optimal optimization for a target instance. To address this limitation, we propose a novel reinforcement learning (RL) based post-place optimization framework that performs various optimization techniques simultaneously. Moreover, to overcome the persistent headache of chip design, timing and power trade-off, we employ multiple agents that target each specific objective. By leveraging deep RL and graph neural network (GNN), our framework models optimal policies, dynamically selecting the most effective action given a target instance. Consequently, our framework obtained an improved Pareto-frontier set compared to comparison baselines while exhibiting 80% and 43% improvements in total negative slack and leakage power, respectively. The results demonstrate that our method outperforms weighted-sum-based co-optimization methods in optimizing timing and power.

There is a need for electronic design automation (EDA) tools for analog circuit design since analog circuit design requires substantial human effort and expertise. Using reinforcement learning (RL)-inspired methodologies, this study presents MA-Opt, an analog circuit optimizer. We propose MA-Opt to provide multiple predictions of optimized circuit designs through the use of multiple actors. Multiple actors can be exploited effectively by sharing a memory that affects the loss function of network training, resulting in an accelerated optimization of circuits. Furthermore, we introduce a cooperative near-sampling method deploying a synergistic effect and then optimizing the design. The efficiency of MA-Opt was demonstrated by simulating three analog circuits and comparing the results to other methods. In the experiment, the use of multiple actors with a shared elite solution set and the cooperative near-sampling method proved to be effective. MA-Opt achieved minimum target metrics up to 34$\%$better than DNN-Opt within the same number of simulations while satisfying all given constraints. Moreover, at identical runtime, MA-Opt exhibited better Figure of Merits (FoMs) in comparison to DNN-Opt.

The proliferation of various devices in our daily lives generates an immense volume of data, necessitating significant storage space and resources. Cloud storage systems are required to manage and meet the storage demands of large-scale data, such as big data. However, cloud storage systems come with security concerns, including the potential exposure of sensitive personal information during user access, threats of data tampering, and data breaches. This paper proposes a system that stores the hash value of cloud data on the blockchain for verification purposes to thwart the risk of data tampering, thereby ensuring data reliability. The proposed system is designed to grant access permissions through smart contracts to data uploaded to the cloud, allowing data owners to control access. It effectively mitigates the risks of keylogger attacks, data leakage, and breaches in cloud storage systems. In this paper, it was designed to be easily verifiable and reliable for any cloud storage and was applied to Dropbox, IPFS, etc., to measure the performance of each cloud storage and the verification performance of the proposed system. For KB-size uploads, IPFS performed up to about 10 times better than the comparable storage, and for downloads, Dropbox performed up to about 1.8 times better than the comparable storage. However, no matter which storage the proposed verification process is applied to, it showed consistent and fast verification performance with an average of 131 ms. In other words, this paper provides factors for selecting appropriate storage according to file size and user requirements and proposes that the verification system can be applied to various storage options, contributing to the development of a more secure storage system with low latency.

Bitcoin is a blockchain-based network where thousands of nodes are directly connected and communicate through a gossip-based flooding protocol. Mined blocks are propagated to all participating nodes in the network through compact block relay (CBR) protocol. Therefore, reducing the block relay time between nodes can reduce the block propagation time to all nodes and ultimately improve the performance of Bitcoin. In order to reduce the block relay time, the delivery time between nodes must be measured and analyzed to find the cause of the delay and provide ways to resolve it. Therefore, in this paper, we measure the CBR time between directly connected Bitcoin nodes and analyze the cause of the relay delay. Our results show that the delivery time delay is affected by whether or not a transaction is requested when assembling the compact block. In addition, the reason for requesting a transaction is due to the transaction propagation method and the characteristics of the transaction itself.

The Ethereum network uses Kademlia, a well-known P2P network, which allows the search for new nodes, and the change of connection with neighboring nodes. The Ethereum network must cope with security attacks such as DDoS attacks, 51% attacks, and Sybil attacks, and scalability issues, which slows down the transaction processing speed per second (TPS) as the network expands. A deep analysis of the dynamically changing topology and the connection between the nodes constituting the topology is needed to solve these problems. Therefore, in this paper, we measure the topology in the Ethereum network using a passive probing data collection to search for active nodes in the network and an active probing method to check the activity of nodes participating in the Ethereum network. Our results give a clear insight into the topology properties and topology visualization.

This paper proposes a method of using a blockchain to solve the privacy problem and forgery of black box image data, which plays an essential role in determining the responsibility for traffic accidents and preventing accidents. Blockchain that can operate inside vehicle black box IoT or connected car is used, and for this purpose, the size is reduced for operation in low-power, low-capacity devices. By enabling consensus, security problems can be solved through a lightweight blockchain that can operate inside a black box device. As a result of the experiment, it was confirmed that the IPFS upload and download delay time increased linearly, and the proposed consensus algorithm decreased 63% compared to PBFT.

Bitcoin is the most representative UTXO-based blockchain platform, and many studies have been conducted related to it. However, account-based blockchains such as Ethereum are not yet profoundly analyzed. There is an urgent need to track all cryptocurrency transactions involved with illegal activities to deanonymize and identify malicious users. To link users' accounts to real identities in both networks, we first need to examine the differences between Ethereum and Bitcoin to propose an efficient deanonymizing method. Therefore, this paper compares and analyzes the wallet address clustering method of Bitcoin and Ethereum.

Recently Distributed Denial-of-Service (DDoS) are becoming more and more sophisticated, which makes the existing defence systems not capable of tolerating by themselves against wide-ranging attacks. Thus, collaborative protection mitigation has become a needed alternative to extend defence mechanisms. However, the existing coordinated DDoS mitigation approaches either they require a complex configuration or are highly-priced. Blockchain technology offers a solution that reduces the complexity of signalling DDoS system, as well as a platform where many autonomous systems (Ass) can share hardware resources and defence capabilities for an effective DDoS defence. In this work, we also used a Deep learning DDoS detection system; we identify individual DDoS attack class and also define whether the incoming traffic is legitimate or attack. By classifying the attack traffic flow separately, our proposed mitigation technique could deny only the specific traffic causing the attack, instead of blocking all the traffic coming towards the victim(s).