• ETRI Journal
• /
• v.43 no.5
• /
• pp.775-786
• /
• 2021

To translate in real time, a simultaneous translation system should determine when to stop reading source tokens and generate target tokens corresponding to a partial source sentence read up to that point. However, conventional attention-based neural machine translation (NMT) models cannot produce translations with adequate latency in online scenarios because they wait until a source sentence is completed to compute alignment between the source and target tokens. To address this issue, we propose a reinforced learning (RL)-based attention mechanism, the reinforced attention mechanism, which allows a neural translation model to jointly train the stopping criterion and a partial translation model. The proposed attention mechanism comprises two modules, one to ensure translation quality and the other to address latency. Different from previous RL-based simultaneous translation systems, which learn the stopping criterion from a fixed NMT model, the modules can be trained jointly with a novel reward function. In our experiments, the proposed model has better translation quality and comparable latency compared to previous models.

• Structural Engineering and Mechanics
• /
• v.78 no.3
• /
• pp.369-378
• /
• 2021

This paper reviewed a few output-only system identification algorithms and identified the shortcomings of those popular blind source separation methods. To address the issues such as less sensors than the targeted modal modes (under-determinate problem), repeated natural frequencies as well as systems with complex mode shapes, this paper proposed a complex wavelet modified second order blind identification method (CWMSOBI) by transforming the time domain problem into time-frequency domain. The wavelet coefficients with different dominant frequencies can be used to address the under-determinate problem, while complex mode shapes are addressed by introducing the complex wavelet transformation. Numerical simulations with both high and low signal-to-noise ratios validate that CWMSOBI can overcome the above-mentioned issues while obtaining more accurate identified results than other blind identification methods.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.14 no.2
• /
• pp.71-78
• /
• 2019

As the capacity of SSDs rapidly increases, the amount of DRAM to keep a mapping table size in SSDs becomes very huge. To address a Demand-based FTL (DFTL) scheme that caches part of mapping entries in DRAM is considered to be a feasible alternative. However, owing to its unpredictable behaviors, DFTL fails to provide consistent I/O response times. In this paper, we a) analyze a root cause that results in fluctuation on read latency and b) propose a new demand-based FTL scheme that ensures guaranteed read response time with low write amplification. By preventing mapping evictions while serving reads, the proposed technique guarantees every host read requests to be done in 2 NAND read operations. Moreover, only with 25% of a cache ratio, the proposed scheme improves random write performance and random mixed performance by 1.65x and 1.15x, respectively, over the traditional DFTL.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.3
• /
• pp.861-876
• /
• 2022

Cryptoassets such as Bitcoin and Ethereum are widely traded around the world. Cryptocurrencies are also transferred between investors. Cryptocurrency has become a new and attractive means of remittance. Thus, blockchain-based smart contracts also attract attention when central banks design digital currencies. However, it has been discovered that a significant amount of cryptoassets on blockchain are lost or stranded for a variety of reasons, including the loss of the private key or the owner's death. To address this issue, we propose a method for recoverable transactions that would replace the traditional transaction by allowing cryptoassets to be sent to a backup account address after a deadline has passed. We provide the computational workload required for our method by analyzing the prototype. The method proposed in this paper can be considered as a good model for digital currency design, including central bank digital currency (CBDC).

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.2
• /
• pp.405-423
• /
• 2022

The distance vector-hop (DV-Hop) is one of the emblematic algorithms that use node connectivity for locating, which often accompanies by a large positioning error. To reduce positioning error, the bio-inspired algorithm and weight optimization model are introduced to address positioning. Most scholars argue that the weight value decreases as the hop counts increases. However, this point of view ignores the intrinsic relationship between the error and weight. To address this issue, this paper constructs the relationship model between error and hop counts based on actual communication characteristics of sensor nodes in wireless sensor network. Additionally, we prove that the error converges to 1/6CR when the hop count increase and tendency to infinity. Finally, this paper presents a modified error-oriented weight positioning model, and implements it with genetic algorithm. The experimental results demonstrate excellent robustness and error removal.

• Journal of applied mathematics & informatics
• /
• v.42 no.3
• /
• pp.583-605
• /
• 2024

This paper discusses the properties of attractors of Type 1 IFS which construct self similar fractals on product spaces. General results like continuity theorem and Collage theorem for Type 1 IFS are established. An algebraic equivalent condition for the open set condition is studied to characterize the points outside a feasible open set. Connectedness properties of Type 1 IFS are mainly discussed. Equivalence condition for connectedness, arc wise connectedness and locally connectedness of a Type 1 IFS is established. A relation connecting separation properties and topological properties of Type 1 IFS attractors is studied using a generalized address system in product spaces. A construction of 3D fractal images is proposed as an application of the Type 1 IFS theory.

• KIPS Transactions on Computer and Communication Systems
• /
• v.2 no.5
• /
• pp.217-222
• /
• 2013

Return-Oriented Programming(ROP) is an advanced code-reuse attack like a return-to-libc attack. ROP attacks combine gadgets in program code area and make functions like a Turing-complete language. Some of previous defense methods against ROP attacks show high performance overhead because of dynamic execution flow analysis and can defend against only certain types of ROP attacks. In this paper, we propose Indirect Branch Target Address Verification (IBTAV). IBTAV detects ROP attacks by checking if target addresses of indirect branches are valid. IBTAV can defends against almost all ROP attacks because it verifies a target address of every indirect branch instruction. Since IBTAV does not require dynamic execution flow analysis, the performance overhead of IBTAV is relatively low. Our evaluation of IBTAV on SPEC CPU 2006 shows less than 15% performance overhead.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.45 no.12
• /
• pp.83-88
• /
• 2008

The QPP interleaver has been gaining attention since it provides contention-free interleaving functionality for high speed parallel turbo decoders. In this paper we first show that the quadratic term $f_2x^2%K$ of $f(x)=(f_1x+f_2x^2)%K$, the address generating function, is periodic. We then introduce a low-power address generator which utilizes this periodic characteristic. This generator follows the conventional method to generate the interleaving addresses and also to save the quadratic term values during the first half of the first period. The saved values are then reused for generating further interleaved addresses, resulting in reduced number of logical operations. Power consumption is reduced by 27.38% in the design with fixed-K and 5.54% in the design with unfixed-K on average for various values of K, when compared with the traditional designs.

• International Journal of Control, Automation, and Systems
• /
• v.4 no.6
• /
• pp.782-787
• /
• 2006

The role of Unmanned Aircraft Vehicles(UAVs) has been increasing significantly in both military and civilian operations. Many complex systems, such as UAVs, are difficult to model accurately because they exhibit nonlinearity and show variations with time. Therefore, the control system must address the issues of uncertainty, nonlinearity, and complexity. Hence, identification of the mathematical model is an important process in controller design. In this paper, attitude dynamics identification of UAV is investigated. Using the flight data, nonlinear state space model for attitude dynamics of UAV is derived and verified. Real time simulation results show that the model dynamics match experimental data.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.11a
• /
• pp.973-975
• /
• 2005

With the continuous growth of e-commerce there is a huge amount of products information available online. Shop managers expect to apply information techniques to increase profit and perfect service. Hence many e-commerce systems use association rule mining to further refine their management. However previous association rule algorithms have two limitations. Firstly, they only use the number to weight item's essentiality and ignore essentiality of item profit. Secondly, they did not consider the relationship between number and profit of item when they do mining. We address a novel algorithm, profit-based association rule algorithm that uses profit-based technique to generate 1-itemsets and the multiple minimum supports mining technique to generate N-items large itemsets.