Chulsoon Hwang

Chulsoon Hwang

Auto-generated profile, based on publicly available information

Chulsoon Hwang

Electromagnetic Compatibility Laboratory, Missouri University of Science and Technology, Rolla, MO, USA

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 September 2022 in IEEE Transactions on Microwave Theory and Techniques
Verified
Journal Article
Closed access
HPHyunwook Park
Hyunwook Park
MKMinsu Kim
Minsu Kim
SKSeongguk Kim
Seongguk Kim
KKKeunwoo Kim
Keunwoo Kim
HKHaeyeon Kim
Haeyeon Kim
JKJoungho Kim
Joungho Kim
In this article, for the first time, we propose a transformer network-based reinforcement learning (RL) method for power distribution network (PDN) optimization of high bandwidth memory (HBM). The proposed method can provide an optimal decoupling capacitor (decap) design to maximize the reduction of PDN self-and transfer impedances seen at multiple ports. An attention-based transformer network is implemented to directly parameterize decap optimization policy. The optimality performance is significantly improved since the attention mechanism has powerful expression to explore massive combinatorial space for decap assignments. Moreover, it can capture sequential relationships between the decap assignments. The computing time for optimization is dramatically reduced due to the reusable network on the positions of probing ports and decap assignment candidates. This is because the transformer network has a context embedding process to capture meta-features including probing ports positions. In addition, the network is trained with randomly generated datasets. The computing time for training and data cost are critically decreased due to the scalability of the network. Due to its shared weight property and the context embedding process, the network can adapt to a larger scale of problems without additional training. For verification, the results are compared with conventional genetic algorithm (GA), random search (RS), and all the previous RL-based methods. As a result, the proposed method outperforms in all the following aspects: optimality performance, computing time, and data efficiency.

3 March 2023 in IEEE Transactions on Microwave Theory and Techniques
Verified
Journal Article
Closed access
LZLing Zhang
Ling Zhang
LJLi Jiang
Li Jiang
JJJack Juang
Jack Juang
ZYZhiping Yang
Zhiping Yang
ELEr-Ping Li
Er-Ping Li
CHChulsoon Hwang
Chulsoon Hwang
This article presents a new optimization method for complex power distribution networks (PDNs) with irregular shapes and multilayer structures using deep reinforcement learning (DRL), which has not been considered before. A fast boundary integration method is applied to compute the impedance matrix of a PDN structure. Subsequently, a new DRL algorithm based on proximal policy optimization (PPO) is proposed to optimize the decoupling capacitor (decap) placement by minimizing the number of decaps while satisfying the desired target impedance. In the proposed approach, the PDN structure information is encoded into matrices and serves as the input of the DRL algorithm, which increases the flexibility of the method to be extended and generalized to different PDN configurations. Also, the output of the algorithm determines the selection of decap types and locations collaboratively, making it easier to find the optimal solution in a huge search space. The proposed method is compared with the state-of-the-art approaches and shows consistent advantages in reducing the number of decaps in different testing cases.

1 March 2021 in IEEE Transactions on Microwave Theory and Techniques
Verified
Journal Article
Closed access
ZXZhifei Xu
Zhifei Xu
RHRunbing Hua
Runbing Hua
JJJack Juang
Jack Juang
SXShengxuan Xia
Shengxuan Xia
JFJun Fan
Jun Fan
CHChulsoon Hwang
Chulsoon Hwang
This article demonstrates an inaudible attack on smart speakers using electromagnetic interference (EMI). The EMI induces voltages on the order of a few millivolts on conductors, which are then converted into baseband signals by exploiting the inherent nonlinearity of microphones. The EMI signal is specially preprocessed to minimize the useless harmonics generation at the microphone output signals, which significantly improves the recognition rate as well as nullify the previous countermeasures based on the harmonics detection. The sensitive carrier frequency found by our proposed method can improve the attack distance as well. A measurement-based methodology is applied to locate the sensitive regions for noise coupling without knowing the layout of the printed circuit board (PCB), and the transfer function is also obtained to insure the main coupling location. Our experiments show that in open space, intentional EMI under 2.5 W can inject commands at distances up to 2.5 m on smart speakers.

1 July 2019
Verified
Conference Paper
Closed access
LZLing Zhang
Ling Zhang
ZZZhongyang Zhang
Zhongyang Zhang
Chenxi HuangChenxi Huang
Chenxi Huang
HDHan Deng
Han Deng
HLHank Lin
Hank Lin
CHChulsoon Hwang
Chulsoon Hwang
Selection of decoupling capacitors (decaps) is important for power distribution network (PDN) design in terms of lowering impedance and saving cost. Good PDN designs typically mean satisfying a target impedance with as less decaps as possible. In this paper, an inductance-based method is utilized to calculate the port priority fist, and afterwards deep reinforcement learning (DRL) with deep neural network (DNN) is applied to optimize the assignment of decaps on the prioritized locations. The DRL algorithm can explore by itself without any prior physical knowledge, and the DNN is trained with the exploration experience and eventually converges to an optimum state. The proposed hybrid method was tested on a printed-circuit-board (PCB) example. After some iterations of training the DNN successfully reached to an optimum design, which turned out to be the minimum number of decaps that can satisfy the target impedance. The usage of DRL with DNN makes the algorithm promising to include more variables as input and handle more complicated cases in the future.

9 November 2017 in IEEE Transactions on Electromagnetic Compatibility
Verified
Journal Article
Closed access
HKHeegon Kim
Heegon Kim
JKJingook Kim
Jingook Kim
JFJun Fan
Jun Fan
CHChulsoon Hwang
Chulsoon Hwang
Precise analytical models of power supply noise induced jitter (PSIJ) at inverter chains are proposed. Based on the piecewise linear approximated I–V curve model, analytical models of local PSIJ transfer functions at local rising and falling edges are derived. The total PSIJ transfer function of an inverter chain output is then estimated by alternately accumulating the local PSIJ transfer function at local rising and falling edges. Based on several assumptions, the full PSIJ transfer function model is significantly simplified, which provides physical insights of PSIJ at inverter chains. Accuracy of the proposed analytical model is successfully validated by SPICE simulations with 130 nm CMOS technology. In addition, properties of the PSIJ transfer function at inverter chains are analyzed based on the proposed model.

21 June 2021 in IEEE Transactions on Microwave Theory and Techniques
Verified
Journal Article
Closed access
ZXZhifei Xu
Zhifei Xu
ZWZihao Wang
Zihao Wang
YSYin Sun
Yin Sun
CHChulsoon Hwang
Chulsoon Hwang
HDHerve Delingette
Herve Delingette
JFJun Fan
Jun Fan
This article proposes a jitter-aware decoupling capacitors placement optimization method that uses the genetic algorithm (GA). A novel method for defining the optimization target function in regard to power delivery network (PDN) and power source-induced jitter (PSIJ) optimization based on the GA-based tool is proposed. The proposed method can provide an optimum and economic solution for the number of decoupling capacitors to use in a PDN to reach the target impedance. Then, by modifying the optimization target function with our proposed method, an optimum solution of the number of decoupling capacitors regarding the PSIJ can be obtained. The PSIJ analytical expressions are derived in conjunction with a resonant cavity model that includes the coordinates of the decoupling capacitors and the PSIJ transfer function. The GA-based optimization algorithm with the proposed target function is first applied to optimize the number of decoupling capacitors regarding the PSIJ. Finally, the measured jitters from HSPICE simulation results are used to verify our optimization method such that both the simulated results and analytically calculated results support the efficiency of our proposed optimization method.