Earticle

In this paper, we propose a PTQ (Post Training Quantization) static with QO (Quantization Only) technique to efficiently deploy and execute deep learning models. A comparative performance evaluation of PTQ static with QDQ (Quantize and DeQuantize) and the proposed quantization method was conducted using the MNIST (Modified National Institute of Standards and Technology database) dataset and 8- bit quantization. Experimental results indicate that the PTQ static with QO method reduces the size of the model by approximately 33%, increases the inference speed by 1.5 times, and minimizes the accuracy loss, similar to the PTQ static with QDQ method. The proposed PTQ static with QO method offers a significant technical enhancement to facilitate the efficient deployment and execution of AI (Artificial Intelligence) models through the quantization of deep learning models. We have shown that the PTQ static with QO method is a beneficial and efficient approach to decrease the size and computation of deep learning models. This study makes novel contributions to the quantization of deep learning models. The practical potential of the PTQ static with QO method lies in its ability to be more suitably deployed for the purposes of AI hardware.

As environmental damage increase by a highly developed material civilization of today, many companies take a growing immensely interest in the influence of environment for beginning a new paradigm year by year. The previous assessments dose not run the gamut of industry but is confined within a certain facility or an area. Industrial processes and operations can not be accomplished independently but are connected with each others through suppliers and customer, and these ideas are fundamental notions of Life Cycle Assessment(LCA). This paper will introduce Life Cycle Assessment(LCA) in environment which is rising, and would like to build environmental management system using approach of Quality Function Deployment(QFD) and Safety Function Deployment(SFD) belonging to the assessment method.

Vehicular density is usually low in a highway environment. Consequently, connectivity of the vehicular ad hoc networks (VANETs) might be poor. We are investigating the problem of deploying the approximation optimal roadside units (RSUs) on the highway covered by VANETs, which employs VANETs to provide excellent connectivity. The goal is to estimate the minimal number of deployed RSUs to guarantee the connectivity probability of the VANET within a given threshold considering that RSUs are to be allocated equidistantly. We apply an approximation algorithm to distribute RSUs locations in the VANETs. Thereafter, performance of the proposed scheme is evaluated by calculating the connectivity probability of the VANET. The simulation results show that there is the threshold value M of implemented RSUs corresponding to each vehicular network with N vehicles. The connectivity probability increases slowly with the number of RSUs getting larger.

Wireless sensor network acquires information from the surroundings using sensor, processes it, and then delivers it to other sensors or sink nodes through wireless interface. This kind of application of wireless sensor network can be used for various areas such as army, private sector, hospital and so on. To improve the efficiency of wireless sensor network, the study on the effective sensor placement methods is being conducted in many ways. Most of these studies do not mention that the environmental factors affect nodes when placing nodes. The reason is because there are too many environmental factors or sensors to be considered that affect other sensors, and because these factors are not easy to be recognized or classified. However, this study classifies and recognizes the factors that we need to consider when placing wireless sensor nodes. In addition, it finds how the sensor’s coverage affects those recognized items. By doing this, this study can be a foundation for the effective placement of sensor nodes in the future when wireless sensor network is built.

This study proposes a pre-deployment key management scheme that requires a few memory capacities and CPU computations to address secure data transmissions in Wireless Sensor Networks (WSNs). The proposed scheme exploits threshold key management mechanisms by Lagrange Interpolation polynomial generating a key set for sensor nodes, and uses symmetric and irreversible cryptography schemes to encrypt transmitted data by the generated keys with Message Authentication Code (MAC). The sensor nodes merely have to aggregate and encrypt received data without complicated cryptography operations. The proposed approach can achieve rapid and efficient secure data transmissions with low communications, and is proper to be implemented on large-scale sensor networks.

Recent advancements in MEMSs technology, VLSI and Wireless Private Area Networks have contributed in developing enormous interest in wireless sensor networks by both academia and research community.WSN is formed by large number of small, cheap and low energy sensors that work together in monitoring regions. WSN find its application in many areas such as object tracking, disaster relief and smart environments. Most of these applications need location awareness of sensor nodes to estimate the origin of events. Localization has remained a challenge in successful implementation of WSN and so, it is a hotspot of research. Range free localization techniques are given lot of importance in recent research works because they suit well for low cost, low energy and less infrastructure requirements of WSN and also involve lesser computation complexity. DV-HOP is one of the range free techniques that have attracted researchers’ interest. In this paper, DV-HOP algorithm is implemented by deploying anchor nodes in a structured way at various radio ranges. Structured deployment of anchors give more accurate location estimates than random deployment that was followed in basic DV-HOP, which has been demonstrated in a detailed manner by the simulation results.

This study examines the use of score auctions in spectrum allocation as a policy instrument for promoting efficient network deployment. Because mobile networks generate substantial positive externalities, including broader connectivity benefits, industrial spillovers, and public-interest gains, market incentives alone may lead to insufficient investment relative to the social optimum. Governments have traditionally addressed this problem through direct regulation, typically by imposing mandatory build-out obligations as a condition of spectrum assignment. In South Korea, for example, regulators have commonly required assignees to deploy a fixed number of base stations within specified time periods. While this command-and-control approach is effective in achieving predetermined deployment targets, it is often inefficient because it lacks flexibility and cannot accommodate heterogeneity in operators’ deployment costs and business conditions. This paper shows that an indirect regulatory approach based on score auctions can improve the efficiency of network deployment. In a score auction, bidders submit both a price bid and a quality commitment, and the winner is selected according to a scoring rule that combines both components. In our model, each bidder’s private information is a two-dimensional vector consisting of a valuation for the spectrum and a cost parameter governing the cost of deployment. We analyze the first-score and second-score auction formats and show that both can be reduced to a tractable one-dimensional model. Our main results are twofold. First, we show that by designing the scoring rule to internalize network externalities, a score auction can align bidders’ private incentives with the social objective, thereby attaining the socially optimal outcome. Implementing this optimal mechanism, however, requires an estimate of the entire externality function, which is often unavailable to the regulator. To mitigate this informational burden, our second result shows that even when the global externality function is unknown, a score auction based on a point estimate of the marginal externality yields higher welfare than traditional direct regulation. This result provides a practical compromise between efficiency and implementability. Overall, our findings offer a theoretical foundation for shifting from direct to indirect regulation to enhance network deployment efficiency.

무선 센서네트워크는 센싱 지역에 분산되어 있는 초소형 센서 노드들이 감지/처리한 데이터를 수집노드로 전송하여 원격의 사용자가 센싱 지역의 상황인지를 가능하게 하는 유비쿼터스 컹퓨팅의 기반 네트워크이다. 전력 사용이 극히 제한된 센서 노드를 이용하여 무선 센서 망을 구성, 유지하며 데이터를 수집하기 위해서는 효율적인 망의 형태와 이에 따른 네트워크 배치 전략을 필요로 한다. 클러스터 기반 네트워크의 형태는 밀집도가 높은 센서네트워크에서 데이터병합을 수행하고 노드간 에너지 소비 균형을 이루기 위한 효과적인 구조이다. 본 논문에서는 클러스터 기반의 단일 홉 전송 구조에서 데이터 수집률을 높이고 노드간 에너지 사용의 균형을 이루게 하여 네트워크의 생존시간을 최대화할 수 있는 네트워크 배치 방법을 제안하였다. 클러스터링 기법에 따른 에너지 소비 모델을 분석하고 이를 통해 노드의 적절한 밀집도를 산출하였다. 싱크로부터 멀어질수록, 센서필드의 중앙보다는 외곽에 노드의 배치를 조밀하게 배치하여 네트워크의 생존시간을 늘렸다. 시뮬레이션을 통해 효율적인 네트워크의 배치가 노드간 에너지 소비의 균형을 이루도록 하여 네트워크의 생존시간을 늘일 수 있음을 확인하였다.

우리는 센서 영역에서 노드들을 효율적으로 배치하기 위한 새로운 관점의 문제를 제안하고자 한다. 기존 연구에서는 센서 영역에서 발생하는 이벤트를 잘 감지할 수 있도록 센서를 배치하기 위해서 임의의 지역에서의 센싱 능력 목표량(Threshold) T의 값 이상을 보장하도록 배치하는 방법을 사용하였다. 하지만, 센서 영역의 형태가 달라지거나 센서 영역 내에 중요성이 다른 지역이 있다면 위의 방법이 효율적이지 않다. 따라서 본 연구에서는 센서들의 평균적인 센싱 능력이 최대가 되도록 배치하는 접근방안과 그 필요성을 제안한다.

We propose the new approach for the efficient sensor deployment in sensor networks. In the existing works, they use the approach that guarantee the sensing ability T for any point in the sensor field. However, if the shape of the sensor field is changed or it is composed of the sections that have different importance each other, the previous approach is not efficient. So, in this work, we propose the approach that maximize the average sensing ability and it's necessity.

As we have understood the commonly launched T-DMB service from December 2005, we kindly recommend to apply LCX cable installing solution which can be covering underground subway and tunnel T-DMB signal shadow area with most efficient network implementation method. Subway passengers would be enjoyable continuous T-DMB services through LCX cable installing solution.

현재 아날로그 방식의 마을방송 시스템은 기술 적용의 문제 및 시설 노후화로 인한 성능 저하 등의 문제로 인하여 무선 디지털 시스템으로의 전환과 이에 따른 시스템 표준이 요구되고 있다. 그러나 주파수 부족으로 인해 효율적인 디지털 무선 마을 방송 시스템의 구축을 위해서는 이웃마을간의 간섭분석이 중요하므로 여기서는 대표적인 간이무선국 dPMR(digital private mobile radio) 및 DMR(digital mobile radio) 방식을 고려하여 동일채널 간섭과 인접채널 간섭에 대한 시뮬레이션을 수행하였다. 동일채널 간섭분석에서는 마을간 이격거리와 인접채널 간섭분석에서는 주파수 오프셋을 사용하여 주파수 재사용 및 채널분리에 대한 본 논문의 결과는 향후 마을방송 표준개발과 시범 사업화를 위한 자료로 활용될 것이다.

Since the existing analog village broadcasting system has some technical problems in applying and degradations in performance due to its old equipments, it had been required recently to be changed to a wireless digital system and to develop the standardization. However, it is important to analyze the interference between villages in deploying the efficient digital wireless village broadcasting system. In this paper, simulations for co-channel and adjacent channel interference have been carried out considering digital private mobile radio(dPMR) and digital mobile radio(DMR) systems as a representative mobile radio. These results for frequency reuse and channel separation drawn from the separation distance between villages in co-channel interference and the frequency offset in adjacent interference can be helpful to establish a standard and the testing service in the near future.

IT기술의 급격한 발전을 토대로 유비쿼터스 사회로 변화함에 따라, USN(Ubiquitous Sensor Network) 기술이 매우 활성화되어있으며 연구 분야로 주목을 받고 있다. 그러나 USN 환경 구축시 효울적인지 실험을 하기위해서는 상당한 시간 및 비용이 따르게 된다. 본 논문에서는 USN 환경을 GML로 구성하고, 장애물에 대해 Map Object 여부 설정을 통해 타겟 지역 설정뿐만 아니라, MSN(Mobile Sensor Node)과 FSN(Fixed Sensor Node)에 대하여 동적 및 다양한 정적 배치가 가능한 멀티형 시뮬레이터인 IS_WSN을 제안한다.

IT기술의 급격한 발전을 토대로 유비쿼터스 사회로 변화함에 따라, USN(Ubiquitous Sensor Network) 기술이 매우 활성화되어있으며 연구 분야로 주목을 받고 있다. 그러나 USN 환경 구축시 효울적인지 실험을 하기위해서는 상당한 시간 및 비용이 따르게 된다. 본 논문에서는 USN 환경을 GML로 구성하고, 장애물에 대해 Map Object 여부 설정을 통해 타겟 지역 설정뿐만 아니라, MSN(Mobile Sensor Node)과 FSN(Fixed Sensor Node)에 대하여 동적 및 다양한 정적 배치가 가능한 멀티형 시뮬레이터인 IS_WSN을 제안한다.

Solar arrays and antennas of the satellite are usually stowed within the dimensions of the launch-vehicle fairing and deployed in the orbit. To solve such multibody dynamic problems, differential equations and algebraic equations are simultaneously solved, and special solution techniques are required. In this paper, Lagrange multipliers associated with the constraints are iteratively computed by monotonically reducing an appropriately defined constraint error vector, and the resulting equation of motion is solved by a well-established ODE technique. Defomable bodies as well as rigid bodies are treated, and applications to satellite solar arrays are explained.