• Journal of Science and Technology Studies
• /
• v.2 no.2 s.4
• /
• pp.1-23
• /
• 2002

This Article examines the main approaches to the public understanding of science(PUS) in light of the changing relation of science and public. Traditional approach called deficit model recognizes scientific knowledge as a entity, unidirectionally diffusing to public. This View basically presupposes the gap between science and public. Meanwhile, this approach has an aspiration to reduce the gap. So there is a paradoxical situation in the traditional PUS. Public perception of environmental crisis and risk of science has been risen since late 1980's. And the science technology studies (STS) have criticized traditional approach. STS proposes new approach of constructive PUS. Constructive PUS conceives the public to have heterogeneous, local characteristic. This approach has been very fertile both in theory and practice. But most recently, newly proposed approach, so called heterogeneous PUS, criticizes constructive PUS. Main point of criticism is that constructive PUS has dichotomy between science and public, and romanticizes the public. It is uncertain whether heterogeneous PUS can take place of constructive PUS. But this trend has the implication that the relation of science and public is constantly changing.

• Journal of Korea Technology Innovation Society
• /
• v.6 no.2
• /
• pp.137-158
• /
• 2003

This paper examines theoretical trends and policy issues concerning science and public. The dominant perspective about it was changed from “popularization of science” to “public understanding of science (PUS)”. PUS pays attention to active roles of public in the understanding of science based on the contextual model. And recently various concepts are attempted to complement PUS such as “heterogeneous PUS”, “quantitative PUS”, “public understanding of research”, and “public participation in science”. PUS related activities in Korea can be analyzed using the concept of “science and technology culture system”. The characteristics of science and technology culture system in Korea can be summarized as follows: Its purpose is confined to support existing national innovation system; Its constituents have been unevenly and separately developed; Its material and institutional infrastructure is not sufficiently prepared; Its organizing mechanism is strongly influenced by government's execution power.

• Journal of Science and Technology Studies
• /
• v.2 no.2 s.4
• /
• pp.25-54
• /
• 2002

Without much communication, large-scale surveys of public knowledge of, and attitudes to, science (quantitative PUS) and case-study analyses of the public's understandings of science in particular (constructivist PUS) have dominated in the public understanding of science (PUS) area. Not only methodological preference but also a strong antipathy against value-orientations that each approach presumed to have and support has been barriers for quantitative PUS and constructivist PUS to benefit each other. In order to overcome such barriers, this paper demonstrates that value orientations guiding quantitative PUS have been much more diverse than what constructivist PUS researchers might think, and that quantitative PUS has indeed yielded the results consistent with and complementary to constructivist PUS. Finally this paper proposes that (1)quantitative PUS should test propositions provided by constructivist PUS, so that it can contribute much to the construction of more generalizable PUS theories and policies, and (2)constructivist PUS uses the outcome of quantitative PUS to develop more complex case studies which consider heterogeneous publics, trends of public evaluations of science, and how public attitudes to science in the abstract and public attitudes to science in particular in a specific context have effect on each other.

• Journal of Communications and Networks
• /
• v.13 no.4
• /
• pp.360-366
• /
• 2011

In this paper, we investigate the problem of power allocation in a heterogeneous network that is composed of a pair of cognitive users (CUs) and an infrastructure-based primary network. Since CUs have only limited effective spectrum-sensing ability and primary users (PUs) are not active all the time in all locations and licensed bands, we set up a new multi-area model to characterize the heterogeneous network. A novel combined interference-avoidance policy corresponding to different PU-appearance situations is introduced to protect the primary network from unacceptable disturbance and to increase the spectrum secondary-reuse efficiency. We use dual decomposition to transform the original power allocation problem into a two-layer optimization problem. We propose a low-complexity joint power-optimizing method to maximize the transmission rate between CUs, taking into account both the individual power-transmission constraints and the combined interference power constraint of the PUs. Numerical results show that for various values of the system parameters, the proposed joint optimization method with combined PU protection is significantly better than the opportunistic spectrum access mode and other heuristic approaches.

• Journal of Communications and Networks
• /
• v.18 no.1
• /
• pp.112-122
• /
• 2016

Cognitive radio (CR) is considered an attractive technology to deal with the spectrum scarcity problem. Multi-radio access technology (multi-RAT) can improve network capacity because data are transmitted by multiple RANs (radio access networks) concurrently. Thus, multi-RAT embedded in a cognitive radio network (CRN) is a promising paradigm for developing spectrum efficiency and network capacity in future wireless networks. In this study, we consider a new CRN model in which the primary user networks consist of heterogeneous primary users (PUs). Specifically, we focus on the energy-efficient resource allocation (EERA) problem for CR users with a special location coverage overlapping region in which heterogeneous PUs operate simultaneously via multi-RAT. We propose a two-tier crossover genetic algorithm-based search scheme to obtain an optimal solution in terms of the power and bandwidth. In addition, we introduce a radio environment map to manage the resource allocation and network synchronization. The simulation results show the proposed algorithm is stable and has faster convergence. Our proposal can significantly increase the energy efficiency.

• ETRI Journal
• /
• v.45 no.2
• /
• pp.318-328
• /
• 2023

Recently, embedded systems, such as mobile platforms, have multiple processing units that can operate in parallel, such as centralized processing units (CPUs) and neural processing units (NPUs). We can use deep-learning compilers to generate machine code optimized for these embedded systems from a deep neural network (DNN). However, the deep-learning compilers proposed so far generate codes that sequentially execute DNN operators on a single processing unit or parallel codes for graphic processing units (GPUs). In this study, we propose PartitionTuner, an operator scheduler for deep-learning compilers that supports multiple heterogeneous PUs including CPUs and NPUs. PartitionTuner can generate an operator-scheduling plan that uses all available PUs simultaneously to minimize overall DNN inference time. Operator scheduling is based on the analysis of DNN architecture and the performance profiles of individual and group operators measured on heterogeneous processing units. By the experiments for seven DNNs, PartitionTuner generates scheduling plans that perform 5.03% better than a static type-based operator-scheduling technique for SqueezeNet. In addition, PartitionTuner outperforms recent profiling-based operator-scheduling techniques for ResNet50, ResNet18, and SqueezeNet by 7.18%, 5.36%, and 2.73%, respectively.