• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.6
• /
• pp.531-539
• /
• 2013

Fingerprinting technique uses the radio signal strength measured reference locations is typically used, although there are many Wi-Fi based location tracking techniques. However, it needs numerous reference locations for precision and accuracy. This paper the analyzes problems of previous techniques and proposes a fingerprinting system using reliability based on a signal strength map. The system collects the signal strength data from a number of reference locations designated by the developer. And then it generates path-loss models to one of the access points for each reference location. These models calculate the predicted signal strength and reliability for a lattice. To evaluate proposed method and system performance, We perform experiments in a $20m{\times}22m$ real indoor environment installed access points. According to the result, the proposed system reduced distance error than RADAR. Comparing the existing system, it reduced about 1.74m.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.38 no.2
• /
• pp.149-156
• /
• 2014

Floor number determination has attracted considerable attention because many applications require accurate floor number information for providing better localization services in multistory buildings. This paper describes a Wi-Fi-based, low-complexity floor number determination method for multistory buildings. In our scheme, with the exception of floor ID and received signal strength of each WAP, detailed information on the wireless access point (WAP) coordinates is not needed. The multi-wall-floor model is used for the simulation and analysis. The simulation results show that the floor determination accuracy is nearly 100 given an adequate deployment density of WAPs on each floor. The results also show that the proposed method provides a good estimate of floor number even when only a few WAPs are implemented on each floor. Specifically, the proposed method can work under extreme conditions, i.e., where there are no WAPs on a floor.

• Journal of Platform Technology
• /
• v.12 no.2
• /
• pp.12-22
• /
• 2024

With the rapid increase in digital data in modern society, digital forensics plays a crucial role, and file type identification is one of its integral components. Research on the development of identification models utilizing artificial intelligence is underway to identify file types swiftly and accurately. However, existing studies do not support the identification of file types with high domestic usage rates, making them unsuitable for use within the country. Therefore, this paper proposes a more accurate file type identification model using Convolutional Neural Networks (CNN) and Gated Recurrent Units (GRU). To overcome limitations of existing methods, the proposed model demonstrates superior performance on the FFT-75 dataset, effectively identifying file types with high domestic usage rates such as HWP, ALZ, and EGG. The model's performance is validated by comparing it with three existing research models (CNN-CO, FiFTy, CNN-LSTM). Ultimately, the CNN and GRU based file type identification and classification model achieved 68.2% accuracy on 512-byte file fragments and 81.4% accuracy on 4096-byte file fragments.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.5
• /
• pp.2437-2442
• /
• 2013

OFDM Wi-Fi AP is susceptible to the co-channel interference. As a countermeasure, the insertion of a smart has been addressed. Despite of the guaranteed efficiency, the complexity of the post-FFT algorithm often keeps itself from being selected as the countermeasure. Instead, simply constructed pre-FFT smart antenna of which the algorithm is based on the received signal covariance matrix is commonly used and the mathematical modeling of it has been deployed. Computer simulations evaluating the improved BER characteristics of the proposed pre-FFT using the covariance matrix of channel estimator output have been carried out. It has been demonstrated that channel matrix output based smart antenna is superior to that using received signal covariance matrix.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.264-266
• /
• 2022

In order to apply the RNN model to the radio fingerprint-based indoor path generation technology, the data set must be continuous and sequential. However, Wi-Fi radio fingerprint data is not suitable as RNN data because continuity is not guaranteed as characteristic information about a specific location at the time of collection. Therefore, continuity information of sequential positions should be given. For this purpose, clustering is possible through classification of each region based on signal data. At this time, the continuity information between the clusters does not contain information on whether actual movement is possible due to the limitation of radio signals. Therefore, correlation information on whether movement between adjacent clusters is possible is required. In this paper, a deep learning network, a recurrent neural network (RNN) model, is used to predict the path of a moving object, and it reduces errors that may occur when predicting the path of an object by generating continuous location information for path generation in an indoor environment. We propose a method of giving correlation between clustering for generating an improved moving path that can avoid erroneous path prediction that cannot move on the predicted path.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.1
• /
• pp.163-168
• /
• 2016

For the quality increase of agricultural and aquatic products, it is necessary to the efficient management system of integrated control and accurate data for improving the water quality of reservoirs. This may be using the USN save time and money. In this paper, the water surface robot was designed and manufactured for measuring and announcing the environment informations, and it's network was constructed. The robot had more excellent durability and was more inexpensive than previous model. The network was designed considering connection with ZigBee, TRS, Wi-Fi or LTE network. Also the exploration function of the robot was confirmed by the simulation and the user interface was programmed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38B no.7
• /
• pp.553-563
• /
• 2013

When the IEEE 802.11 wireless LAN-based broadcasting services, we use broadcast packets to broadcast multimedia contents to a large number of users using limited wireless resources. However, broadcast transmission is difficult to recover the loss packets compared with unicast transmission. Therefore, analysis of packet loss characteristics is required to perform efficient packet recovery. The packet loss in wireless transmissions is often bursty with high loss data rate. Even if loss patterns have the same average packet loss, they are different in the recovery rate of random loss and burst loss depending on the nature. Therefore, the analysis and research of the nature of the loss are needed to recover loss packets considering bursty characteristics. In this paper, we experimented Wi-Fi broadcast transmission according to transmission rate and analyzed bursty characteristics of loss patterns using 4-state markov model.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.36 no.1
• /
• pp.17-24
• /
• 2012

There has been little study on the performance characteristics of the effective heat pump systems in the cooling mode using the two-stage compression cycles have hardly performed. In this study, the performance of the two-stage compression $CO_2$ cycle with an FI (flash intercooler) and the FGB (flash gas bypass) was investigated by using a theoretical method. The performance analysis was carried out with aby varyingiation (the indoor temperature, outdoor temperature, and 1st- and 2nd-stage EEV openings. As of a result, the coefficients of performance (COPs) of the Bbasic, FI, and FGB cycles were decreased by 28.5%, 22.1%, and 24.5%, respectively, for various outdoor temperature conditions. In addition to, the performance variation of the two-stage compression cycle was smaller than that of the single-stage compression cycle. The performance of the FI and FGB cycles was improved by 13.5%, and 6.9%, respectively, when the 1st-stage EEV opening was increased from 32% to 48%, and by 0.9%, and 2.6%, respectively, when the 1st- andthe 2nd-stage EEV opening was increased from 32% to 48%, andwas increased from 42% to 58%, respectively. The FI cycle showed anthe most improved performance for any given operating conditions.

• Phonetics and Speech Sciences
• /
• v.16 no.1
• /
• pp.67-76
• /
• 2024

Recent advances in text-to-speech (TTS) technology have significantly improved the quality of synthesized speech, reaching a level where it can closely imitate natural human speech. Especially, TTS models offering various voice characteristics and personalized speech, are widely utilized in fields such as artificial intelligence (AI) tutors, advertising, and video dubbing. Accordingly, in this paper, we propose a one-shot multi-speaker TTS system that can ensure acoustic diversity and synthesize personalized voice by generating speech using unseen target speakers' utterances. The proposed model integrates a speaker encoder into a TTS model consisting of the FastSpeech2 acoustic model and the HiFi-GAN vocoder. The speaker encoder, based on the pre-trained RawNet3, extracts speaker-specific voice features. Furthermore, the proposed approach not only includes an English one-shot multi-speaker TTS but also introduces a Korean one-shot multi-speaker TTS. We evaluate naturalness and speaker similarity of the generated speech using objective and subjective metrics. In the subjective evaluation, the proposed Korean one-shot multi-speaker TTS obtained naturalness mean opinion score (NMOS) of 3.36 and similarity MOS (SMOS) of 3.16. The objective evaluation of the proposed English and Korean one-shot multi-speaker TTS showed a prediction MOS (P-MOS) of 2.54 and 3.74, respectively. These results indicate that the performance of our proposed model is improved over the baseline models in terms of both naturalness and speaker similarity.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.10
• /
• pp.2316-2322
• /
• 2013

Due to the development of ICT, with using hardwares such as WiFi, 3G and GPS and so on, smartphone could have provided a lot of applications with novel functions rapidly. Through such applications, lots of personal information such as personal location, personal images, and list of phone calls is created, saved and widely used. Because there is lots of leakage of the stored personal information due to loss of phone and application, privacy violation have been important issue nowadays. Smartphone with GPS and Internet provides location information. To protect the information, the technologies that only the authorized user can access it while inquiring the location information have been proposed. In this paper, to minimize the identification information for location information subject and information user and anonymize the identifiable information such as phone number, we proposed a model that can reduce the leakage of information and avoid the wrong usage of the stored information in the server. This technique will be used for protecting privacy when developing the application that provides routing service through location history information.