Find Research Output

Research Output
  • All
  • Scholar Profiles
  • Research Units
  • Research Output
Filter
Department Publication Year Content Type Data Sources

SELECTED FILTERS

Department of Communications and Networking
Clear all

1.Automatic Building and Floor Classification using Two Consecutive Multi-layer Perceptron

Author:Cha, J;Lee, S;Kim, KS

Source:2018 18TH INTERNATIONAL CONFERENCE ON CONTROL, AUTOMATION AND SYSTEMS (ICCAS),2018,Vol.2018-October

Abstract:Key issues of indoor localization is taking full advantages and overcoming its disadvantages. indoor localization based on Wi-Fi fingerprinting attracts researchers' attentions since it does not require new infrastructure and devices. Many devices such as smart phones and laptops, which have a function to capture Wi-Fi signals, can be used for Wi-Fi fingerprinting. However, due to unreliable Wi-Fi signals, there are still difficulty to achieve high positioning accuracy. The unreliable signal disturbs devices to find their locations. As a result, getting localization with devices sometimes makes a wrong decision in building classification. It is useless for people to find a destination floor if they are in different building. In this paper, we propose two consecutive multi-layer perceptrons to get more precise localization. With sumple structure, we get better performance and show precise decision results in building classification, which is critical in Wi-Fi fingerprinting. We use UJIndoorLoc dataset which is open dataset.

2.An Implantable and Conformal Antenna for Wireless Capsule Endoscopy

Author:Wang, JC;Leach, M;Lim, EG;Wang, Z;Pei, R;Huang, Y

Source:IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS,2018,Vol.17

Abstract:This letter proposes an implantable antenna with ultrawide bandwidth operating in the medical device radio communications service band (401-406 MHz) for the wireless capsule endoscopy (WCE). The simulation and experimental results show the proposed antenna has a good performance in terms of the return loss and hence the bandwidth from 284 to 825 MHz. The maximum realized gain of this antenna is -31.5 dBi at 403 MHz. The maximum simulated input power is <1.7 mW in order to satisfy the specific absorption rate (SAR) regulations in the IEEE standard. The tolerance of the antenna owing to bendability and different WCE shell thicknesses is investigated. These indicate that the proposed antenna is a good candidate for the WCE.

3.Fourier-transform-based two-stage camera calibration method with simple periodical pattern

Author:Chen, XC;Fan, RM;Wu, J;Song, XK;Liu, Q;Wang, YW;Wang, YJ;Tao, B

Source:OPTICS AND LASERS IN ENGINEERING,2020,Vol.133

Abstract:Clear and focused pattern images are essential prerequisites for accurate feature detection in traditional camera calibration methods, which introduce numerous limitations in various areas, such as long-distance photogrammetry. A feature detection method robust against defocusing is proposed for extracting the centers or corners of a planar square periodic target. A Fourier transform is employed to calculate two wrapped phase maps from the periodic target images, which are then used to accurately extract the feature points. The calibration procedure is divided into two stages to obtain more accurate results. A rough calibration is performed to calculate the rotation angles between the target and the camera. If the tilt angle is larger than 12 degrees, the corresponding images are removed. Subsequently, the remaining images are used for precise calibration. The simulations and the experiments demonstrate that the proposed method can accurately calibrate a camera with a planar square periodic pattern, even in the case of severe defocusing.

4.Chromosome Classification with Convolutional Neural Network based Deep Learning

Author:Zhang, WB;Song, SF;Bai, TM;Zhao, YX;Ma, F;Su, JL;Yu, LM

Source:2018 11TH INTERNATIONAL CONGRESS ON IMAGE AND SIGNAL PROCESSING, BIOMEDICAL ENGINEERING AND INFORMATICS (CISP-BMEI 2018),2018,Vol.

Abstract:Karyotyping plays a crucial role in genetic disorder diagnosis. Currently Karyotyping requires considerable manual efforts, domain expertise and experience, and is very time consuming. Automating the karyotyping process has been an important and popular task. This study focuses on classification of chromosomes into 23 types, a step towards fully automatic karyotyping. This study proposes a convolutional neural network (CNN) based deep learning network to automatically classify chromosomes. The proposed method was trained and tested on a dataset containing 10304 chromosome images, and was further tested on a dataset containing 4830 chromosomes. The proposed method achieved an accuracy of 92.5%%, outperforming three other methods appeared in the literature. To investigate how applicable the proposed method is to the doctors, a metric named proportion of well classified karyotype was also designed. An result of 91.3%% was achieved on this metric, indicating that the proposed classification method could be used to aid doctors in genetic disorder diagnosis.

5.Multi-GHz Microstrip Transmission Lines Realised by Screen Printing on Flexible Substrates

Author:Shi, YZ;Jiang, ZZ;Lam, S;Leach, M;Wang, JC;Lim, EG

Source:2017 IEEE ELECTRICAL DESIGN OF ADVANCED PACKAGING AND SYSTEMS SYMPOSIUM (EDAPS),2017,Vol.2018-January

Abstract:This paper reports experimental work on 50 SI microstrip transmission lines implemented by screen-printing low-cost silver paste onto thin flexible polyethylene terephthalate (PET) substrates of varying thickness. The microstrip line designs are based on PET substrates with thicknesses of 1.4 mm, 0.7 mm and 0.5 mm, leading to conductive track widths of 3.8 mm, 1.7 mm and 1.2 mm respectively for a 50 SI line; these designs were then realised. The S-parameter measurements show that the insertion loss of the microstrip transmission lines on each substrate can be as low as 0.2 dB/cm, 0.17 dB/cm, and 0.14 dB/cm up to a frequency of 5 GHz in spite of the average quality of the silver paste used. The experimental results also show that the screen-printed transmission lines still work quite well in bent condition and wearable electronics application at GHz is possible.

6.3D VIDEO CODING USING MOTION INFORMATION AND DEPTH MAP

Author:Cheng, F;Xiao, JM;Tillo, T

Source:2015 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA & EXPO (ICME),2015,Vol.2015-

Abstract:In this paper, a motion-information-based 3D video coding method is proposed for the texture plus depth 3D video format. The synchronized global motion information of camcorder is sampled to assist the encoder to improve its rate-distortion performance. This approach works by projecting temporal previous frames into the position of the current frame using the depth and motion information. These projected frames are added in the reference buffer as virtual reference frames. As these virtual reference frames are more similar to the current frame than the conventional reference frames, the required residual information is reduced. The experimental results demonstrate that the proposed scheme enhances the coding performance in various motion conditions including rotational and translational motions.

7.XJTLUIndoorLoc: A New Fingerprinting Database for Indoor Localization and Trajectory Estimation Based on Wi-Fi RSS and Geomagnetic Field

Author:Zhong, ZH;Tang, Z;Li, XX;Yuan, TC;Yang, Y;Wei, M;Zhang, YY;Sheng, RZ;Grant, N;Ling, CF;Huan, XT;Kim, KS;Lee, S

Source:2018 SIXTH INTERNATIONAL SYMPOSIUM ON COMPUTING AND NETWORKING WORKSHOPS (CANDARW 2018),2018,Vol.

Abstract:In this paper, we present a new location fingerprinting database comprised of Wi-Fi received signal strength (RSS) and geomagnetic field intensity measured with multiple devices at a multi-floor building in Xi'an Jiatong-Liverpool University, Suzhou, China. We also provide preliminary results of localization and trajectory estimation based on convolutional neural network (CNN) and long short-term memory (LSTM) network with this database. For localization, we map RSS data for a reference point to an image-like, two-dimensional array and then apply CNN which is popular in image and video analysis and recognition. For trajectory estimation, we use a modified random way point model to efficiently generate continuous step traces imitating human walking and train a stacked two-layer LSTM network with the generated data to remember the changing pattern of geomagnetic field intensity against (x, y) coordinates. Experimental results demonstrate the usefulness of our new database and the feasibility of the CNN and LSTMbased localization and trajectory estimation with the database.

8.Determine the Permittivity of the Plastic Materials

Author:Lim, EG;Wang, Z;Leach, MP;Gray, D;Man, KL;Zhang, N

Source:2014 INTERNATIONAL SYMPOSIUM ON COMPUTER, CONSUMER AND CONTROL (IS3C 2014),2014,Vol.

Abstract:Microwave dielectric measurements are difficult to make, each method involves a compromise between accuracy, experimental simplicity, and complexity of the analysis. It is with these ideas in mind that we present a technique in which a waveguide is completely filled with the dielectric of interest to determine the dielectric constant for plastic materials (e.g. Rexolite, Lacqrene and PTFE). Good agreements between measured and manufacturer material specifications have been obtained.

9.Analysis of a Similarity Measure for Non-Overlapped Data

Author:Lee, S;Cha, J;Theera-Umpon, N;Kim, KS

Source:SYMMETRY-BASEL,2017,Vol.9

Abstract:A similarity measure is a measure evaluating the degree of similarity between two fuzzy data sets and has become an essential tool in many applications including data mining, pattern recognition, and clustering. In this paper, we propose a similarity measure capable of handling non-overlapped data as well as overlapped data and analyze its characteristics on data distributions. We first design the similarity measure based on a distance measure and apply it to overlapped data distributions. From the calculations for example data distributions, we find that, though the similarity calculation is effective, the designed similarity measure cannot distinguish two non-overlapped data distributions, thus resulting in the same value for both data sets. To obtain discriminative similarity values for non-overlapped data, we consider two approaches. The first one is to use a conventional similarity measure after preprocessing non-overlapped data. The second one is to take into account neighbor data information in designing the similarity measure, where we consider the relation to specific data and residual data information. Two artificial patterns of non-overlapped data are analyzed in an illustrative example. The calculation results demonstrate that the proposed similarity measures can discriminate non-overlapped data.

10.S(2)Net: A Security Framework for Software Defined Intelligent Building Networks

Author:Xue, N;Huang, X;Zhang, J

Source:2016 IEEE TRUSTCOM/BIGDATASE/ISPA,2016,Vol.

Abstract:During the last ten years, the progress of the Internet of Things (IoT) has continued unabated and gradually expanded into almost every corner of our daily lives. One very typical use case of IoT is the intelligent building (IB). Along with the remarkable rise of IoT, an increasing number of smart devices will be connected to IB networks, which brings many challenges to current IB network architectures. SDN (Software Defined Networking) is a potential solution to future IB networks; however, there are still few studies on SDN-based IB networks. In this paper, we have proposed a secure SDN framework named as S(2)Net, offering authentication and key agreement between the central controller and smart devices. In addition, we have designed a secure OpenFlow message issuing protocol for IB networks; it is used to securely exchange messages between the controller and smart devices. Also, these security protocols are evaluated regarding their security and performance. In doing so, we take an important step towards integrating SDN with extant IB networks.

11.Prediction of Days in Hospital for Children Using Random Forest

Author:Wang, CG;Dong, XL;Yu, LM;Ye, LS;Zhuang, WF;Ma, F

Source:2017 10TH INTERNATIONAL CONGRESS ON IMAGE AND SIGNAL PROCESSING, BIOMEDICAL ENGINEERING AND INFORMATICS (CISP-BMEI),2017,Vol.2018-January

Abstract:In this study, a method was developed to predict the number of hospitalization days of infant patients. The random forest algorithm, along with a data set consisted by records extracted from a hospital information system, was utilized to develop a model to predict the days in hospital. When half of randomly selected records was used as training set to train the random forest algorithm and the other half was used as testing set to test the trained model, the random forest method achieved good predictive accuracy with RMSE being 0.314, R-2 being 0.706, IR1 being 0.545, and Acc 1 being 71%%, which is better than the results obtained by Adaboost method and Bagging method. Experiment on three subgroups of records: a group with all data, a group with records having less than or equal to 14 days in hospital, and a group with records having greater than 14 days in hospital, shows that the prediction of the developed method on the group having more than 14 days in hospital was better than predictions on other groups. Analysis to the importance of three different types of feature sets to the accuracy of prediction reveals that the feature set relating to personal information contribute more to the prediction than other types of features.

12.Two-wavelength phase-shifting method with four patterns for three-dimensional shape measurement

Author:Wu, J;Zhou, ZH;Liu, Q;Wang, YJ;Wang, YW;Gu, YH;Chen, XC

Source:OPTICAL ENGINEERING,2020,Vol.59

Abstract:Fringe projection technique has been widely used for three-dimensional (3D) shape measurement. However, it remains challenging to achieve high-speed measurement. A two-wavelength phase-shifting profilometry method with only four patterns is presented. Specifically, all these four patterns contain two wavelength components. The short wavelength component was used to compute the wrapped phase map, while the long one was used to unwrap the wrapped phase map. The performance of the proposed method was validated by both simulation study and experimental results. (C) 2020 Society of Photo-Optical Instrumentation Engineers (SPIE)

13.Design of An Arduino-based Smart Car

Author:Wang, Z;Lim, EG;Wang, WW;Leach, M;Man, KL

Source:2014 INTERNATIONAL SOC DESIGN CONFERENCE (ISOCC),2014,Vol.

Abstract:Remote-controlled cars are one of the most popular toy products currently on the mass market. Each series of car has a specific remote-control unit. This presents the consumer with a critical problem; obtaining a substitution controller where the original control unit has broken down. In this work a robot based on an external Arduino microcontroller, controllable by an Android application via Bluetooth, which can be recommended as a prototype for the combination of embedded systems with Android mobile devices is investigated.

14.Unbalancing Computations for SM2 Key Exchange Protocols

Author:Zhao, CX;Zhang, J;Huang, X

Source:PROCEEDINGS OF 2020 IEEE 4TH INFORMATION TECHNOLOGY, NETWORKING, ELECTRONIC AND AUTOMATION CONTROL CONFERENCE (ITNEC 2020),2020,Vol.

Abstract:SM2 is a national ecliptic curve public key algorithm standard promulgated by the State Cryptography Administration (SCA) of China. It contains an authenticated key exchange (AKE) protocol which can establish shared secret keys for two communicating parties. However, the computational burden is generally too heavy for capability-limited devices in Internet of Things (IoT). This paper alleviates its computational burden on limited devices by transferring some time-consuming operations from a limited device to its more powerful communicating partner. Two computationally unbalanced protocols are designed. Compared with the original protocols, the computationally unbalanced protocols are more friendly to limited devices.

15.Neural-Network-Based Building Energy Consumption Prediction with Training Data Generation

Author:Lee, S;Cha, J;Kim, MK;Kim, KS;Pham, VH;Leach, M

Source:PROCESSES,2019,Vol.7

Abstract:The importance of neural network (NN) modelling is evident from its performance benefits in a myriad of applications, where, unlike conventional techniques, NN modeling provides superior performance without relying on complex filtering and/or time-consuming parameter tuning specific to applications and their wider ranges of conditions. In this paper, we employ NN modelling with training data generation based on sensitivity analysis for the prediction of building energy consumption to improve performance and reliability. Unlike our previous work, where insignificant input variables are successively screened out based on their mean impact values (MIVs) during the training process, we use the receiver operating characteristic (ROC) plot to generate reliable data with a conservative or progressive point of view, which overcomes the issue of data insufficiency of the MIV method: By properly setting boundaries for input variables based on the ROC plot and their statistics, instead of completely screening them out as in the MIV-based method, we can generate new training data that maximize true positive and false negative numbers from the partial data set. Then a NN model is constructed and trained with the generated training data using Levenberg-Marquardt back propagation (LM-BP) to perform electricity prediction for commercial buildings. The performance of the proposed data generation methods is compared with that of the MIV method through experiments, whose results show that data generation using successive and cross pattern provides satisfactory performance, following energy consumption trends with good phase. Among the two options in data generation, i.e., successive and two data combination, the successive option shows lower root mean square error (RMSE) than the combination one by around 400 similar to 900 kWh (i.e., 30%%similar to 75%%).

16.UWB antenna with superior low-frequency performance

Author:Leach, M;Wang, Z;Juans, G;Lim, EG

Source:MICROWAVE AND OPTICAL TECHNOLOGY LETTERS,2016,Vol.58

Abstract:This article presents the design of a compact, low-profile planar antenna with a circular radiator. With careful design of the radiator and ground plane, the antenna's low-frequency range extends to 1.5 GHz, in addition to operating as an ultrawideband (UWB) antenna across the entire Federal Communications Commission allocated UWB range. (c) 2016 Wiley Periodicals, Inc. Microwave Opt Technol Lett 58:118-121, 2016

17.Review of wearable antennas for WBAN applications

Author:Wang, J.C. ; Lim, E.G. ; Leach, M. ; Wang, Z. ; Man, K.L.

Source:IAENG International Journal of Computer Science,2016,Vol.43

Abstract:Recent research into wearable antennas has provoked increasingly widespread concern about the use of wireless body area network (WBAN) applications. This paper reviews state-of-the-art wearable antennas on textile materials focusing on designs with dual band and UWB and wearable antennas with metamaterials with single band and dual band. This paper also presents the challenges and considerations when designing a suitable wearable antenna. © 2016, IAENG International Journal of Computer Science.

18.Toward Fully-Shared Access: Designing ISP Service Plans Leveraging Excess Bandwidth Allocation

Author:Kim, KS

Source:2014 INTERNATIONAL CONFERENCE ON INFORMATION AND COMMUNICATION TECHNOLOGY CONVERGENCE (ICTC),2014,Vol.

Abstract:Shaping subscriber traffic based on token bucket filter (TBF) by Internet service providers (ISPs) results in waste of network resources in shared access when there are few active subscribers, because it cannot allocate excess bandwidth in the long term. New traffic control schemes have been recently proposed to allocate excess bandwidth among active subscribers proportional to their token generation rates. In this paper we report the current status of our research on designing flexible yet practical service plans exploiting excess bandwidth allocation enabled by the new traffic control schemes in shared access networks, which are attractive to both ISP and its subscribers in terms of revenue and quality of service (QoS) and serve as a stepping stone to fully-shared access in the future.

19.MTFS: Merkle-Tree-Based File System

Author:Kan, J;Kim, KS

Source:2019 IEEE INTERNATIONAL CONFERENCE ON BLOCKCHAIN AND CRYPTOCURRENCY (ICBC),2019,Vol.

Abstract:The blockchain technology has been changing our daily lives since Bitcoin-i.e., the first decentralized cryptocurrency-was invented and released as open-source software by an unidentified person or a group called Satoshi Nakamoto in 2009. Of many applications which can be implemented based on the blockchain, storage is an important one, a notable example of which is the InterPlanetary File System (IPFS). IPFS is a distributed web based on a peer-to-peer hypermedia protocol to make the web faster, safer, and more open and focuses on public accessible files. To provide a solution for private file storage in the blockchain way, in this paper we propose a Merkle-tree-based File System (MTFS). In MTFS, the blockchain is more than a trust machine; it is an abstract of a cluster system. Distributed random nodes form a tree network cluster without a central controller to provide a secure private storage service and faster message propagation. Advance proxy re-encryption algorithm is applied to guarantee secure file exchanges under permission. Merkle tree will make sure that the files are distributed among the service nodes in a balanced way. The proposed MTFS can be used not only for personal file storage and exchange but also for industry requiring mutual trust in file uploading and downloading in making contracts like insurances.

20.Large-Scale Location-Aware Services in Access: Hierarchical Building/Floor Classification and Location Estimation using Wi-Fi Fingerprinting Based on Deep Neural Networks

Author:Kim, KS;Wang, R;Zhong, Z;Tan, Z;Song, H;Cha, J;Lee, S

Source:2017 INTERNATIONAL WORKSHOP ON FIBER OPTICS IN ACCESS NETWORK (FOAN),2017,Vol.2017-December

Abstract:One of key technologies for future large-scale location-aware services in access is a scalable indoor localization technique. In this paper, we report preliminary results from our investigation on the use of deep neural networks (DNNs) for hierarchical building/floor classification and floor-level location estimation based on Wi-Fi fingerprinting, which we carried out as part of a feasibility study project on Xi'an Jiaotong-Liverpool University (XJTLU) Campus Information and Visitor Service System. To take into account the hierarchical nature of the building/floor classification problem, we propose a new DNN architecture based on a stacked autoencoder for the reduction of feature space dimension and a feed-forward classifier for multi-label classification with argmax functions to convert multi-label classification results into multi-class classification ones. We also describe the demonstration of a prototype DNN-based indoor localization system for floor-level location estimation using real received signal strength (RSS) data collected at one of the buildings on the XJTLU campus. The preliminary results for both building/floor classification and floor-level location estimation clearly show the strengths of DNN-based approaches, which can provide near state-of-the-art performance with less parameter tuning and higher scalability.
Total 121 results found
Copyright 2006-2020 © Xi'an Jiaotong-Liverpool University 苏ICP备07016150号-1 京公网安备 11010102002019号