Papers

Management of OCDMA auto-correlation function distorted by dispersion effects

Md S. Ahmed, M. Abuhelalad and I. Glesk
19th IEEE International Conference on Transparent Optical Networks (ICTON), Girona, Spain, 2-6 July 2017. (invited paper)

Paper

Managing dispersion affected OCDMA auto-correlation based on PS multi-wavelength code carriers using SOA

Md S. Ahmed, M. Abuhelalad and I. Glesk
IEEE/OSA Journal of Optical Communications and Networking, 9 (8). pp. 693-698, 2017

Paper

User Activity Detection in Massive Random Access: Compressed Sensing vs. Coded Slotted ALOHA

V. Boljanović, D. Vukobratović, P. Popovski and Č. Stefanović
18th IEEE International Workshop on Signal Processing Advances in Wireless Communications, Hokkaido, Japan, July 2017

Paper

Estimating heart rate and rhythm via 3D motion tracking in depth video

C. Yang, G. Cheung, and V. Stankovic
IEEE Transactions on Multimedia, vol. 19, pp. 1625-1636, July 2017.

Paper IEEE published version

Gait phase classification for in-home gait assessment

M. Ye, C. Yang, V. Stankovic, L. Stankovic, and S. Cheng
ICME-2017 IEEE International Conference on Multimedia and Expo, Hong Kong, July 2017
Finalist of the 10K Best Paper Award

Paper Award

Seminars

post img

Talk at EU NILM Workshop 2017

Lina Stankovic presented SENSIBLE aims and the latest results pertaining to Non Intrusive Load Monitoring (NILM) at the NILM EU 2017 Workshop on 6-7 November 2017 in London. The collaborative work of ESR Srdjan Lulic on Deep Learning based NILM with the Strathclyde researchers during his secondment was presented, as well as results from ESR Bochao Zhao's secondment to Tongji University in Shanghai on Graph Signal Processing based NILM. ESR Kanghang He also presented his research findings on NILM accuracy estimation. The workshop was attended by a growing European NILM community of academics, research intensive SMEs and energy industry, and provided the opportunity to network with like-minded researchers and industrial stakeholders with connections made for future collaboration.

Talks (External)
post img

Seminar on Deep Learning AI

On November 3rd, 2017, Srdjan Lucic, from PANONIT, gave a seminar to the Advanced Image & Video Processing class and other academics and researchers at Strathclyde. Srdjan first gave a brief tutorial on deep learning and its application to Computer Vision. Then, he spoke about the work he conducted at Strathclyde as part of SENSIBLE, including novel energy disaggregation methods using convolutional neural networks (CNN). Srdjan presented the results obtained with the proposed CNN-based method that are comparable with the state of the art for the energy dissagregation application, that use more complex deep learning architectures. After the presentation, Srdjan answered numerous questions from staff and students.


post img

Invited Seminar at 26th International Electrotechnical and Computer Science Conference ERK 2017

On 25th September, 2017, Dr Dejan Vukobratovic, Associate Professor at University of Novi Sad, gave a seminar at 26th International Electrotechnical and Computer Science Conference ERK 2017, organized by IEEE Slovenia Section and Faculty of Electrical Engineering University of Ljubljana. The talk presented SENSIBLE project and FTN’s contribution on Machine-type Communications (MTC) and Mobile Edge Computing (MEC) technologies for future Internet of Things (IoT) services in 5G.

Talks (External)
post img

Internet of Things in 5G: From massive machine-type communications to large-scale information processing

On 8th August, 2017, Dr Dejan Vukobratovic, Associate Professor at University of Novi Sad, gave a talk entitled “Internet of Things in 5G: From massive machine-type communications to large-scale information processing” at University of Strathclyde. In the talk, Dejan discussed evolution of Machine-type Communications (MTC) and Mobile Edge Computing (MEC) technologies for future Internet of Things (IoT) services in 5G, focusing mainly on research on massive MTC and large-scale distributed information processing conducted in his research group. The talk also covered how these technologies will be used for future large scale monitoring of critical infrastructures.


post img

Distributed optimization methods with variable number of working nodes

On August 2nd, 2017, Dr Dusan Jakovetic, Assistant Professor from Dept Mathematics and Informatics, University of Novi Sad, gave a seminar entitled “Distributed optimization methods with variable number of working nodes” at the University of Strathclyde, Glasgow. The seminar was attended by university’s staff and students from three different departments - Electronic and Electrical Engineering, Computer and Information Systems, and Mathematics, as well as SENSIBLE visiting researchers. Dusan presented his latest work on distributed optimization problems that often arise in various applications, including inference tasks in sensor networks and smart grid, coordination and control of multi-robot networks, as well as machine learning in distributed environments.


post img

Seminar during secondment at Tongji University

During their secondment at Tongji University in July 2017, Shanghai, Drs Lina and Vladimir Stankovic with their PhD students Bochao Zhao and Minxiang Ye gave a range of seminars on the topics of Internet of Things (IoT) for smart buildings, specifically focusing on energy efficiency sensing and load disaggregation algorithms as well as health and motion assessment. Our host, Dr Samuel Cheng and his PhD students also gave seminars on Deep Learning Networks and Hypergraphs.
During their time at the lab, the seconded researchers jointly set up, trained and demonstrated the use of aerial drones for mapping terrain, EV3 swarm robots and the NAO interactive humanoid robot that can learn human motion and respond to simple commands. All these devices are equipped with a multitude of sensors, motors and communications capapbility.


post img

Seminar at the Institute of Aurel Stodola, Liptovsky Mikulas, Slovakia

On July 11th 2017, Dr Helena Gleskova and Prof Ivan Glesk presented the work of their groups on the development of low-voltage organic transistors for wearable electronics and the application of the transistors to temperature and pressure sensors based on ferro/pyroelectric polymer P(VDF-TrFE). Both presentations were part of a half-day seminar organized by Prof. Mullerova from the Institute of Aurel Stodola of University of Zilina in Liptovsky Milulas. The third presenter of the day was Dr Pavel Cheben from the National Research Council in Ottawa, Canada.



post img

Scene Reconstruction

On June 22nd 2017, Dr Cheng Yang from NII visited University of Strathclyde and gave a seminar on his latest work on scene reconstruction to SENSIBLE team from University of Strathclyde and University of Novi Sad.






post img

Detection of Random Walks on Graphs with Heterogeneous Sensors

Dr Dragana Bajovic, Assistant Professor from University of Novi Sad, gave a seminar entitled “Detection of Random Walks on Graphs with Heterogeneous Sensors,“ at the University of Strathclyde, Glasgow on June 14th 2017. Dragana presented her latest results on the lower bounds of the error exponent, efficient ways of computing the bound via convex optimization, and highlighted a range of possible applications, to staff and research students from three different departments - Electronic and Electrical Engineering, Computer and Information Systems, and Mathematics.


post img

H2020 SENSIBLE project: Big Data Analytics for Intelligent Buildings

On 1 June 2017, Dr Dejan Vukobratovic (University of Novi Sad) and David Murray (University of Strathclyde) gave a talk at the The Special Interest Group Mathematics for Big Data, on the SENSIBLE project objectives, challenges for big data analytics for SENSIBLE and examples of algorithms and data analytics tools used by the SENSIBLE team.

Talks (External)
post img

Low-voltage organic thin-film transistors and sensors

On 9 May 2017, Dr Helena Gleskova (University of Strathclyde) gave a seminar at McGill University in Canada, on the topic of wearable sensors. Wearable sensors are likely to be powered by batteries or energy-harvesting devices. Consequently, the sensors need to operate at low voltages. Gate dielectrics with sub-10-nm thickness can provide organic transistors with operating voltage as low as 1.5 V. The transistor can serve as the first-stage signal amplifier and a converter of the targeted external stimulus to voltage. This talk summarized the steps and advanced in the fabrication of such low-voltage organic transistors and presented recent results of pressure and temperature sensors based on organic transistors. which examined the key disruptive milestones which led to a transformation of IT into ICE, then discussed the resulting challenges, looked at possible solutions and showed why advanced all-optical devices may help to enable more scalable solutions via advanced optical code-division multiple-access approaches.


post img

IEEE Local Chapter seminar at McGill: “From IT to ICE, the resulting challenges and the search for solutions”

On 2 May 2017, Prof Ivan Glesk (University of Strathclyde) gave a talk, which examined the key disruptive milestones which led to a transformation of IT into ICE, then discussed the resulting challenges, looked at possible solutions and showed why advanced all-optical devices may help to enable more scalable solutions via advanced optical code-division multiple-access approaches.


Wireless Networks for Industrial Internet

Branka Vucetic, University of Sydney
Seminar FTN, University of Novi Sad, Feburary 2017

Talk (External)