AIT Open Cellular Training Center

//AIT Open Cellular Training Center
AIT Open Cellular Training Center2019-04-23T02:11:15+00:00

Project Description

Project Title: “AIT Open Cellular Training Center, Thailand”

  • Provides the development of Open Cellular Software Defined Radio (SDR) device (Wireless Access Platform Design))
  • Project sponsored by Telecom Infra
  • Project Led by Facebook

Project Head

  • Full Name: Attaphongse Taparugssanagorn
  • Department: Information and Communication Technologies (ICT)
  • Designation: Full-time faculty/associate Professor
  • Achievement: (Dr. Tech)/University of Oulu, Finland.

Main Focus

This project research is based on using the following fundamental knowledges: signal processing, statistical signal processing, namely estimation and detection techniques, communications theory, and artificial intelligence to solve various problems.

With the introduction of Digital Television (DTV), the switchover from analog to digital broadcasting began around 2006 in some countries, and many countries have now completed the switchover, while other countries are in various stages of switchover. Due to this switchover some frequencies originally allocated for TV broadcasting are not utilized. These unused frequencies are called TV White Spaces (TVWS). TVWS can take advantage of the improved propagation capabilities to provide internet access in rural areas. Studies shows the existence of white spaces in densely populated areas and these can be used for other wireless communication services. Different DTV broadcasting standards have been adopted in different parts of the world e.g., Digital Video Broadcasting (DVB), Advanced Television System Committee (ATSC), Integrated Services Digital Broadcasting (ISDB), Digital Terrestrial Multimedia Broadcasting (DTMB) and Digital Multimedia Broadcasting (DMB).

In Thailand, the National Broadcasting and Telecommunications Commission (NBTC) is playing an important role in promoting and implementing the transition from analogue to Digital Terrestrial Television (DTT). In 2012, the transition road map was developed and Digital Video Broadcasting Second Generation Terrestrial (DVB-T2) was selected as a national standard for DTV. With the recent transition from analog to digital television, there are new free spectrum which was once the preserve for TV broadcasters in Thailand. This freed up new spectrum are called TVWS as mentioned before.

Many researchers around the world have already recognized the potential benefits of this newly available frequencies. The TVWS has a lot of potentials to provide the affordable broadband access in the rural areas due to its excellent propagation characteristics and abundance in rural and developing regions. As recommended by the Federal Communications Commission (FCC), the TVWS can be used by the secondary users (unlicensed) when they are not being used by the licensed users (e.g., national TV broadcasting). The traditional spectrum measurement devices are quite expensive and inconvenient for field measurements due to their size and weight. To enable cost-effective, realistic and widespread assessment of TVWS spectrum availability, the low-cost spectrum analyzers that are also easier to operate have been recently proposed in many research projects.

The Cognitive Radio (CR) devices are required to perform sensing decision as quickly as possible. With the concept of compressive sensing theory, it is possible to reconstruct the sparse signal even with the sampling rate quite lower than the Nyquist-sampling rate. This concept helps CR device to quickly acquire the signal samples and make decision whether the channel is busy or idle quickly which is not possible in conventional sensing approaches. Using the samples obtained via compressive sensing we then use supervised machine learning and deep learning algorithms, i.e., Support Vector Machine (SVM), Convolutional Neutral Network (CNN), Recurrent Neural Network (RNN), Long-Short Term Memory (LSTM), just to name a few. The huge amount of data available via measurement is the main motivation of using such algorithms with some important features, including received power, Max, Min, Peak-to-Average Power Ratio (PAPR), etc.

Objective and Aim of Project

The main objective of the project is to presents major results of radio spectrum utilization measurements in TV band that have been carried out in urban area in the city of Bangkok, suburban area in the city of Pathum Thani, and rural area in the Mae Kasa, Tak province in the Thailand. The contribution of this project are as follows:

  • To propose a novel CR architecture that is aimed at wideband operation in alien RF environments. The proposed model can sense a wide frequency band of interest and detect the ongoing RF activities, without any prior knowledge about the active signals, and with less sensing time. This is achieved using a compressive sensing theory and the mentioned machine learning and deep learning techniques.
  • To conduct spectrum measurement of TV band via a low-cost spectrum analyzer in various location in Thailand.
  • To determine spectrum occupancy using the proposed improvement of the adaptive threshold setting techniques to form a knowledge base for the CR engine.
  • To determine the sparsity of TV signal in UHF band all three areas.

This project was funded by one-year project “TV White Space” sponsored by the National Broadcasting and Telecommunications Commission (NBTC).

For this project, teamwork was very essential. The team consisted of Dr. Adisorn Lertsinsrubtavee, Miss Nisarat Tansakul, and the teams from the Internet Education and Research Laboratory (intERLab), Mr. Bipun Man Pati, my PhD student from Nepal, and myself. Whereas, the project head was Prof. Kanchana Kanchanasut.

Limitations

However, there are some limitations and risks involved in this project, such as:

  • Given the low spectrum utilization, most of existing wideband sensing techniques assumed that the wideband signal is sparse in the frequency domain, i.e., the sparsity basis is a Fourier matrix. The use of compressive sensing when the signal is not sparse is not possible.
  • We consider only the UHF TV band for our analysis if the signal is sparse in frequency domain. But as the spectrum utilization increases, we cannot consider the TV signal sparse in frequency domain and hence we cannot use compressive sensing techniques. Thus, novel techniques in compressive sensing when the signal is not sparse need to be carried out.

Dr. Attaphongse summarized TV White Space (TVWS) in a rap rhyme as he believes that it is the easiest and a fun way to understand the concept. The rap goes like:

The unused TV channels, called “TV White Space”

Between the active ones in the VHF and UHF.

Can be used to provide broadband Internet access.

With surrounding TV channels, it is harmoniously operated.

 

Better coverage than 100 m limited range of Wi-Fi.

One hundred times the distance, it should be fine.

Superior range and ability to penetrate obstacles.

Trees, buildings, and rough terrain, everything is possible.

 

A new broadband connection in rural communities.

Function as a backhaul to networks with low population density.

Utilize less infra to provide further coverage saving customers money.

An option for densely populated areas is also from this technology.