Msc Project- Dvb :interim Report

School of Engineering & Design Electronic and Computer Engineering (ECE)

MSc Project: Implementation of Digital Communication System on a Texas Instruments DSP

Project Interim Report MSc Programme Ivan Cheung

Msc Project Interim Report

Contents Page 1. Introduction

3

2. Background to the Project

4

3. Initial Survey

7

4. Aims and Objectives

9

5. Experimental/Investigative methods to be adopted

10

6. Time-Plan

12

7. Deliverables or specific Outcomes

13

8. Bibliography

14

2

Msc Project Interim Report

Introduction This project will explore how a Digital Signal Processor (DSP) is used to compute the signal processing used in mainstream communication systems especially wireless systems. It will show just why DSPs and other embedded processors such as FPGAs are extremely important in playing the role of the workhorse, shaping the ever increasing speed of growth of wireless communication technology. Most of the wireless communications today are becoming more advanced, and is maturing at an exponential rate. The electronics industry in general is growing fast, and because it is aiding the development of wireless communications technology since now it’s the hot topic, wireless is growing even faster. Mobile phones for example were once just used for voice communication, nothing else. Now there are plenty more media applications such as video messaging and internet browsing, and the word ‘communications’ has gained a new definition. The graphical user interface on the phone screens has become much more user friendly, a lot like Windows on a PC. Even though the processing power of mobile phones is yet to reach the standards of PDAs and PCs, the day will come soon. If not for DSPs, mobile phones and PDAs just wouldn’t have enough processing power for this luxury. This project will essentially be a demonstration on how the hardware (more specifically speaking it should be firmware) for a basic Digital Video Broadcast (DVB) transmitter/receiver system is implemented onto DSP. It will involve standard modulation and demodulation techniques and will demonstrate just how quick designing and implementing is with using embedded technology such as a DSP. This topic of DVB on DSP was selected because it combines knowledge gained from the student’s first degree, “Electronic and Microelectronic Engineering” and the present Masters degree, “Wireless Communications”, both obtained from Brunel University.

3

Msc Project Interim Report

Background to the Project When initiating such a project it’s always recommended that there is some background research done. Although the nature of this project involves just the implementation and evaluation of the DVB system, a bit of research will fill in the gaps of knowledge and there will be a better understanding on how the design functions and how to deal with the problems which arise during the project cycle. Note this was not meant to be a research topic project similar to a phD thesis, and much of the implementation revolves around utilising standard codes already designed which are available from the DSP software packages and the internet, and then tweaking with them for providing the output performance needed for this project. This background section of the Interim report will provide a more detailed review of the technical field, and will be accompanied by a quick preliminary initial survey. The technical fields discussed here will soon be converted to survey questions used for the identifying the information and knowledge required to successfully complete the making and testing. That main survey will be there to aid the whole project cycle within its duration, and help guide the direction so that it won’t steer off to the wrong course. The most important aspect of a project progress is making the correct decisions to what should happen next. This is a project which involves writing codes in a high level language for the functionality of the DVB system, simulating it with a software package, and when all is verified, it will then be downloaded onto the DSP from the software application on the PC via a programming cable. The DSP integrated circuit is situated in a printed circuit board with other built in peripherals such as memory, output components, and interfacing, known as a development board. This sort of design process is different to the traditional ASIC (Application Specific Integrated circuit) approach where the time it takes to implement is much faster, but won’t consider the minute timing issues which will optimise the performance of the system. This is all about working in the Physical layer of communications design, the actual electronic components. The fact that the necessary needs and fundamental sequence of events is already laid out means all that’s left is to choose; what type of DSP and kit is most appropriate to work with, and which codes to use. Knowledge gained from the two engineering degrees is important. During the last degree there has been experience with designing embedded systems on microcontroller and FPGA boards which will help with the grasping of how to use this DSP. In the present degree much has been learnt on modulation techniques on the Physical layer for wireless transmission/reception such as QAM (Quadrature Amplitude Modulation). Now is the time to take all the theory and make the real thing. The following paragraph is the description of the functionality of this system given as the project specifications from the project supervisor. There won’t be specifically detailed descriptions on each of the components’ functionalities. The design blocks specified by the supervisor are standard blocks already used commercially in many modern devices, but the responsibility of the DSP and other

4

Msc Project Interim Report major components and which blocks they encompass is to be decided. Basically, the method of modulation used will be QAM, where the binary input (working at baseband frequency) will be converted into not one but 2 analogue signals I and Q, which are used for modulating in respect to varying the phase and amplitude to differentiate the different n bit input codes fed into the modulator. I and Q will be fed into a Digital Up Converter (DUC), which produces a single output and fed to an additional Digital Up Mixer before transmission. The signal at output is digital and represents a sine wave with the phase and amplitude modulating (working at passband or IF frequency). When receiving, the signal is down mixed, down converted, and converted back into the original binary bits. At the moment there are standard already built boxes which conduct the DUC/DDC and Up/Down mixing, used to simulate the transmission and reception, hence it has been suggested that the DSP should implement the QAM conversions as shown in this diagram, although it’ll be desirable if the DSP was able to incorporate the whole system, this has never been attempted before. DSP

Binary input

Binary output

I DUC

Digital Up mixer

DDC

Digital Down mixer

Q

I Q

TX

RX

In reality, the signals sent through the air are analogue, hence it’ll be desirable if a Digital to Analogue and Analogue to Digital converters be used before transmission and after reception, shown in the next diagram. This was discussed as the possible extension to the project.

5

Msc Project Interim Report DSP Binary input

Binary output

I DUC

Digital Up mixer

DAC

DDC

Digital Down mixer

ADC

Q

I Q

TX

RX

INCLUDEPICTURE "http://tbn0.google.com/images?q=tbn:

Agilent oscilloscope picture sourced from http://www.triosmartcal.com.au/product_info.php?products_id=1050

When the code has been downloaded onto the DSP board, I and Q wires have connected the DSP board to the DUC/DDC box, and the Transmitter port connected to the Receiver port of the box (to simulate a wireless channel), then it is ready for the main evaluation. Oscilloscopes and other testing equipment from Agilent Technologies have been chosen as they advertised their products when they came for a seminar, and the equipment are readily available in the labs of the university. To evaluate the performance of the system, the oscilloscopes and spectrum analysers will tap off from the I and Q inputs/outputs and from the transmission cable, for observing the actual waveforms and the frequency responses. Comparing the randomly generated binary inputs with the outputs will show if the bits have transmitted with no errors, but comparing the frequency response waveforms with the theory suggested will also indicate if it is functioning properly. So what are the technical fields of knowledge required to develop this DVB transmitter/receiver? To summarise, the technical fields described in the above on implementing and evaluating are: How DSPs work, How the selected development board works Knowledge on DVB and QAM modulation theory How to use the software packages How to use the evaluation equipment

6

Msc Project Interim Report

Initial Survey In the main survey, the set of questions thought out are simple and useful for; 1. identifying the key abstracts, journals, books, reports, etc needed for research 2. What needs to be purchased 3. setting up the timing of events, which will aid in producing the time-plan section shown later on 4. being another tool to keep the project progress on track. This initial survey will give kick start to what’s expected of the project. It will only have questions which cover the beginnings of the project life, although it has the potential of keeping the timing on track right until evaluation, because its focus is to plan what resources are used, and when they are used. 1. What is the decided development kit? Texas Instruments TMS320C6416T DSP Starter Kit (DSK). Inside the package there is; 1. The TMS320C6416T DSP development board and it’s necessary cables 2. Texas Instruments Code Composer Studio v3.3 3. Matlab & Simulink 30 day trial for use with TI DSPs 4. Manuals on the DSK, and how to use the Code Composer Studio. 2. What are the topics to research about? This project covers many areas of electronics engineering and wireless communications, and depending on the student’s background knowledge on QAM modulation and DSPs, this may mean much research to do. Initially this project was open for any suggestions in an appropriate wireless technology, whether it be 3G, WiMAX, WiFi or DVB. After discussions and some research it was found that the algorithms for WiMAX and 3G are much more challenging to implement than DVB or WiFi. Due to the time limit, it would be best for DVB to be used here. Hence the research field narrowed to just concentrating on DVB. The research topics that help with understanding how the certain parts of the system function are; How to convert a normal binary signal into I (In-Phase) and Q (Quadrature) signals Theories on QAM modulation more specific Digital Up/Down converting interpolation, How to extract the frequency domain response of QAM and in the special case of QPSK, How to implement QAM Physical layer, What is Digital mixing, more specific on multipliers and CORDIC (COordinate Rotation DIgital Computer)

Analogue to Digital and vice versa circuitry for modulation. The research topics that help with the understanding of how to implement/evaluate the project are:

7

Msc Project Interim Report How to use the Code Composer Studio v3.3 (provided in the starter kit) Revision of C programming How to use the Agilent evaluation equipment 3. Are there any books/journals that are useful to the project? A few books obtained from the past degree about general signal processing are useful for their general modulation theories: “Signals & Systems” second edition – Alan V. Oppenheim and Alan S. Willsky, Prentice Hall Signal Processing Series ISBN – 0-13-651175-9 “Modern Digital and Analogue Communication Systems” third editionB.P. Lathi ISBN – 0-19-511009-9 One book is quite useful for wireless theories: “Wireless Communications- Principles and Practice” second editionTheodore S. Rappaport ISBN- 0-13-042232-0 This is the book recommended by the supervisor. It accompanies with the kit entirely because it teaches how to develop signal processing on this specific Texas Instruments DSK (Development Starter Kit) which has that specific DSP model: “Real-Time Digital Signal Processing, from Matlab to C with the TMS320c6x DSK” Thad B. Welch, Cameron H.G. Wright, Michael G. Morrow. ISBN- 0-8493-7382-4 4. How long will it take to master the Code Composer studio? With most unfamiliar software there will always be some time in the beginning to familiarise oneself with it. This would mean following the manual provided, or a tutorial provided by the software. Due to the fact the student has had experience with programming with C, the learning curve would not be too steep. With a few hours of concentrated tutorial learning this can be achieved. 5. What extra components will be bought, and when? At the moment there had been discussions into buying a daughter card which fits onto the development board called DSK6XXXHPI. This will be the ADC/DAC interface and will provide the parallel port interfacing with a PC. Due to the fact shipping might be delayed this order should be placed as soon as possible, maybe even before the project cycle begins.

6. What is the budget for purchasing additional components? At the moment the Starter Kit is borrowed from the supervisor. The rules state that for every student the budget is £100, but this is flexible with the supervisor.

8

Msc Project Interim Report

Aims and Objectives The aims of this project are; To get a greater understanding of how DSPs are used as communication workforces, and get a feel of how to build and test a DVB system. This task will enhance the student’s knowledge and expertise in implementing wireless systems on DSP, which will refresh the knowledge gained from the first degree. The Objectives of this project are; To produce a transmitter/Receiver System which is able to successfully transmit and receive a randomly generated bit sequence. If there are bit errors when comparing the input to the output, the code should be modified so that it can deal with timing errors. The nature and source of errors should be fully understood. The output of the system should produce an appropriate frequency response curve when analysed with evaluation equipment.

9

Msc Project Interim Report

Experimental/Investigative methods to be adopted This is the list of all the key activities necessary to complete the project, categorising in which are for Research, Purchasing, Implementation + Evaluation, and within that which are experimental methods /investigative techniques. 1. Report done throughout duration Research 2. 3. 4. 5. 6.

Learn how DVB worksinvestigative technique Learn the Architecture of the chosen DSP- investigative technique Reading manuals that come from kitinvestigative technique Online reading investigative technique Mastering ISE: Mastering Code Composer Studio from TI Experimental method 7. Small research on what the structure of the code for implementing 3G will be like investigative technique 8. Recap on modulation and how to retrieve QPSK frequency domain graph investigative technique 9. Lecture notes investigative technique 10. Research and choose possible ADC from TI which will suit this board. investigative technique 11. How are they connected together? investigative technique Purchasing 12. Get chosen DSP Kit investigative technique 13. Get any outstanding equipment needed for the board by going to the Spectrum Digital website investigative technique 14. Get chosen TI ADC

experimental technique

Implementation + Evaluation 1) Preliminary Code Composer familiarisation experimental technique Write simple code given from help example (can be anything basic, flashing lights etc) and test it Write simple modulation code from help example and test it. 2) Preliminary board familiarisation experimental technique Set up equipment Test simple modulation code from help example with board and equipment

10

Msc Project Interim Report 3) Main Retrieving the C code experimental/investigative technique Understand the standard codes from supervisor for implementing DVB (harder option due to translation from Matlab to C?) OR Get the codes from web for implementing a 3G system on DSP from web OR Codes from TI’s website? Making sense of the code + Testing

experimental/investigative technique

Re-arranging and re-writing code i.e. enhancement + Testing experimental/investigative technique

11

Msc Project Interim Report

Time-Plan The time plan is basically all of the activities described above but translated onto Gantt chart in Microsoft Project, I won’t involve all the activities until the final submission in September e.g. dissertation writing, but it will concentrate on the activities taken from the initial survey:

Unfortunately due to the lack of time to complete this interim report due to deadlines of other assignments, the timings of these events were estimated inaccurately. However it is to show that alot of research activity will commence in the Easter holidays of 2008. 12

Msc Project Interim Report

Deliverables or specific Outcomes To summarise, this is an observation type of project. The deliverables are mainly to do with the simulation results on the software application, and the output results of the DSP in hardware. There will be a series of tests where randomly generated binary inputs are fed to the system and then the output at the receiver will track the bits that are received correctly and incorrectly. The Bit Error Rate calculations are an important value to show the efficiency of the system. The graphs recorded either from the PC or oscilloscopes during the evaluation phase will be the main deliverables, and will be screen dumped or captured on photo to be put in the main project dissertation. The whole point of this experiment is to show that modulation circuitry of the Physical layer could be implemented on embedded technology easily, it’s not something that hasn’t been achieved before, hence as mentioned before it’s not similar to a phD type of research project. However there is a possible interesting outcome to look out for, and that is whether implementing the whole system meaning the modulation part as well as the up-conversion is possible, because this has never been achieved before.

13

Msc Project Interim Report

Bibliography Websites used for this Interim report: http://www.triosmartcal.com.au/product_info.php?products_id=1050

No actual books or journals were used to write this interim report, however it’ll be useful to indicate what books will be used in the progress of this project. “Signals & Systems” second edition – Alan V. Oppenheim and Alan S. Willsky, Prentice Hall Signal Processing Series ISBN – 0-13-651175-9 “Modern Digital and Analogue Communication Systems” third editionB.P. Lathi ISBN – 0-19-511009-9 One book is quite useful for wireless theories: “Wireless Communications- Principles and Practice” second editionTheodore S. Rappaport ISBN- 0-13-042232-0 This is the book recommended by the supervisor. It accompanies with the kit entirely because it teaches how to develop signal processing on this specific Texas Instruments DSK (Development Starter Kit) which has that specific DSP model: “Real-Time Digital Signal Processing, from Matlab to C with the TMS320c6x DSK” Thad B. Welch, Cameron H.G. Wright, Michael G. Morrow. ISBN- 0-8493-7382-4

14