CIS 632 / EEC 687 Mobile Computing

Transcription

1 CIS 632 / EEC 687 Mobile Computing MC Platform #4 USRP & GNU Radio Chansu Yu 1

2 Tutorial at IEEE DySpan Conference, 2007 Understanding the Issues in SD Cognitive Radio Jeffrey H. Reed, Charles W. Bostian, VT Software Defined Radio (SDR) Termed coined by Mitola in 1992 Radio s physical layer behavior is primarily defined in software Accepts fully programmable traffic & control information Supports broad range of frequencies, air interfaces, and application software Changes its initial configuration to satisfy user requirements 2

3 Cognitive Radio Term coined by Mitola in 1999 Mitola s definition: Software radio that is aware of its environment and its capabilities Alters its physical layer behavior Capable of following complex adaptation strategies A radio or system that senses, and is aware of, its operational environment and can dynamically and autonomously adjust its radio operating parameters accordingly Learns from previous experiences Deals with situations not planned at the initial time of design 3

4 GNU Radio Origin: MIT s SpectrumWare (mid 90s) Software toolkit for signal processing Software radio construction Rapid development Cognitive radio USRP (Universal Software Radio Peripheral) Hardware frontend for sending and receiving waveforms

5 GNU Radio Components Hardware Frontend Host Computer RF Frontend (Daugtherboard) ADC/DAC and Digital Frontend (USRP) GNU Radio Software

6 Development Architecture Python C++ Application management (e.g., GUI) Flow graph construction Non-streaming code (e.g., MAC-layer) Signal processing blocks Certain routines also coded in assembly Python Application development Flow graph construction C++ Signal processing blocks

7 Basics: Blocks Signal Processing Block Accepts 0 or more input streams Produces 0 or more output streams Source: No input noise_source, signal_source, usrp_source Sink: No outputs audio_alsa_sink, usrp_sink

8 Basics: Data Streams Blocks operate on streams of data

9 Basics: Data Types Blocks operate on certain data types char, short, int, float, complex Vectors Input Signature: Data types for input streams Output Signature: Data types for output streams Two streams of float One stream of complex

10 Basics: Flow Graph Blocks composed as a flow graph Data stream flowing from sources to sinks

11 Dial Tone Example Generates two sine waves and outputs them to the sound card Importing necessary module Generates two sine waves Writes sampling _freq input to the sound card connect the blocks together

12

13 GRC (GNU Radio Companion) * Generate the corresponding python file in GRC to see how it is programmed.

14 USRP RF Front-end

15 USRP (Universal Software Radio Peripheral) Serves as a digital baseband and IF section of a radio communication system All waveform-specific processing (e.g., modulation) is done on the host CPU. All the high speed general purpose operations (digital up/down conversion, decimation, ad interpolation) are done on the FPGA.

16 USRP

17 USRP Motherboard

18

19 USRP Daughter-boards LFTX LFRX RFX2400

20 USRP Daughterboards Base-band boards: BasicTX/BasicRX, LFTX/LFRX Bandwidth: 250 MHz (Basic), 30 MHz (LF) Can not connect directly to antenna WBX - Wide-band transceiver 50 MHz 2.2 GHz Relatively poor performance TVRX MHz receiver Bandwidth: 6 MHz Suitable for receiving TV signals RFX-series - General full-duplex transceivers Bandwidth: 30 MHz Bands: MHz, MHz, MHz, GHz, GHz

21 21

22 1. Open a terminal and type gnuradio-companion and open a new project. 2. Double click the Options block and name the project title and author. (Generate Options to WX GUI, Run to Autostart, and Realtime Scheduling to Off) Observe the default sample rate is set to samples/sec in the Variable block, which will be used in other blocks. 3. On the right side, find Signal Source from Sources category. Bring it to the main window. Double click Signal Source block and change the output type to Float and frequency to Find WXGUI Scope Sink from WXGUI Widgets category to bring it to the main window. Change the type to Float and window size (500,100). 5. Click on the out port of Signal Source and in port of Scope Sink to make a connection. 6. Save the grc file and Click Generate the flow graph icon. Notice that there is a warning message at the bottom of the window. Add Throttle block in Misc category (where/why?) in between the two. 7. Click Generate the flow graph icon again. No warning now. Then, click Execute the flow graph icon to execute it. 8. Add Audio sink block to hear the 1KHz sound.

23 9. In the scope window, change the channel option to Dot Large to observe actual samples (32000 samples/sec).how many dots are there in one cycle? 10. Change the sample rate (how?) and execute again to observe the actual samples. How low can you drop the sample rate? Recall the Nyquist sampling theorem. 11. Add FFT sink block as another sink. Change the type float and window size to (500,100) to observe the FFT plot. 12. Add another Signal Source of frequency 800. Combine the two sources using Add block. 13. Replace the Add block with Multiply block to observe and hear the differences. 14. Add Low Pass Filter to block the 1.8KHz component and pass the 200Hz component. Set the low pass filter to have a cutoff frequency of 1KHz and a transition width of 200 Hz. Use a Rectangular Window. 15. Repeat with the High Pass Filter. Choose the parameters and observe the Amplitude in FFT plot (change db/div option to easily see) 16. With the Low Pass Filter, change the Decimation to 2. A decimation factor of two means that the output of the filter will have a sample rate equal to one-half of the input sample rate, or in this case only samples/sec. 17. Observe the FFT plot to see the peak frequency, which is 400Hz (not 200Hz, why?). Double click on the FFT Sink block and change the sample rate to samp_rate/2.