Fundamentals of Networking Lab Seattle Spectrum Measurements MHz
Fundamentals of Networking Lab Purpose for the measurements Sweeping thru a wideband (bandwidth>900MHz), and use this wideband data to validate our wideband detection architectures and algorithms for cognitive radio Monitor a set of channels continuously to investigate the temporal correlation for channel modeling
Fundamentals of Networking Lab Hardware settings
Fundamentals of Networking Lab System block diagram
Fundamentals of Networking Lab USRP2 RF Chain (1) 50MHz-2GHz (2) down converted to a center frequency of 126MHz (3) MHz (4) baseband: 0- 1MHz
Fundamentals of Networking Lab Data Collection Parameters ParameterValue Frequency Range MHz Channel Bandwidth8 MHz Number of Channels6 Sample Rate16.67 MS/s Samples per Channel2.048x10^9 Collection Times20 Dec 11 15:00, 21 Dec 11 00:01, 21 Dec 11 07:00, 21 Dec 11 12:00 Total Data Recorded3.93x10^11 Bytes 6
Fundamentals of Networking Lab Example Collection 7
Fundamentals of Networking Lab Example Collection 8
Fundamentals of Networking Lab Processing Steps Initial 832 point FFT and magnitude squared Receiver Characteristic Compensation Harmonic and DC Clipping Result – 2000x416 matrix –Rows are time –Columns are frequency 9
Fundamentals of Networking Lab Processing Diagram 10
Fundamentals of Networking Lab Receiver Compensation 11 Left half multiplied by: Right Half multiplied by:
Fundamentals of Networking Lab Harmonic Clipping 12 -Harmonics associates with the USRP are clipped to the noise floor
Fundamentals of Networking Lab Channel Four After Processing 13
Fundamentals of Networking Lab Channel Five After Processing 14
Fundamentals of Networking Lab Summary Table 15 PropertyValue B/W per channel MHz Channels per recording time6 Total recording times4 Total files24 Time between rows of the matrix seconds Total time per channel seconds Total time sweeps2000 FFT size416 Noise Floor1.047E-5 Frequency Per Bin20KHz
Fundamentals of Networking Lab Channel Occupancy Probability Estimate the noise floor Calculate threshold for occupation Compare 2000x416 with threshold value –one for occupied and zero otherwise Sum over 2000 rows and divide by 2000 Result – matrix representing P(i, j) Repeat for different channel and subchannel sizes 16
Fundamentals of Networking Lab Noise Floor Channel three is unoccupied Sum all of the values in channel three Divide by number of values Result 1.047x10^-5 17
Fundamentals of Networking Lab Threshold Calculation 18 [1]
Fundamentals of Networking Lab Pij Across The Spectrum Probabilities of occupancy line up well with FCC spectrum allocation Busy channels occur at paging bands MHz Additional activity from MHz (Business radios 19
Fundamentals of Networking Lab Results Most activity recorded on Paging bands These bands are and MHz 20
Fundamentals of Networking Lab , , Bands support paging at UW and Seattle area Services the UW medical center Use 4-FSK at 6400 bps Flex Protocol – standard protocol for paging in the U.S. Transmission power is 2000 Watts 21
Fundamentals of Networking Lab Unidentified Paging Signals Unidentified Signals: – – – – Exact locations inferred from FCC registration data 22
Fundamentals of Networking Lab Calibration 23
Fundamentals of Networking Lab USRP Noise Figure Plot 24 [2]
Fundamentals of Networking Lab Calibration Results New noise level about 2 dB off -174 dBm 25
Fundamentals of Networking Lab References 1.Danijela Cabric, Artem Tkachenko, and Robert W. Brodersen Experimental study of spectrum sensing based on energy detection and network cooperation. In Proceedings of the first international workshop on Technology and policy for accessing spectrum (TAPAS '06). ACM, New York, NY, USA,, Article 12. DOI= /