Ghassan Dahman / Fredrik Tufvesson Department of Electrical and Information Technology Lund University, Sweden Channel Modelling – ETIM10 Lecture no: 2013-02-11 Fredrik Tufvesson - ETIN10 1 7 Directional channel models and Channel sounding
Ghassan Dahman / Fredrik TufvessonDepartment of Electrical and Information Technology
Lund University, Sweden
Channel Modelling – ETIM10Lecture no:
2013-02-11 Fredrik Tufvesson - ETIN10 1
7
Directional channel modelsand
Channel sounding
2013-02-11 Fredrik Tufvesson - ETIN10 2
Why directional channel models?
• The spatial domain can be used to increase the spectral efficiency of the system– Smart antennas– MIMO systems
• Need to know directional properties– How many significant reflection points?– Which directions?
• Model independent on specific antenna pattern
2013-02-11 Fredrik Tufvesson - ETIN10 3
Double directional impulse response
2013-02-11 Fredrik Tufvesson - ETIN10 4
Physical interpretation
l
2013-02-11 Fredrik Tufvesson - ETIN10 5
Angular spread
l
double directional delay power spectrum
angular delay power spectrum
angular power spectrum
power
E s , , , s , , , P s , , ,
DDDPS , , P s , , , d
ADPS , DDDPS , , GMS d
APS APDS , d
P APS d
2013-02-11 Fredrik Tufvesson - ETIN10 6
Directional models
• The double directional delay power spectrum is sometimes factorized w.r.t. DoD, DoA and delay.
• Often in reality there are groups of scatterers with similar DoD and DoA – clusters
DDDPS , , APSBS APSMS PDP
DDDPS , ,k
P kc APS k
c,BS APSkc,MS PDP k
c
2013-02-11 Fredrik Tufvesson - ETIN10 7
Angular dispersion
• At the base station the angular spread is often modeled as Laplacian
• Typical rms angular spread:– Indoor office: 10-20 deg– Industrial: 20-30 deg– Microcell 5-20 deg LOS, 10-40 deg NLOS– Rural: 1-5 deg
0( ) exp( 2 )APSS
2013-02-11 Fredrik Tufvesson - ETIN10 8
Laplacian distribution, example
• Angular spreads 5, 10, 20, 40 degrees
2013-02-11 Fredrik Tufvesson - ETIN10 9
Geometry-based stochastic channel models
• Assign positions for scatterers according to given distributions
• Derive impulse response given the scatterers and distributions for the signal properties.
• Used in the COST 259model, the COST 273model the 3GPP spatial channel model, and the WINNER model
10
Geometry-Based Stochastic Channel Model (GSCM)
BS
MS 1
Cluster
Local cluster
Local cluster
Cluster
MS 2
• Create an ”imaginary” map for radio wave scatterers (clusters)
2013-02-11 Fredrik Tufvesson - ETIN10
Courtesey: K. Haneda, Aalto Uni.
2013-02-11 Fredrik Tufvesson - ETIN10 11
Properties of channel models used for Beyond-3G (B3G) MIMO simulations
Copied from: Narandzic, M.; Schneider, C.; Thoma, R.; Jamsa, T.; Kyosti, P.; Xiongwen Zhao; , "Comparison of SCM, SCME, and WINNER Channel Models," Vehicular Technology Conference, 2007. VTC2007-Spring. IEEE 65th , vol., no., pp.413-417, 22-25 April 2007
• 3GPP Spatial Channel Model (SCM)
• SCM extension (SCME)
• WINNER I , and WINNER II
2013-02-11 Fredrik Tufvesson - ETIN10 12
Properties of channel models used for Beyond-3G (B3G) MIMO simulations
Copied from: Narandzic, M.; Schneider, C.; Thoma, R.; Jamsa, T.; Kyosti, P.; Xiongwen Zhao; , "Comparison of SCM, SCME, and WINNER Channel Models," Vehicular Technology Conference, 2007. VTC2007-Spring. IEEE 65th , vol., no., pp.413-417, 22-25 April 2007
• 3GPP Spatial Channel Model (SCM)
• SCM extension (SCME)
• WINNER I , and WINNER II
2013-02-11 Fredrik Tufvesson - ETIN10 13
Considered scenarios (WINNER II)
A1 – Indoor officeA2 – Indoor to outdoorB1 – Urban micro-cellB2 – Bad Urban micro-cellB3 – Indoor hotspotB4 – Outdoor to indoorB5 – Stationary FeederC1 – Suburban macro-cellC2 – Urban macro-cellC3 – Bad urban macro-cellC4 – Urban macro outdoor to indoorD1 – Rural macro-cellD2 – Moving networks
2013-02-11 Fredrik Tufvesson - ETIN10 14
Channel measurements
In order to model the channel behavior we need to measureits properties
– Time domain measurements• impulse sounder• correlative sounder
– Frequency domain measurements• Vector network analyzer
• Directional measurements– directional antennas– real antenna arrays– multiplexed arrays– virtual arrays
2013-02-11 Fredrik Tufvesson - ETIN10 15
Working principle
Courtesy of MEDAV
Requirement for efficient measurements:• Large bandwidth (high delay resolution)• Large time bandwidth TW product (i.e., (signal duration × (1/W))>1), to achieve high SNR by
transmitting more energy even if the allowed transmit power is limited) • Signal duration should be adaptive to channel properties (e.g., TW product vs. coherence time)• Power spectral density should be uniform across the bandwidth to have same quality for the
channel estimation• Low Crest Factor (Peak to Average ratio)• Good correlation properties
2013-02-11 Fredrik Tufvesson - ETIN10 16
Can the channel be measured in a unique way?
• A channel that fulfills these requirements is known as underspread.• Furtionately, the majority of wirless channels are underspread (i.e., slow
time variant channels)
1/2
1
2013-02-11 Fredrik Tufvesson - ETIN10 17
Impulse sounder
impulse responseof sounder impulse response
of channel
• Comparable with the impulse radar
• Sends a sequence of short pulses to achieve good spatial resolution
2013-02-11 Fredrik Tufvesson - ETIN10 18
Correlative sounder
• Transmit a pseudo-noise sequence and correlate with the same sequence at the receiver– Compare conventional CDMA systems– Correlation peak for each delayed multipath component
p( )
Tc-Tc
ˆ( )h( )h
correlation peak impulse response measured impulse response
2013-02-11 Fredrik Tufvesson - ETIN10 19
Frequency domain measurements
• Use a vector network analyzer or similar (sounders) are used to determine the transfer function of the channel
• Time domain properties via FFT• Using a large frequency band it is possible to get good time
resolution
( ) ( )* ( )* ( )meas TXantenna channel RXantennaH f H f H f H f
2013-02-11 Fredrik Tufvesson - ETIN10 20
Channel sounding – directional antenna
• Measure one impulse response for each antenna orientation
2013-02-11 Fredrik Tufvesson - ETIN10 21
Channel sounding – antenna array
• Measure one impulse response for each antenna element
• Ambiguity with linear array
d d d
h( ) h( ) h( )
spatially resolved impulse response
Signal processing
linear array
x=0 x=d x=2d x=(M-1)d
h( )
2013-02-11 Fredrik Tufvesson - ETIN10 22
Real, multiplexed, and virtual arrays
• Real array: simultaneous measurement at all antenna elements
• Multiplexed array: short time intervals between measurements at different elements
• Virtual array: long delay
RX RX
RX
RX
RX
Digital Signal Processing
Digital Signal Processing
Digital Signal Processing
2013-02-11 Fredrik Tufvesson - ETIN10 23
Directional analysis
• The DoA can, e.g., be estimated by correlating the received signals with steering vectors.
• An element spacing of d=5.8 cm and an angle of arrival of =20 degrees gives a time delay of 6.6·10-11 s between neighboring elements
d
d sin
a
1exp jk0dcos
exp j2k0dcos
exp j M 1 k 0d cos
2013-02-11 Fredrik Tufvesson - ETIN10 24
High resolution algorithms
• In order to get better angular resolution, other techniques for estimating the angles are used, e.g.:– MUSIC, subspace method using spectral search– ESPRIT, subspace method– MVM (Capon’s beamformer), rather easy spectral search method– SAGE, iterative maximum likelihood method
• Rather complex, one measurement point may take few minutes on a decent computer
2013-02-11 Fredrik Tufvesson - ETIN10 25
RUSK LUND, our broadband MIMO channel sounder
• A fast switched measurement system for radio propagation investigations at 300 MHz, 2 GHz and 5 GHz.
• Financed by Knut and Alice Wallenbergs stiftelse, FOI and LTH
2013-02-11 Fredrik Tufvesson - ETIN10 26
It’s all about measuring some delays...
• In MIMO systems we use the fact that there are several paths between the transmitter and receiver
• These paths are characterized by a– time delay, – phase shift, – attenuation, – angle of departure and – angle of arrival
• The angle of departure and angle of arrival result in a slight difference in time delay (phase) for each of the antenna elements
2013-02-11 Fredrik Tufvesson - ETIN10 27
It’s all about measuring some delays...
• In practice we measure the transfer functions between each of the antenna elements, and we calculate the parameters of interest
2013-02-11 Fredrik Tufvesson - ETIN10 28
Working principle
Courtesy
MEDAV
2013-02-11 Fredrik Tufvesson - ETIN10 29
Timing diagram
2013-02-11 Fredrik Tufvesson - ETIN10 30
Multicarrier spread spectrum sequence (MCSSS)
0 0.5 1 1.5
-0.5
0
0.5
1
time [µs]
norm
.mag
nitu
de
Tx signalin time
0 0.5 1 1.5
-0.5
0
0.5
1
time [µs]
norm
.mag
nitu
de
Tx signalin time
5.1 5.15 5.2 5.25 5.30
0.2
0.4
0.6
0.8
1
frequency[GHz]
norm
.mag
nitu
de
Tx signalin frequency
5.1 5.15 5.2 5.25 5.30
0.2
0.4
0.6
0.8
1
frequency
norm
.mag
nitu
de
Tx signalin frequency
MCSSS - Test Squence• periodic broadband Signal• high Correlation Gain• low Crest Factor or peak-to-average ratio • band limited (almost rectangular shape in frequency domain)• flexible in generation• simultaneous sounding of different bands (Up-/Downlink)
2013-02-11 Fredrik Tufvesson - ETIN10 31
The measurement system
• 200 kg of batteries to allow for 6 hours of mobile measurements• 640 MHz sampling frequency, to allow high Doppler frequencies• 2 separate PCs to manage the data flow from the A/D converters• Oven controlled rubidium clocks to maintain synchronization
during wireless measurements• GPS and wheel sensors to position the system• Broadband patch antennas with 128 antenna ports at 2.6 GHz• Circular 300 MHz antennas with a diameter of 1.5 m• Cylindrical 2 GHz array with 128 elements
2013-02-11 Fredrik Tufvesson - ETIN10 32
RUSK LUND transmitter
• Baseband (Arbitrary Wave Form) Signal Generator
• Frequency Synthesizer• Rubidium Reference• Modulator• Power Amplifier• MIMO Control Unit• GPS
• bandwidths: up to 240 MHz• frequency grid 10 MHz• max. power 500 mW, with
possibility for 10 W external amplifier
• carrier frequency ranges – 2200 – 2700 MHz, – 5150 – 5750 MHz– 235-387 MHz (20W)
• Power Supply 24 V DC and 230 V AC
2013-02-11 Fredrik Tufvesson - ETIN10 33
RUSK LUND receiver
• RF-Tuner• High Speed ADC• Automatic Gain Control (AGC)• MIMO Control Unit• Rubidium Reference• High Speed Data Recorder 320
MByte/s, 500 GByte
• GPS Receiver• Odometer Interface• total amplification 72 dB• AGC dynamic range 51 dB ,
adjustable in 3 dB steps, • intermediate frequency 160 MHz• bandwidth 240 MHz
2013-02-11 Fredrik Tufvesson - ETIN10 34
Antennas
4x16 dual polarized circular patch array
4x8 dual polarized rectangular array
• To get good resolution we want large size arrays
Some real world examples
2013-02-11 Fredrik Tufvesson - ETIN10 35