Cover Sheet for Presentation to IEEE 802.16 Broadband Wireless Access Working Group (Rev.

0)
Document Number: IEEE 802.16.1pp-00/15 Title:

Proposed System Impairment Models

Date Submitted: 2000-03-08 Source: John Liebetreu Sicom Inc. 785 East Redfield Road Scottsdale, Arizona 85260 Voice: Fax: E-mail: 408-607-4830 408-607-4806 john@sicom.com

Co-presenters: David Falconer, Carleton University, Tom Kolze, Broadcom, Yigal Leiba, Breezecom Venue: IEEE 802.16 meeting, Albuquerque, NM, March 6-10, 2000 Base Document: 802.16.1pc-00/15 http://ieee802.org/16/phy/docs/802161pc-00_15.pdf Purpose: Aid in the PHY Task Groups preparation of a detailed evaluation table for performance of PHY layer air interface proposals. Notice: This document has been prepared to assist the IEEE 802.16. It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor acknowledges and accepts that this contribution may be made public by 802.16. IEEE Patent Policy: The contributor is familiar with the IEEE Patent Policy, which is set forth in the IEEE-SA Standards Board Bylaws <http://standards.ieee.org/guides/bylaws> and includes the statement: IEEE standards may include the known use of patent(s), including patent applications, if there is technical justification in the opinion of the standards-developing committee and provided the IEEE receives assurance from the patent holder that it will license applicants under reasonable terms and conditions for the purpose of implementing the standard.

System Impairment Model

Ad hoc modelling committee: David Falconer, Carleton University Tom Kolze, Broadcom Yigal Leiba, Breezecom John Liebetreu, Sicom With thanks also to:
Naftali Chayat, (Breezecom) Bruce Cochran, (Sicom) Scott Enserink, (Sicom) Lucille Rouault, (ENST/NIST) Val Rhodes, (Intel) Benoit Verbaere, (ENST/NIST)

Process
Identify primary performance degradation sources Model and parameterize these sources Establish performance metrics Establish baseline characterization techniques

Performance degradation sources

Phase noise Power amplifier Multi-path Model parameters may be
Set by group and simulated by contributors Stated and simulated by contributors

Power Amplifier Models

Saleh Model
Uses simple two-parameter functions to model the AM-to-PM and AM-to-AM characteristics of nonlinear amplifiers. Originally developed to specify the behavior of TWTAs. Appropriate selections for the amplitude and phase coefficients (s and s) provide a suitable model for solid state amplifiers as well. It is a frequency-independent model. Can be made frequency-dependent by adding filters that mirror how the coefficients change with frequency.

Saleh Model
Input signal:
x(t)=r(t)cos[0t+(t)]
0 is the carrier frequency, r(t) is the modulated envelope (t) is the modulated phase

The output of the nonlinear amplifier is:

y(t)=A[r(t)]cos{0t+(t)+(r(t))} A(r) represents the AM-to-AM conversion (r) represents the AM-to-PM conversion.

Saleh Model
The specific forms of the two functions:

A(r)=ar/(1+ar2) (r)=r2/(1+r2)
As an example, the set of parameters that closely matches TWTA data [1] is, a= 2.1587 a= 1.1517 = 9.1040 = 4.033

Saleh Model
Saleh model with parameters:
a= 2.1587, a= 1.1517, = 4.033, = 9.1040
1.2 Kaye, George, and Eric 1 70 0.8 60 50 0.6 40 0.4 AM-AM AM-PM 30 20 0.2 10 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Input Amplitude 0
Output Phase (deg) Output Amplitude

90 80

Saleh Model
Saleh model with simplified parameters:
a= 2, a= 1, = 2 and = 1
1.2 Simplified 1
Output Amplitude

100 90 80 70 60
Output Phase (deg)

0.8

0.6 AM-AM 0.4 AM-PM

50 40 30 20 10

0.2

0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Input Amplitude

Saleh Model Summary

Uses simple two-parameter functions to model the AM-to-PM and AM-to-AM characteristics of nonlinear amplifiers. Appropriate selections for the amplitude and phase coefficients (s and s) can provide a suitable model for solid state amplifiers well. Salehs models for TWTAs are shown to accurately match actual measured data . Can be altered to a frequency-dependent model.

Rapp Model
Developed for solid-state power amplifiers. Produces a smooth transition for the envelope characteristic as the input amplitude approaches saturation. Vout = Vin/(1 + (|Vin|/Vsat)2P)1/(2P) Where Vsat is the saturation voltage of the power amplifier and P is the smoothness factor.

Rapp Model
Curves for various smoothness factors P:
Output Amplitude vs. Input Amplitude for Rapp Model of HPA
1.2

Output Amplitude

0.8

0.6

P = 100 0.4 P = 10 P = 2.0 P = 1.0 P = 0.5

0.2

0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Normalized Input Amplitude

Rapp Model- Modified

Honkanen and Haggman altered the low-level portion of the AM/AM characteristic in order to better mirror the exponential relationships of bipolar junction devices. Included AM/PM model as well. Their AM/AM and AM/PM models matched measurements of actual class AB mobile phone amplifier. Resulted in more accurate portrayal of intermodulation effects than the Rapp model when compared to a class AB mobile phone amplifier. They do not list their models parameters.

Ghorbani model
Similar approach to Saleh. Claimed more suitable for SSPAs then Saleh. PA output : y(t)=A(r(t))cos{0t+(t)+(r(t))} where, A(r) = x1rx2/(1+x3rx2) + x4r (r) = y1ry2/(1+y3ry2) + y4r For the GaAs FET SSPA characterized by Ghorbani:
x1 = 8.1081 x2 = 1.5413 x3 = 6.5202 x4 = -0.0718 y1 = 4.6645 y2 = 2.0965 y3 = 10.88 y4 = -0.003

Ghorbani model compared to Rapp

Ghorbani model AM/AM curve, customized to a FET, and Rapps AM/AM curve:
1.2 1
Ghorbani Parameters: x1 = 8.1081 x2 = 1.5413 x3 = 6.5202 x4 = -0.0718 Rapp Parameters: P = 10 Vsat = 1.0

Normalized Output Voltage

0.8

0.6

0.4 Ghorbani Output Voltage Rapp Output Voltage 0.2

0 0 0.5 1 1.5 2 2.5 3 3.5 Normalized Input Voltage

Ghorbani Compared to Saleh

Ghorbani model AM/AM curve, customized to a FET, and Saleh models best fit to that curve:
1.2 1
Ghorbani Parameters: x1 = 8.1081 x2 = 1.5413 x3 = 6.5202 x4 = -0.0718 Saleh Parameters: = 1.3325 = 0.3403

Normalized Output Voltage

0.8

0.6

0.4 Ghorbani Output Voltage Saleh Output Voltage 0.2

0 0 0.5 1 1.5 2 2.5 3 3.5 Normalized Input Voltage

Ghorbani Compared to Saleh

Ghorbani model AM/PM curve, customized to a FET, and Saleh models best fit to that curve:
0.45 0.4

0.35

Output Phase Shift (rad)

0.3

0.25

0.2

Ghorbani Parameters: y1 = 4.6645 y2 = 2.0965 y3 = 10.88 y4 = -0.003 Saleh Parameters: = 5.4514 = 12.9957

0.15

0.1 GhorbaniOutput Phase Shift Saleh Output Phase Shift 0.05

0 0 0.5 1 1.5 2 2.5 3 3.5 Normalized Input Voltage

Modified Saleh (Enserink)

Ghorbani model AM/AM curve, customized to a FET, and modified Saleh AM/AM fit to that curve:
1.2

Normalized Output Voltage

1
Ghorbani Parameters: x1 = 8.1081 x2 = 1.5413 x3 = 6.5202 x4 = -0.0718 Saleh-Modified Parameters: = 20.0 = 19.5 (Used same form as Saleh's AM/PM for the AM/AM as well.)

0.8

0.6

0.4

Ghorbani Output Voltage 0.2 Saleh-Modified Output Voltage 0 0.5 1 1.5 2 2.5 3 3.5

0 Normalized Input Voltage

Ghorbani Compared to Saleh

Saleh model matches the GaAs FET amplifiers AM/PM characteristic well. Saleh model does not match the FET amplifiers AM/AM characteristic very well. Can improve the match by changing the Saleh AM/AM equation to have the same form as the Saleh AM/PM equation. Ghorbani model is better suited to the FET amplifiers characteristics and matches them closely.

Recommendation
Adopt the well-known Saleh model as a comparison baseline. Baseline model serves as a reference point for comparison with other power amplifier models, (e.g., Ghorbani model).

References
A.A.M. Saleh, Frequency-independent and frequency-dependent nonlinear models of TWT amplifiers, IEEE Trans. Communications, vol. COM-29, pp.1715-1720, November 1981. A.R. Kaye, D.A. George, and M.J. Eric, Analysis and compensation of bandpass nonlinearities for communications, IEEE Trans. Communications Technology, vol. COM-20, pp.965-972, October 1972 C. Rapp, Effects of HPA-Nonlinearity on a 4-DPSK/OFDM-Signal for a Digitial Sound Broadcasting System, in Proceedings of the Second European Conference on Satellite Communications, Liege, Belgium, Oct. 22-24, 1991, pp. 179-184. M. Honkanen and Sven-Gustav Haggman, New Aspects on Nonlinear Power Amplifier Modeling in Radio Communication System Simulations, Proc. IEEE Int. Symp. On Personal, Indoor, and Mobile Comm.,PIMRC 97, Helsinki, Finland, Sep. 1-4, 1997, pp. 844-848. A. Ghorbani, and M. Sheikhan, The effect of Solid State Power Amplifiers (SSPAs) Nonlinearities on MPSK and M-QAM Signal Transmission, Sixth Intl Conference on Digital Processing of Signals in Comm., 1991, pp. 193-197.

Phase noise assumptions

Purpose: weighing sensitivity of different proposals to phase noise not an interface specification Transmitter mmW up-converter and receiver mmW down-converter are expected to dominate phase noise Based on PLL-oscillator model

SSB phase noise PSD, L(f)

log(PSD) 20dB/dec Lloop 20dB/dec Lfloor 20dB/dec

1Hz

fxtal

floop

100MHz

log(f)

Phase noise model

The model has four parameters
Corner frequency for crystal phase noise Corner frequency for PLL loop LO noise floor level PLL phase noise level

Two parameters for ease of simulation are a zero at 1Hz, and a pole at 100MHz To ease simulation, 1/f noise is not accounted for

Phase noise notes

Thermal noise, discrete spurs and demodulator induced phase noise are NOT included in this model. Model is to be used for comparison purposes, NOT for precise performance evaluation

ETSI/BRAN Multipath Models

ETSI/BRAN document HAPHY151TL03, Channel model suitable for bands over 20 GHz, 21 Sept. 1999.

T1:
0.96 -0.19 0 40 ns.

T2:

0.96 0.19exp(-j) 0 40 ns.

=(1-0.8(40 ns.)/Tsymbol))

Based on measurements in Europe by Telia

ETSI/BRAN (cont.) and Papazian

G3:
0.8 0.4 0 20 ns. 0

G4:

0.8 0.4exp(-j) 20 ns.

L7/Papazian:
0.65 -0.47 -0.1 0 3.8 ns. 15.2 ns.

=(1-0.8(20 ns.)/Tsymbol))

Some Measured Kanata Responses

Kanata channel kab23002 0.7 e d u 0.6 t i n g 0.5 a m e 0.4 s n o 0.3 p s e R 0.2 0.7 e d u 0.6 t i n g 0.5 a m e 0.4 s n o 0.3 p s e R 0.2 Kanata channel kgb30044

0.1 0.1 0

-100 -80 -60 -40 -20 0 20 Time (ns.)

40

60

80 100

-100 -80 -60 -40 -20 0 20 Time (ns.)

40

60

80 100

Kanata channel kab29069 Kanata channel kab23031 0.7 e d u 0.6 t i n g 0.5 a m e 0.4 s n o 0.3 p s e R 0.2 0.7 e d u 0.6 t i n g 0.5 a m e 0.4 s n o 0.3 p s e R 0.2

0.1 0.1 0 0 -100 -80 -60 -40 -20 0 20 Time (ns.) 40 60 80 100

-100 -80 -60 -40 -20 0 20 Time (ns.)

40

60

80 100

Proposed Multipath Models

Model A1:
.929 .371 exp(-j) 0 0 Variables: -<< 0<<50 ns.

Model A2:
.929

Model A3:
.995 .0995 exp(-j) 0

.371 exp(-j)

Proposed Multipath Models (cont.)

Model B:
.908 Variable: -<< -0.279 0.279 exp(-j) -0.140

-20 ns.

20 ns.

50 ns.

Time variation? -- slow compared to symbol rate

Frequency Responses
) B d ( e s n o p s e r Examples of Channels A1, A2 and B Frequency Responses 5 A1, A2, =20 ns., =0 deg. 0

-5

=12 ns., =144 deg.

l e n n -10 a h C -15

B, =144 deg.

-20 -50

-40

-30 -20 -10 0 10 20 30 Frequency relative to center frequency (MHz)

40

50

SNR Degradation for (8,1) DFE (50 Megasymbols/s)

7 6 5 SNR 4 degradation 3 (dB) 2 1 0 0 45 90 135 180 Phase shift (degrees) MODEL A1 MODEL A2 MODEL B

SNR Degradation for (8,1) DFE (25 Megasymbols/s)

7 6 5 4 SNR 3 degradation 2 (dB) 1 0 -1 -2 Phase shift (degrees ) 0 45 90 135 180 MODEL A1 MODEL A2 MODEL B

Conclusions on Multipath Modeling

Three 2-tap and one 3-tap models proposed for PHY evaluation purposes, with variable phase and delay parameters. Worst case channels, including some with precursors (non-minimum phase). Examples of equalizer performance (not optimized). Fairly consistent with others models in terms of delay spread and echo magnitudes.