CC BY
45
Research of influence of DRM broadcast transmitter nonlinearities onto the output signal parameters
Keywords: DRM, Digital Radio Mondiale, broadcast transmitter, nonlinear distortion, MER, out-of-band emissions, spectrum mask.
Digital Radio Mondiale (DRM) is an OFDM-based digital radio standard for long-, medium- and shortwave bands. Modern AM broadcast transmitters (with PDM or PSM modulator) with minor modifications or linear (SSB) transmitters may be used. Quality of transmitter is described by MER in output signal and accordance of out-of-band emissions spectrum mask. Nonlinearities in transmitter lead to increase out-of-band emissions in the adjacent channel and degradation of desired signal quality. A computer simulation of nonlinearities in transmitter is chosen as a research method of analysis the MER parameter and spectrum of output signal of DRM broadcast transmitter. The analysis procedures have been done as a real time, using real DRM signal, software simulator of nonlinear transmitter and software DRM measuring receiver and spectrum analyzer. Most often-used types of transmitter's AM-AM and AM-PM nonlinearities were simulated. Critical levels for different types of non-linearity were found.Analysis show that influence of nonlinearities on MER is less critical that on output spectrum. Accordance of spectrum mask must be using during transmitter tuning.
Oleg Varlamov,
Moscow Technical University of Communications and Informatics, senior staff scientist, Ph.D. [email protected]
Introduction
Digital Radio Mondiale (DRM) is a new OFDM-based digital radio standard for long-, medium- and short-wave bands [1]. It was designed to use the same frequency allocation as the current broadcast analog systems to offer a high degree of compatibility. DRM also assumes using the same modern AM broadcast transmitters (with PDM or PSM modulator) with minor modifications. Linear (SSB) transmitters also may be used but with esser efficiency.
In general DRM signal is required more linearity of transmitter than from traditional AM broadcast signal. Nonlinearities in transmitter lead to increase out-of-band emissions in the adjacent channel [2] and degradation of desired signal quality. Specific requirements for PMD transmitters (such as DC offset error, admissible time delay between envelope and phase signals, minimum envelope bandwidth) are described in [3].
The aim of this report is nonlinearity influence research of transmitter on the DRM output signal quality. They apply to both linear and PDM transmitters.
Choice of analysis procedure
The most often-used measure of performance for digital communication systems is average biterror rate (BER). BER measurements are complex and long in time process and so it is not comfortable during transmitter tuning process. More comfortable is to use of signal/noise ratio (SNR) measurement.
Products of transmitter nonlinearities is of
Gaussian distribution [4] and it's influence on DRM signal is alike AWGN channel noises. That is why SNR and BER are related. Minimal required SNR evels on receiver input are given in [1]. SNR on transmitter output must be higher on 10-15 dB to eliminate transmitter nonlinearities influence on link budget. So SNR on transmitter output must be greater than 30-35 dB.
International DRM standardization documents using analog of SNR, called Modulation Error Ratio (MER) [5]. This parameter is a single "figure of merit" analysis of the transmitted signal and shall be not less than 30 dB. Also transmitter out-of-band emissions must lie under spectrum mask described at [6]. Russian Federation DRM standardization documents are described in [7].
The bandwidth of a DRM signal is smaller than 20 kHz and the number of carriers used in the OFDM-modulation is relatively small (Max 460). This characteristic motivates to do a real-time software simulation of nonlinear transmitter on a conventional personal computer (PC) with using the soundcard as the input and output device. But more wider bandwidth required for out-of-band emissions observation. So soundcard with 96 kHz sample rate was used.
"DReaM" open source software implementation of DRM receiver [8] with MER measurement function was used as a receiver. PC-based FFT spectrum analyzer with spectrum mask function [9] was used for measurement out-of-band emissions.
Nonlinear transmitter was modeled through the description of their AM/AM and AM/PM characteristics. For that, real DRM signal (leading from file in I/Q format on IF 12 kHz) is converted to polar representation. Then, nonlinearities are added and backward conversion to I/Q format on zero IF is executing. In that way we have DRM signal with transmitter's distortions, which is feeding to spectrum analyzer and receiver for MER measurement.
Simulation
Most often-used types of transmitter nonlinearities were simulated.
For AM-AM distortion it was:
rr fx-(1+a)-a, when x> a/(1+a)l
• cutoff: y(x)= \ y J v ;
|0,when x<a/(1+a) J
• quadratic nonlinearity: y(x) = ax2 + x
• cubic nonlinearity: y(x) = ax3 + x
• S-type nonlinearity: y(x) = ax3 + bx2 + x, when a < 0, b > 0;
AM-AM nonlinearity
AM-AM nonlinearity
40
30
20
m-t
50 g 40 g
30
20
/33 V 34. 1
-0.4 -0.2 0 0.2 0.4
a
—quadratic nonlinearity ~ ■ cubic nonlinearity
Rg. 2. AM-AM distortion: quadratic and cubic nonlinearity
dli 30 40 50 60 70 80 90 -100 -110 -120
-2
-1
0
b=-a
|----S-type nonlinearity ■ -N-type nonlinearity
Fig. 3. AM-AM distortion: S-type and N-type nonlinearity
•'YWv'A
\
0 2 4 6 8 10 1214 1618 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48kII/
Fig. 4. Example of simulated output spectrum with critical mask compliance for S-type nonlinearity with a = -0,7
AM-PM unevenness
50
40
30
20
1
0.1
0.2 b, radian
0.3
0.4
----quadratic unevenness
----cubic unevenness
----unevenness view as: y(x)=a(l-x)A2
----unevenness view as: y(x)=a(1-x)A3
----unevenness view as: y(x)=axA2+bxA3, a=b
----unevenness view as: y(x)=a(1-x)A2+b(1-x)A3, a=b
Fig. 5. AM-PM distortion
• N-type nonlinearity: y(x) = = ax3 + bx2 + x, when a > 0, b < 0, where a and b are parameters, variable during simulation. Graphical representations of simu-ation results for AM-AM distortion are shown on Fig. 1-3 respectively. Example of simulated output spectrum with critical mask compliance for S-type nonlinearity with a = -0,7 is shown on Fig. 4.
Critical values of parameters a and b for accordance of DRM out-of-band emissions spectrum mask are shown on all Figures as a blue labeled points.
For AM-PM distortion it was:
• quadratic unevenness: y(x) = ax2;
• cubic unevenness: y(x) = ax3;
• unevenness view as: y(x) = a(1 - x)2;
• unevenness view as: y(x) = a(1 - x)3;
• unevenness view as: y(x) = ax2 + bx3;
• unevenness view as: y(x) = a(1 -x)2+b(1 -x)3.
Graphical representations of simulation results
for AM-AP distortion are shown on Fig. 5.
Comparison of these two criteria (MER and accordance of DRM spectrum mask) show, that influence of transmitter nonlinearities on output spectrum is more critical. That is why, accordance of ITU spectrum mask must be used as a criteria during transmitter tuning. In that case MER on transmitter output will be sufficiently high (around 30 dB) automatically.
In general, DRM signal is more sensitive to distortion at low amplitude.
Conclusions
A computer simulation of nonlinearities in transmitter on MER and DRM out-of-band emissions spectrum mask in broadcast DRM transmitter output signal was done. Critical levels for different types of non-linearity were found. Analysis show that influence of nonlinearities on MER is less critical that on output spectrum. Accordance of spectrum mask must be using during transmitter tuning.
References
1. ETSI ES 201 980 V3.1.1 (2009-08) Digital Radio Mondiale (DRM); System Specification.
2. Rec. ITU-R SM.328-10. Spectra and bandwidth of emissions.
3. DRM Introduction and Implementation Guide. Revision 2. September 2013. http://www.drm.org /wp-content/uploads/2013/09/DRM-guide-artwork-9-2013-1.pdf, date of access 31.01.2013.
4. Chris van den Bos, Michiel H. L. Kouwenhoven, Wouter A. Serdijn "Effect of Smooth Nonlinear Distortion on OFDM Symbol Error Rate", IEEE Transactions on communications, Vol. 49, No. 9, pp. 1510 - 1514, Sep. 2001.
5. ETSI EN 302 245-1 V1.1.1 (2005-01) Electromagnetic compatibility and Radio spectrum Matters (ERM); Transmitting equipment for the Digital Radio Mondiale (DRM) broadcasting service; Part 1: Technical characteristics and test methods.
6. ETSI EN 302 245-2 V1.1.1 (2005-01) Electromagnetic compatibility and Radio spectrum Matters (ERM); Transmitting equipment for the Digital Radio Mondiale (DRM) broadcasting service; Part 2: Harmonized EN under article 3.2 of the R&TTE Directive.
7. Oleg Varlamov "Development of national regulatory framework for DRM digital broadcasting" / T-Comm — Telecommunications and Transport, 2013. No9. pp. 47-50.
8. DReaM receiver download page on Sourceforge: http://sourceforge.net/projects/drm/, date of access 31.01.2013.
9. Oleg Varlamov, Viktor Gromorushkin, Valeri Lavrushenkov, Igor Chugunov "Generator of test signals for measuring characteristics of EER SSB switching power amplifiers" / T-Comm — Telecommunications and Transport, 2011. No 9. pp. 47-49.
