See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261073792

Performance of comparison metrics on M-QAM OFDM systems with high power

amplifier

Conference Paper · October 2012

DOI: 10.1109/WICT.2012.6409204

CITATIONS READS
5 381

2 authors:

Deepa Thangavel R. Kumar

SRM Institute of Science and Technology University of Northern Iowa
33 PUBLICATIONS   43 CITATIONS    40 PUBLICATIONS   125 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Visible Light D2D Communication via LED View project

All content following this page was uploaded by Deepa Thangavel on 18 July 2014.

The user has requested enhancement of the downloaded file.

 Performance of Comparison Metrics on M-QAM
OFDM systems with High Power Amplifier
Deepa.T1
1 2
Department of Telecommunication Engineering
SRM University.
Chennai,India
deepa.t@ktr.srmuniv.ac.in
Abstract—Orthogonal frequency division multiplexing (OFDM)
has been adopted by several wireless transmission standards. A
major drawback of OFDM is the large dynamic range of the
transmit waveforms, making OFDM vulnerable to nonlinearities
of the power amplifier (PA) and causing the PA to yield low
power efficiency on the RF to dc power conversion. A commonly
used performance metric to characterize a dynamic range of
OFDM signal is the peak-to-average power ratio (PAPR). To
suppress the nonlinear effects, one may want to reduce the PAPR
value. However, this results in the increase of error vector
magnitude (EVM) and decrease of modulation error ratio
(MER). EVM and MER analysis is useful tool for prediction of
dynamic performance. In this paper, we compare among the
different performance metrics such as EVM and MER and show
that they can be equivalently useful as signal to noise ratio(SNR)
and bit error rate(BER). Simulation results are shown to
illustrate the performance of EVM and MER based on these
relationships. Simulation result shows that by modulating a
signal at fixed subcarriers, the MER increases and EVM
decreases.
Keywords-Orthogonal frequency division multiplexing
(OFDM), Error vector magnitude (EVM) vector magnitude(EVM),
Modulation error rate(MER).
I. INTRODUCTION
Orthogonal Frequency Division Multiplexing (OFDM) is an
efficient modulation technique, because of its high data rate ,
high spectral efficiency and mitigating wireless impairments .
This technology has been proposed as the standard for
broadcasting both audio and video digital signals and for
wideband wireless communication systems [1], [2]. However
their consisting of large numbers of independent QAM
subcarriers, means the composite signal’s peak to average
power ratio (PAPR) can be significant. Nonlinear distortion is
a source of major degradation of modulation fidelity in
multicarrier systems, such as OFDM systems. The
fundamental band of the RF nonlinear power amplifier(PA)
transmitter system output for an OFDM signal input is
composed of large numbers of inter-modulation products
(IMPs) superimposed on the inband amplified OFDM
subcarriers and appearing in adjacent channels as spurious
out-of-band emissions[3]. The IMP impairment of the inband
signal modulation fidelity is measured by error vector
978-1-4673-4804-1 2012
c IEEE
Kumar R 2
Department of Eectronics and Communication
Engineering
SRM University
Chennai,India
kumar.r@ktr.srmuniv.ac.in
magnitude (EVM).The performance of the OFDM scheme is
also severely affected by the phase noise of the oscillators at
the receiver. The modulation error rate (MER) is widely
applied as a measure of OFDM signals' performance, which
integrates impairments such as nonlinearity, IQ imbalance,
phase noise, as well as Doppler frequency shift [4].
The paper is structured as follows. Section 2 describes OFDM
system model with its algorithm . Section 3 describes briefly
the performance metrics such as PAPR,EVM,MER and BER.
Section 4 describes the evaluation of CCDF and the effect of
performance metric on OFDM systems are simulated using M
ary QAM modulation techniques. Conclusions are given in
Section 5.
II. OFDM SYSTEM MODEL
A. Description of OFDM system Algorithm
In OFDM system, a random bit sequence is generated, and
then the bits are mapped into 64-QAM symbols. The block
diagram of OFDM transmitter is shown in Fig. 1.The I and Q
~ ~
values of this sequence are stored as the I j and Q j array.
~ ~
The I j and Q j array is next used to form the OFDM
frequency domain signal, X(f). An inverse FFT function is
then used to compute the time domain signal, X(t). The input
time domain signal is then scaled to the required average
input power level. Then the power amplified time domain
signal, Xa(t) is calculated. Using the FFT function, the PA
output frequency domain signal, Xa(f), is calculated. From
Xa(f), the resulting OFDM symbols I j , Q j are mapped. The
algorithm for computing performance matrics is shown in
Fig.2
B. High pwer Amplifier Model
The IFFT output signal is then passed through a nonlinear
power amplifier model. The complex envelope of the input
signal into the amplifier can be expressed as [5],[6],
jφ ( t )
x a (t ) = x (t ) e
(1)
910
 Mapping of 64 IFFT OFDM
QAM symbols & Power Demodulator
~ ~ OFDM Modulator Amplifier &
I j ,Qj
FFT

Fig. 1. Block diagram of OFDM system

The high power amplifier (HPA) is Rapp’s solid state power Where,
amplifier model (SSPA) with amplitude and phase Xn is the mapped information data
characteristics, N is the number of subcarriers
x(k) is the IFFT output signal
k 1 x (t )
A[ x (t ) ] = 1
(2) The complementary cumulative distribution function can be
2p expressed as,
ª § k x (t ) · º 2p

«1 + ¨ 1 ¸ »
CCDF = Probability (PAPR > PAPR0) (6)
« ¨© A o ¸
¹ »
¬ ¼
Where PAPR0 is the threshold level

4 Generate Random data sequence

§ k x(t ) ·
φ [ x(t ) ] = α φ ¨¨ 1 ¸
¸ (3)
© Ao ¹
Map to 64 QAM data
Where, A0 is the saturation output amplitude
K1 is the small signal gain
x(t) is the complex envelope of the input signal ~
Store original reference Ij and
p is a knee factor which controls the smoothness of the
transition from the linear region to the saturation region
ĮɎ is typically set to zero, meaning SSPA adds no phase
distortion Generate OFDM signal in frequency

III. PERFORMANCE METRICS

Compute output signal from SSPA model
The performance metrics such as PAPR, EVM, MER and
BER which quantify the dynamic range of OFDM signals.
A. Peak To Average Power Ratio (PAPR) Use equation 4 to Calculate PAPR of
PAPR for the discrete time OFDM signal x(n) is defined as the
ratio of maximum signal power to average signal power [7].
2 Compute FFT of Xa(t)
max x(n)
PAPR = 10 log 2
(4)
Avg x(n)
Compute measured values I j , Q j from
The PAPR reduction capability is measured by the CCDF amplitude and phase of FFT output signal
which indicates the probability that the PAPR exceeds a
certain threshold value. The CCDF of PAPR can be applied to
determine the bounds for the minimum number of redundancy Calculate EVM and MER from
bits required to identify the PAPR sequences.
The complex representation of OFDM signal can be expressed
as, Demodulate of FFT output signal
N −1 j 2 Π nk
1
x(k ) = ¦X e n
N
, k = 0,1,.....N − 1 (5)
N n =0
Fig. 2. Program algorithm of OFDM system

2012 World Congress on Information and Communication Technologies 911

 the average power per symbol of the constellation. OFDM
B. Constellation Diagrams and EVM
Signals are demodulated before EVM calculations are made.
A constellation diagram is a representation of a signal EVM is a measure of the deviation of the demodulated
modulated by a digital modulation scheme such as quadrature
received symbol ( I j , Q j ) from the original transmitted data
amplitude modulation (QAM) or phase-shift keying (PSK)[8].
It provides a graphical representation of the complex envelope ~ ~
symbol ( I j , Q j ).
of each possible symbol state. The distance between signals on
a constellation diagram indicates how different the modulation
signals are, and how well a receiver can differentiate between
all possible symbols in presence of noise. Measured Measured
constellation diagrams can be used to recognize the type of Signal
interference and distortion in a signal. Figure 3 shows three Q Channel Error Vector
constellation diagrams for 64 QAM-OFDM systems , which
has 64 symbols that modulate the RF carrier in both
magnitude and phase. Constellation diagrams that show the φ
Reference Signal
ideal placement of symbols for a given 64-QAM OFDM
system are often represented by symbols at integer levels as Phase Error
seen in Fig.3(a). Fig.3(c) shows the normalized constellation
diagram representation from Fig. 3(a) and 3(b).
Fig. 4 shows the error vector for one measured symbol. In this I Channel
case with only one measured symbol, the magnitude of this Fig. 4. EVM Definition
small vector equals the EVM. If there were more symbols
acquired than just this one, the EVM would equal the sum of The EVM, magnitude error and phase error can be expressed
the magnitudes of the error vectors for all of the measured as[10],

¦ [(I )]
symbols divided by the total number of measured symbols.
) + (Q
N
1 ~ 2 ~ 2
j −Ij j − Qj
N j =1
EVM rms = (7)
Vmax

2 2 ~2 ~ 2
MagnitudeE rror = I j + Q j − I j + Q j (8)

~
Qj Qj
PhaseError = arctan − arctan ~ (9)
Ij Ij
(a) (b)
Where,
Vmax is the maximum amplitude of the ideal constellation
points
N is the number of points in a measurement
I j , Q j are the ideal and quadrature components of the j-th
measured OFDM signal
~ ~
I j , Q j are the ideal and quadrature components of the j-th
referenced OFDM signal
EVMRMS is the RMS error magnitude
(c)

To calculate and compare EVM efficiently, the diagrams in

Fig. 3. (a) Ideal constellation diagram (b) measured constellation diagram (c)
normalized constellation diagram
Thus, when symbols have been normalized, EVM is defined
as the root-mean-square (RMS) value of the difference
between a collection of measured OFDM symbols and ideal
OFDM symbols [9]. These differences are averaged over a
given, number of symbols and are often shown as a percent of
Figs. 3(a) and 3(b) are scaled to form the normalized
(dimensionless) constellation diagram in Fig. 3(c).When the
received signal is corrected and normalized, the RMS error
between the expected (ideal) and measured signal is
calculated. The EVM calculation not only provides RMS
value of the error vector magnitude but also the gain and
quadrature error of the system, which can be utilized to
characterize the overall path gain and phase shift of the
system. The allowable EVM values for each data rate are

912 2012 World Congress on Information and Communication Technologies

specified by the standard. Large error vector magnitudes are
tolerable for low data rate systems, as long as the received
symbols fall in their respective constellation region. However,
as the system speed increases, these regions shrink and
necessitate a more accurate demodulation for error free
recovery of the transmitted signal. High bit rate systems,
therefore, are required to produce lower EVM values to
guarantee error free demodulation. Error vector magnitude is
normally expressed as a linear measurement in percent.
C. Modulation Error Ratio(MER)
MER is another measurement metric of OFDM system is
closely related to EVM because in a single numerical value it
summarizes the quality of a OFDM transmitter. As shown in
Fig. 5 Modulation error ratio is the ratio of average symbol
power to average error power Modulation Error is used to
Measure of Modulation quality[11].
( )
N 
§ ~2 ~ 2
¨ ¦ I j + Qj
¨
MER(dB) = 10 log¨ N
j =1
[( )]
~ 2 ~
¨¨ ¦ I j − I j + Q j − Q j) (
© j =1
The MER is related to the OFDM constellation used by the
OFDM modulator, which can be QPSK,M-QAM depending
on the modulation scheme chosen for each transmitter site.
The MER is related to how much the constellation radiated by
the OFDM signal transmitted is similar to the ideal
constellation, which would be radiated without any distortion
or intermodulation effects [12]. Any power amplifier system
will introduce distortion over the OFDM spectrum.
Modulation Error
Q Channel Target
Symbol
Transmitter Symbol
I Channel
Fig. 5. MER Definition
The more is the power amplifier distortion, the more the
constellation differs from the reference one and the MER
value is low. So the MER value should be maximized.
Fig. 5 illustrates the vector for a target symbol (ideal symbol )
we want to transmit. Because of one or more impairments, the
2012 World Congress on Information and Communication Technologies
transmitted symbol vector (or received symbol vector) is a
little different than ideal. Modulation error is the vector
difference between the ideal target symbol vector and the
transmitted symbol vector. In the case of MER, the higher the
number, the better.
D. Bit Error Rate
BER measures how often symbols are pushed into boundaries
of neighboring symbols, causing these symbols to be
misinterpreted. BER does not measure the condition of the
modulated signal itself, though a poor BER is an indicator of
poor signal quality [13][14]. Because the BER measurement
detects and counts every misinterpreted bit, it is a sensitive
indicator of problems caused by transient or bursty noise
interference. BER is a commonly used performance metric
which describes the probability of error, Pe in terms of number
of bits per bit transmitted.
·
¸
§ 1·
¸ 2¨1 − ¸
L ¹ §¨ ª 3 log 2 L º 2 Eb
¸ ·
2
¸¸ Pe = © Q
log 2 L ¨© «¬ L2 − 1 »¼ N o
¸ 
¸ (11)
¹ ¹
(10)
Where,
L is the number of levels in each dimension of the M-ary
QAM modulation system
Eb is the energy per bit
N0/2 is the noise power spectral density
Q [.] is the Gaussian co-error function
IV. SIMULATION RESULTS
To show the PAPR ,EVM,MER and BER performance of
OFDM systems, we considered an OFDM system using 1024
subcarriers and different M-QAM modulatios(for example
M=4,16,32,64 M is the modulation order) ,simulated by
randomly generated data. PAPR is measured for OFDM
system using randomly generated data bits with M ary QAM
modulations, the result is shown in Fig.6. In Fig. 6 it can be
seen that by keeping fixed number of subcarrier (N=1024 are
taken for example) and increasing the number of bits the
PAPR value is not increasing rapidly as shown in Fig. 6. The
PAPR of OFDM signal at each modulation changes in small
fractions while keeping the subcarriers constant.The Fig. 7
shows the BER versus SNR performance of different
modulation systems. It can be noticed that the BER decreases
rapidly by increasing SNR. From Fig. 7 , it is clear that the
BER is lesser using 4 ary QAM modulation technique rather
than 16 ,32 and 64 ary QAM ,by increasing the number of bits.
The BER versus EVM curve shown in Fig. 8(a) .It can be
noticed that the inverse relationship that exists between BER
and EVM. From Fig. 8(a),it can be seen that there is a constant
1.5% difference between 4 and 16-QAM, whereas there is a
913
 Fig. 7 BER versus Eb/No performance of OFDM system
Fig. 6.CCDF performance of OFDM system

8(a) 8(b)

8(c) 8(d)

Fig. 8.performance curves (a)BER versus measured EVM(b) BER versus MER (c) Measured EVM versus SNR
performance (d)MER versus SNR performance

914 2012 World Congress on Information and Communication Technologies

 0.5% difference between 16-QAM and 32-QAM at 10-3 BER
level. The BER versus MER curve shown in Fig. 8(b). In Fig.
8(b), we note that there is a constant 20 dB difference between
4 and 16-QAM, whereas there is a 50dB difference between
16-QAM and 32-QAM at 10-2 BER level. Fig. 8(c) shows
EVM versus SNR performance of M –QAM OFDM system.
Fig. 8(c) shows EVM versus SNR performance of M-QAM
OFDM systems. It is clear that the lower the EVM, the better
system performance. Fig. 8(d) shows MER versus SNR
performance of M-QAM OFDM system. From Fig. 8(d),it
can be seen that the higher the MER, the better system
performance.
CONCLUSION
In this paper, we have studied the EVM, MER and BER,
performances of the M-QAM-OFDM systems subject to the
nonlinear HPA in AWGN channels. An analytical description
of the joint effects of this imperfection on the performance of
OFDM systems has been presented. We have derived closed
forms of the EVM, MER and BER expressions for AWGN
channels. Analytical and simulation results show that the
higher the MER and the lower EVM. EVM and MER values of
OFDM systems were computed and simulated in order to relate
them to BER and SNR figures. Using the measured root mean
square error magnitude data in combination with the plots from
Fig. 8(a) and Fig. 8(c), the dynamic performance of the OFDM
receiver can be predicted.
REFERENCES
[1] H. Schulze and C. L¨uders, Theory and Applications of OFDM and
CDMA: Wideband Wireless Communications, Wiley, 2005.
[2] R.V. Nee & R. Prasad, OFDM wireless multimedia communications
(London: Artech House, 2000).
[3] A.Frank H.P Fitzek,Basak Can,Nguyen Cong Haun,Muhammad Imadur
Rahaman, and Ranjee Prasad,”Cross lay er optimization of OFDM
systems for 4G wireless communication,”center for Tele
Infrastructure(CTIF) , Aalberg University.
2012 World Congress on Information and Communication Technologies
View publication stats
[4] J. Lima Pinto, I. Darwazeh, “Phase Distortion and Error Vector
Magnitude for 8-PSK Systems”, Proceedings of London
Communications Symposium, LCS 2000, London, U.K., 14-15
September, pp.37-40.
[5] E. Costa, S. Pupolin, “M-QAM-OFDM System Performance in
thePresence of a Nonlinear Amplifier and Phase Noise” IEEE
Transactions on Communications, , pp 462-472, March 2002.
[6] J. K. Cavers, “The Effect of Data Modulation Format on
Intermodulation Power in Nonlinear Amplifiers”, IEEE Transactions on
Communications, pp. 489-493, March 1994.
[7] S. H. Han, J. H. Lee, “An Overview of Peak – to – Average Power Ratio
Reduction Techniques for Multicarrier Transmission”, IEEE
Transaction on Wireless Communication, April 2005.
[8] Lajos Hanzo, William Webb, and Thomas Keller.Single- and Multi-
Carrier Quadrature Amplitude Modulation. Wiley, Chichester, 2nd
edition, 2000.
[9] Michael D. McKinley, Kate A. Remley, Maciej Mylinski, J. Stevenson
Kenney, Dominique Schreurs,and Bart Nauwelaers. “EVM Calculation
for BroadbandModulated Signals”. Technical Report,2005. Work of
United States Government.
[10] R. Hassun, M. Flaherty, R. Matreci, M. Taylor, "Effective Evaluation of
Link Quality using Error Vector Magnitude Techniques", Proceedings of
1997 Wireless Communications Conference, Boulder, CO, USA 11-13
August 1997, pp. 89-94.
[11] Yin Wing,Zhaowu Chen and Ke Gong,”MER Performance anlaysis of
M-QAM OFDM with Wiener Phase noise”IEEE Press, 18-21 April
2007,doi:10.1109/ICMMT.2007.381517.
[12] K.M. Ghairabeh, K.G. Gard, and M.B. Steer. “Accurate Estimation of
Digital Communication System Metrices - SNR, EVM and _ in a
Nonlinear Amplifier Environment”. IEEE Transactions on
Communications, pages pp.734–739, Sept. 2005.
[13] Lajos Hanzo, William Webb, and Thomas Keller.Single- and Multi-
Carrier Quadrature Amplitude Modulation. Wiley, Chichester, 2nd
edition, 2000.
[14] Andrea Goldsmith. ‘’Wireless Communications’’,Cambridge University
Press, Stanford University, 1st edition edition, 2005.
915