Performance Analysis of PAPR Reduction in MIMO-OFDM

ISSN (ONLINE): 2395-695X
ISSN (PRINT): 2395-695X
International Journal of Advanced Research in Basic Engineering Sciences and Technology (IJARBEST)
Vol. 2, Issue 11, November 2016
Jayaraman.G et al. © IJARBEST PUBLICATIONS
Performance Analysis of PAPR Reduction in
MIMO-OFDM
Jayaraman.G1
, VeeraKumar K2
, Selvakani.S3
Assistant Professor, Department of ECE, Francis Xavier Engineering College, Tirunelveli 1
Associate Professor, Department of ECE, Francis Xavier Engineering College, Tirunelveli 2
PG Scholar, Department of ECE, Francis Xavier Engineering College, Tirunelveli 3
Abstract— In communication system, it is aimed to provide highest possible
transmission rate at the lowest possible power and with the least possible noise. MIMO-
OFDM has been chosen for high data rate communications and widely deployed in many
wireless communication standards. The major drawback in OFDM signal transmission is
high PAPR. In previous, use clipping technique to tackle this problem. In this paper, use
EM-GAMP algorithm to reduce PAPR in considerable amount.
Index Terms— MIMO-OFDM, EM-GAMP, PAPR, MUI, CCDF.
I. INTRODUCTION
Communication is the act of conveying intended meanings from one group to another
through the use of mutually understood signs and semiotic rules. Wireless communication is the
transfer of information or power between two or more points that are not connected by an
electrical conductor. The most common wireless technologies use radio. . In the communication
world of today, high data rate information transmission along with high capacity and reliability
are just some of the requirements which modern system have to meet in order to provide a good
quality of service to the end user. The arenas where wireless communication systems are
deployed, signals usually suffer phenomenon like multipath delay, fading and Inter Symbol
Interference (ISI) due to the frequency selectivity of the channel at the receiver side, the result of
which is the poor performance and high probability of errors.
In order to overcome the above mentioned issues channel coding and equalization
techniques are implemented. But due to the cost of hardware and various technical issues like
delays in coding and equalization process, it is not feasible to employ these techniques where
desired bit rates and the reliability of data expectations are quiet high. The solution of this issue
is to implement an effective scheme like OFDM where the high bit rate over the frequency
selective channel is guaranteed to some extent.
MIMO-OFDM is an efficient technique for high throughput and Qos of upcoming
generation wireless communication systems [1]. MIMO also have to improve the energy
efficiency and power consumption with low power components. Major problem of multicarrier
transmission is high Peak-to-average Power Ratio of the transmit signal [2]. Digital-to-analog
converter and power amplifier are used at the transmitter to handle this high PAPR [3]. Many
techniques are used to reduce the PAPR such as clipping [4], selected mapping (SLM) [5] and
partial transmission sequence (PTS) [6]. PAPR reduction schemes are also used in MIMO
32

systems [7]. A new PAPR reduction method fast iterative truncation algorithm [8] was developed
for MIMO-OFDM systems. This algorithm has low convergence rate. So in this paper, use an
efficient algorithm to reduce PAPR in considerable amount. The proposed algorithm EM-GAMP
result shows a performance improvement than existing methods of both PAPR reduction and
computational complexity. The rest of this paper details are as follows. In Section II, system
model and PAPR reduction are explained in detail. Proposed system is developed in Section III.
Simulation results are provided in Section IV followed by conclusion in Section V.
II. SYSTEM MODEL
The system model of the MIMO-OFDM downlink scenario is depicted in Figure 1, where
the BS is assumed to have M transmit antennas and serve K independent single-antenna users (K
≪ M), and the total number of OFDM tones is N. Then normalize the data vector to satisfy E
{∥sn∥22} = 1. For each tone n ∈ τc
, set 𝑠 𝑛 = 0 𝑘×1such that no signal is transmitted in the guard
band. Since cooperative detection among users is often impossible, precoding must be performed
at the BS to remove multiuser interference (MUI). Usually, the signal vector on the nth tone is
linearly precoded as
𝑤 𝑛 = 𝑃𝑛 𝑠 𝑛 (1)
where 𝑤 𝑛 belongs to 𝐶 𝑀×1
is the precoded vector that contains symbols to be transmitted on the
nth sub-carrier through the M antennas respectively and 𝑃𝑛belongs to 𝐶 𝑀×𝐾
represents the
precoding matrix for the nth OFDM tone.
Fig 1. System model for MIMO-OFDM
After precoding, all precoded vectors are reordered to M antennas for OFDM modulation,
[𝑎1 … . . 𝑎 𝑀] = [𝑤1 … . . 𝑤 𝑁] 𝑇
(2)
Where 𝑎 𝑚belongs to 𝐶 𝑁×1
represents the frequency-domain signal to be transmitted from
the mth antenna. The time-domain signals are obtained through the inverse discrete Fourier
33

transform (IDFT), i.e., â 𝑚 = 𝐹 𝑁
𝐻
𝑎 𝑚, ∀m. Then, a cyclic prefix (CP) is added to the time-domain
samples of each antenna to eliminate inter symbol interference (ISI). Finally, these samples are
converted to analog signals and transmitted via the frequency selective channel.At the receivers,
after removing the CPs of the received signals, the DFT is performed to obtain the frequency-
domain signals.
III. PROPOSED SYSTEM
To facilitate proposed algorithm development, introduce a noise term to model the
mismatch between y and Ax, i.e.
y = Ax + ϵ (3)
where ϵ denotes the noise vector. To reduce the PAPR associated with each transmit
antenna, aim to find a quasi-constant magnitude solution to the above underdetermined linear
system. To encourage a quasi-constant magnitude solution, propose a hierarchical truncated
Gaussian mixture prior for the signal x. In the first layer, coefficients of x are assumed
independent of each other and each entry xi is assigned a truncated Gaussian mixture
distribution.
𝑝(𝑥𝑖) = {𝛱
𝒩(𝑥𝑖; 𝑣, 𝛼𝑖1
−1
)
𝜂𝑖1
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
+ (1 − 𝛱)
𝒩(𝑥𝑖; 𝑣, 𝛼𝑖2
−1
)
𝜂𝑖2
𝑖𝑓 𝑥𝑖
€ [-ν, ν] (4)
The second layer specifies Gamma distributions as hyper priors over the precision
parameters 𝛼1, 𝛼2
𝑝(𝛼1, 𝛼2; 𝑎, 𝑏) = ∏ ∏ 𝐺𝑎𝑚𝑚𝑎(𝛼𝑙𝑖⃒𝑎, 𝑏)𝐼
𝑖=1
2
𝑙=1 (5)
Where
𝐺𝑎𝑚𝑚𝑎 (𝛼 ⃒𝑎, 𝑏) = 𝛤(𝑎)−1
𝑏 𝑎
𝛼 𝑎−1
𝑒−𝑏𝛼
(6)
As a result, the associated entries xi will eventually lie on one of the two boundary points,
leading to a quasi-constant magnitude solution. A variational expectation maximization (EM)
strategy is employed for the Bayesian inference. In proposed model, z ≜{x,α1,α2,κ} are treated
as hidden variables. The noise variance β and the boundary parameter v are unknown
deterministic parameters, i.e., θ≜ {β, v}.It is straightforward to show that the marginal
probability of the observed data can be decomposed into two terms
ln p(y; θ) = F(q, θ) + KL(q║p) (7)
34

where
F(q, θ) = ∫ q(z)ln
p(y,z;θ)
q(z)
dz (8)
and
𝐾𝐿(𝑞║𝑝) = − ∫ 𝑞(𝑧)𝑙𝑛
𝑝(𝑧⃒𝑦;𝜃)
𝑞(𝑧)
𝑑𝑧 (9)
Where q(z) is any probability density function, KL(q║p) is the Kullback-Leibler
divergence between p(z|y; θ) and q(z). Since KL(q║p) ≥ 0, it follows that F(q, θ) is a lower
bound of ln p(y; θ), with the equality holds only when KL(q║p) = 0, which implies p(z|y; θ) =
q(z). The EM algorithm can be viewed as an iterative algorithm which iteratively maximizes the
lower bound F(q, θ) with respect to the distribution q(z) and the parameters θ. Consider the
calculation of q(x).
Since the variables {xi} in the joint likelihood function p(y|x; β) are non-factorizable,
obtaining the posterior q(x) is rather difficult. To overcome this difficulty, we employ the
generalized approximate message passing (GAMP) technique to obtain an amiable
approximation of the joint likelihood function p(y|x; β).Here we approximate the joint likelihood
function p(y|x; β) as a product of approximate marginal likelihoods computed via the GAMP, i.e.
𝑝(𝑦⃒𝑥; 𝛽) ≈ 𝑝̂( 𝑦⃒𝑥; 𝛽)𝛼 ∏ 𝒩(𝑥𝑖⃒𝑟𝑖,̂ 𝜏𝑖
𝑟
)𝐼
𝑖=1 (10)
Where 𝒩(𝑥𝑖⃒𝑟𝑖,̂ 𝜏𝑖
𝑟
) is the approximate marginal likelihood obtained by the GAMP
algorithm PAPR is defined as the ratio of the peak power of the signal to its average power.
Specifically, the PAPR at the mth transmit antenna is defined as
PAPRM ≜
2N║âM║
2
║âM║2
2 (11)
Where the operator ║║
2
is used because RF-chains often process and modulate the real
and imaginary part of time domain samples independently.
The complementary cumulative distribution function (CCDF) is used to evaluate the
PAPR reduction performance. The CCDF denotes the probability that the PAPR of the estimated
signal exceeds a given threshold PAPR0, i.e.
CCDF (PAPR0) = Pr(PAPR > PAPR0) (12)
Also, to evaluate the multiuser interference of the transmit signals, we define the MUI as
MUI =
∑ ║sn−Hnwn║2
2
n∈τ
∑ ║sn║2
2
n∈τ
(13)
35

IV. SIMULATION RESULTS
OFDM is mainly designed to combat the effect of multipath reception, by dividing the
wideband frequency selective fading channel into many narrow flat sub channels. OFDM offers
flexibility in adaptation to time varying channel condition by adopting the parameters at each sub
carriers accurately. To avoid ISI due to multipath, successive OFDM path, successive OFDM
symbols are separated by guard band. This makes the OFDM system resistant to multipath
effects.
Figure 1 shows the original OFDM signal. It has lots of high peak signal. The peak of
original signal is continuously varying. So the largest magnitude produces high Peak to average
power ratio which intrudes the system performance.
Figure 2 shows the OFDM clipped signal. Clipping scheme is the method of clipping the
high peaks of the OFDM signal before passing it through the power amplifier (PA). This is done
with the help of clipper.
Figure 3 shows the EM-GAMP signal. Simulation result shows the proposed algorithm
achieves a substantial performance improvement over existing methods in terms of both the
PAPR reduction and computational complexity.
Figure 4 shows the CCDF of the PAPR in MIMO-OFDM system. In Existing system, the
zero forcing produces PAPR near to 10.6 dB and Clipping produces PAPR near to 6 dB. The
Proposed system produces PAPR near to 0.8dB.
Fig 2. Original signal
36

Fig 3. clipped signal
Fig 4. EM-GAMP signal
37

Fig 5. CCDF of PAPR
V. CONCLUSION
Considered the problem of joint PAPR reduction and multiuser interference (MUI)
cancellation in OFDM based massive MIMO downlink systems. A hierarchical truncated
Gaussian mixture prior model was proposed to encourage a low PAPR solution/signal. A
variational EM algorithm was developed to obtain estimates of the hyper parameters associated
with the prior model, as well as the signal. Specifically, the GAMP technique was embedded into
the variational EM framework to facilitate the algorithm development. The proposed algorithm
only involves simple matrix vector multiplications at each iteration and thus has a low
computational complexity. Simulation results show that the proposed algorithm achieves notable
improvement in PAPR reduction as compared with the Zero forcing and clipping-filtering
algorithm and meanwhile renders better MUI cancellation and lower out-of-band radiation. The
proposed algorithm also demonstrates a fast convergence rate, which makes it attractive for
practical real-time systems.
The EM-GAMP approach produce an efficient result as low PAPR compared to other
PAPR reduction techniques. Therefore, in future, the channel estimation is an area which
required a lot of attention and improper channel estimation degrades the performance of system.
In this work, it is assumed that channel is estimated perfectly. Hence one can evaluate the
performance of proposed work with different channel estimation method. To reduce the PAPR
other precoding techniques may be used for the future work.
38

REFERENCES
[1] F. Rusek, D. Persson, B. K. Lau, E. G. Larsson, T. L. Marzetta, O. Edfors, and F. Tufvesson, “Scaling up
MIMO: Opportunities and challenges with very large arrays,” IEEE Signal Process. Mag., vol. 30, no. 1,
pp. 40–60, Jan. 2013.
[2] G. Wunder, R. F. Fischer, H. Boche, S. Litsyn, J. No, "The PAPR problem in OFDM transmission: New
directions for a long-lasting problem", IEEE Signal Process. Mag., vol. 30, no. 6, pp. 130-144, Jan. 2014.
[3] T. Jiang and Y.Wu, “An overview: Peak-to-average power ratio reduction techniques for OFDM signals,”
IEEE Trans. Broadcasting, vol. 54, no. 2, pp. 257–268, June 2008.
[4] R. ONeill and L. B. Lopes, “Envelope variations and spectral splatter in clipped multicarrier signals,” in
Proc. IEEE PIMRC 95, Toronto, Canada, Sept. 1995, pp. 71–75.
[5] R. W. Bäuml, R.F. Fischer, J. B. Huber, "Reducing the peak-to-average power ratio of multicarrier
modulation by selected mapping", IEE Electron. Lett., vol. 32, no. 22, pp. 2056-2057, Oct. 1996.
[6] S. H. Muller and J. B. Huber, “OFDM with reduced peak-to-average power ratio by optimum combination
of partial transmit sequences,” IEE Elec. Letters, vol. 33, no. 5, pp. 368–369, Feb. 1997.
[7] S. H. Han, J. H. Lee, "An overview of peak-to-average power ratio reduction techniques for multicarrier
transmission", IEEE Wireless Commun., vol. 12, no. 2, pp. 56-65, Apr. 2005.
[8] C. Studer and E. G. Larsson, “PAR-aware large-scale multi-user MIMO-OFDM downlink,” IEEE J. Sel.
Areas Commun., vol. 31, no. 2, pp.303–313, Feb. 2013.
[9] D. Wulich, “Definition of Efficient PAPR in OFDM”, IEEE Communications Letters, vol. 09, no. 9, pp.
832_834, 2005.
[10]J. Vila and P. Schniter, “Expectation-maximization Gaussian-mixture approximate message passing,” IEEE
Trans. Signal Process., vol. 61, no. 19, pp. 4658–4672, Oct 2013.
[11]A.Venkataraman, H. Reddy, and T. M. Duman, “Space-Time Coded OFDM with Low PAPR”, EURASIP
Journal on Applied Signal Processing, pp. 1_9, 2006.
[12]D. G. Tzikas, A. C. Likas, and N. P. Galatsanos, “The variational approximation for Bayesian inference,”
IEEE Signal Process. Mag., vol. 25, no. 6, pp. 131–146, Jan. 2008.
[13]M. Tipping, “Sparse Bayesian learning and the relevance vector machine,” Journal of Machine Learning
Research, vol. 1, pp. 211–244, 2001.
39

Performance Analysis of PAPR Reduction in MIMO-OFDM

More Related Content

What's hot (19)

Viewers also liked (14)

Similar to Performance Analysis of PAPR Reduction in MIMO-OFDM (20)

More from IJARBEST JOURNAL (20)

Recently uploaded (20)

Performance Analysis of PAPR Reduction in MIMO-OFDM