Noname manuscript No.

(will be inserted by the editor)

Analysis of Adaptive Modulation and Coding with

Channel Estimators for WiMAX

B. Siva Kumar Reddy · Dr. B. Lakshmi

Received: date / Accepted: date

Abstract WiMAX has adopted Adaptive Modulation and Coding (AMC) in

OFDM to realize higher data rates and error free transmission. AMC schemes
utilize the channel state information (CSI) to use the channel effectively and
maximize the throughput. In this paper, LSE (Least Square Error), MMSE
(Minimum Mean square Error) and LMMSE (Least Minimum Mean square
Error) channel estimators are suggested and BER execution has been analyzed
over multipath fading channels(AWGN, Rayleigh and Rician fade channels).
The BER execution and throughput aspects are applied as a measure for AMC
execution and explored the Target-BER technique for AMC plots by acquiring
BER Vs. SNR characteristics. Simulation results have demonstrated that the
increment in modulation order causes to increment in throughput and BER
values. These results derived a trade-off among modulation order, FFT length,
throughput, BER esteem and spectral efficiency. The BER changes gradually
for AWGN channel and arbitrarily for Rayleigh and Rician fade channels.
Keywords AMC · CSI · LMMSE · LSE · MMSE · OFDM

1 INTRODUCTION

The growing development in the utilization of advanced systems has headed to

the interest for the design of new communication networks with higher capac-
ity. WiMAX (Worldwide Interoperability for Microwave Access) [1] is particu-
larly mainstream in wireless applications because of its restriction to patterns
Research Scholar
Department of Electronics and Communication Engineering,
National Institute of Technology Warangal, Andhra Pradesh-506004, India.
E-mail: bsivakumar100@gmail.com
Associate Professor
Department of Electronics and Communication Engineering,
National Institute of Technology Warangal, Andhra Pradesh-506004, India.
E-mail: lkodali93@gmail.com
2 B. Siva Kumar Reddy, Dr. B. Lakshmi

of interference and degradation (multipath and delay spread). A transmitted

symbol might achieve the recipient on the following symbol period and Inter
Symbol Interference (ISI) [2]. Due to the symbol time is shorter at higher data
rates; subsequently it expects just a smaller delay to induce ISI. Equaliza-
tion [3] is the formal strategy for taking care of ISI but at high data rates
requires an excess of handling power. OFDM (Orthogonal frequency-division
multiplexing) [4] has turned into the best answer for mitigating ISI in broad-
band frameworks, incorporating WiMAX to lend different guard intervals.
This guard interval removes ISI and ICI (Inter Carrier Interference) [5] by
maintaining orthogonality among all sub carriers.

OFDMA (Orthogonal Frequency Division Multiple Access) [6] is included

with Adaptive Modulation and Coding (AMC) methods [7] [8] to enhance
throughput, spectral efficiency [9] and BER performance of wireless communi-
cation system by adjusting transmission parameter like modulation code-rate
based on the channel state adaptively [10]. When adaptive transmission is em-
ployed to OFDM framework, transmission parameters of every subcarrier are
adjusted to every subchannel. To acquire different code-rate for every subcar-
rier, Rate-Compatible Punctured Convolutional (RCPC) code [11] is utilized.
Received signal-to-noise ratio (SNR) is applied for measure of subchannel state
to select proper modulation and code-rate for every subcarrier. Transmission
power and modulation for every subcarrier is resolved for decrease of the adap-
tation algorithm.

Lately advancement in the field of communication main efforts have been

guided towards improving efficient channel estimation technique [12]- [16] for
utilization in wireless communication frameworks utilizing different modula-
tion and transmission schemes. The significant challenges in wireless communi-
cation are to acquire exact channel state information (CSI) at the receiver end
of the framework for detection of the transmitted signal. In this paper, LSE
(Least Square Error), MMSE (Minimum Mean Square Error) and LMMSE
(Least Minimum Mean square Error) channel estimators [15] are utilized to
give randomly varying channel state information (CSI) to the transmitter
through a feedback, indicated in Fig. 1.

Figure 1....

The rest of the paper is structured as follows. Section 2 describes the basic
OFDM transceiver system and explains clearly the operation of the each block
in the system model. Section 3 narrates basic channel estimation methods and
gives the mathematical analysis for suggested channel estimators. Section 4
introduces the importance of Adaptive Modulation Coding scheme in recent
wireless communication systems. Section 5 reports the results obtained for
WiMAX specifications and derives the trade-off cases among various param-
eters. Finally, the manuscript has been concluded in the Section 6 from the
results derived from Section 5.
Title Suppressed Due to Excessive Length 3

2 OFDM TRANSCEIVER SYSTEM MODEL

Recently OFDM is essentially utilized in broadband remote access due to

its robust against multipath [17], maximizes throughput, robustness against
narrowband interference [18], establishes single-frequency network, frequency
diversity [19]. In OFDM, the sub carriers are not divided but overlapped or-
thogonally. Because of this overlapping, the bandwidth is considerably more
efficiently employed than in the traditional multicarrier frameworks [20]. Hypo-
thetically, m-ary digital modulation schemes utilizing OFDM can accomplish
a bandwidth efficiency [21], defined as bit rate per unit bandwidth, of log2M
bits/sHz (M is the modulation scheme size).
As pointed in Fig. 1, a convolutional encoder [22] first encodes the binary
input data in the transmitter section. In the wake of interleaving, the binary
values are constituted on Quadrature Amplitude Modulation (QAM) values.
In order to adjust the signal in the receiver for a possible phase shift, pilot
carriers can be embedded. In the Serial to Parallel obstruct, the serial QAM
input symbol-stream is changed over to a parallel stream with width equivalent
to the amount of subcarriers. These parallel symbols are modulated onto the
sub carriers by applying the Inverse Fast Fourier Transform. Note that with a
specific end goal to get output spectrum with a relative low out-of-band radia-
tion, the size of the IFFT can be picked bigger than the number of subcarriers
that is really used to transmit the data. After the IFFT block, the parallel
output will be changed over back to serial and cyclic prefix of the time domain
samples is added to take out ISI. IEEE 802.16 grants the insertion of guard
time intervals of a few lengths such 0.25, 0.125, 0.0625 and 0.03125 is lent to
the WiMAX symbol earlier it is transmitted. The guard interval is more than
the maximum delay time and after that the embedded guard interval signal is
D/A converted and passed through frequency selective fading channels.
In the receivers, the guard interval is removed and the inverse transform-
ing could be carried out to transmitter like time domain signals are changed
over by the FFT into complex symbols. The complex symbol is demodulated
adaptively utilizing modulation level information. Demodulated symbols are
block deinterleved and sent to Viterbi decoder and afterwards the bits will be
sent to a particular client and afterwards extracted using the required bit rate
information of client.

3 RAYLEIGH FADING CHANNEL

In wireless telecommunications, multipath is the radio signals attaining the

receiving antenna by two or more paths. The multipath may be because of at-
mospheric ducting, reflection and refraction from diverse objects, for example,
mountains and buildings. The results of multipath include constructive and
destructive interference and phase shifting of the signal. This comes about
Rayleigh fading [23]. The standard statistical model of this manages a distri-
4 B. Siva Kumar Reddy, Dr. B. Lakshmi

bution reputed to be Rayleigh distribution. In the event that all the signals
arriving the receiver are reflected signals, then i.e. Rayleigh fading.
The bit error probability in the presence of channel (h) is [14],
s 
1 2
|h| Eb  1 √
Pb/h = erf c  = erf c( γ) (1)
2 N0 2
2
where γ = Effective bit energy to noise ratio = |h|N0Eb
For a Rayleigh channel, the BER using BPSK modulation is given by
Z ∞  
1 √ − γ
Pb = erf c γe ( ) dγ
2γ̄ 0 γ̄
 r 
1 γ̄
= 1− (2)
2 γ̄ + 1
" s #
1 (Eb /N0 )
= 1−
2 (Eb /N0 ) + 1

3.1 Rayleigh Exponential Distribution

For some type of scattering environments, gI (t) and gQ (t) are at any time t1 are
iid Gaussian random variables with zero mean and variance b0 = E[gI2 (t1 )] =
2
E[gQ (t1 )]. This mainly occurs in NLOS regions, in rich scattering environments
isotropic antennas are used. Under such conditions, it causes to ”Rayleigh
fading” [14]. q
The envelop of received signal α = |g(t1 )| = E[gI2 (t1 )] + E[gQ
2 (t )] is
1
Rayleigh distributed at any time t1 i.e.,
x x2 2x x2
Pα (x) = exp{− }= exp{− }x ≥ 0 (3)
b0 2b0 Ωp Ωp
Where Ωp = E(α2 ) = E[gI2 (t1 )] + E[gQ2
(t1 )] = 2b0 is the average envelop
2
power. The squared envelop α at any time t1 has the exponential distribution
is
1 x
Pα2 (x) = exp(− ) (4)
Ωp Ωp

4 RICIAN CHANNEL

A Rician model is found in a framework with LOS propagation and scattering.

The model is specified by the Rician element (k), which is the ratio of the line
of sight and the scatter power components. The occurrence of Rayleigh and
Rician faded channels by differing Rician element (k) is depicted in the Fig.
2. Rician component is changed in accordance with make both channels event
are equivalent.
Figure 2....
Title Suppressed Due to Excessive Length 5

5 CHANNEL ESTIMATION

Channel Estimation is the procedure of characterising the effect of the physical

medium on the input sequence. The primary point of Channel Estimation is
to evaluate the effect of the channel on known or in part known set of trans-
mittances. OFDM Systems are particularly provided for channel estimation.
The subcarriers are closely spaced. Channel is estimated dependent upon the
preparation arrangement which is known to both transmitter and receiver. The
receiver can utilize the known training bits and the respective received samples
for estimating the channel. In general a few channel estimation techniques are
developed, which could be utilized as a part of different technologies of wireless
communication system. The estimation methods could be separated into three
classes: 1)Pilot-assisted channel estimation methods, 2)Blind and semi-blind
estimation methods and 3)Decision directed channel estimation techniques.
Pilotassisted Channel Estimation Technique , which is also known to be
trainingbased channel estimation technique, is a conventional method for ac-
quiring channel estimation for correspondence systems. In these methods,
training sequences of information known to the receiver are multiplexed with
the transmitted data symbols at a decided position before transmission. Such
training data are utilized at the recipient for estimating the CSI relating to
its position. The CSI relating to the data information position is at that point
acquired by method of interpolating between different channel estimates prior
acquired from the training data sequences.

5.1 LSE channel estimation

LSE estimator reduces the square error between estimation and detection to
estimate channel h[m]. In matrix form, the actual output could be composed
as y = xh and the error is e = ȳ − y. (ȳ is the expected output).
The square error (S) can be defined as [12]- [16]

S = |e|2
= (ȳ − y)2
(5)
= (ȳ − y) ∗ (ȳ − y)t
= (ȳ − Xh) ∗ (ȳ − Xh)t

Where t stands for complex transpose of matrix.

The equation can be minimized by taking its derivative w.r.t h and equating
it equals to zero. The final equation is [12],

h̄ = (X t X)−1 X t y (6)

Where h̄ = hls = X −1 y This equation can be implemented on SISO as well

as MIMO systems.
6 B. Siva Kumar Reddy, Dr. B. Lakshmi

5.2 MMSE channel estimation

The MMSE estimator minimizes the mean-square error.

Mean square error = mean(ȳ − y)2 = E(ȳ − y)2 . Concept of expected value
and correlation can be used to derive the equations for finding the channel
response.
The estimated channel is

Hmmse = F ∗ (RgY ∗ RY−1Y ∗ Y ) (7)

Where RgY = Rgg ∗ F 0 ∗ X 0 , RY Y = X ∗ F ∗ Rgg ∗ F 0 ∗ X 0 +variance of noise

∗ Identity matrix. The equation can be used for both ISO as well as MIMO
systems.

5.3 LMMSE channel estimation

In the presence of channel noise, LMMSE estimation of uncorrelated Gaussian

channel vector g is [12]- [16],
−1
ḡ = Γgy Γyy y, (8)

where Γgy = Γgg F H X H , Γyy = XF Γgg F H X H + σn2 IN For unique minimum

MSE, above covariance matrices should be positive definite. Γgy , Γyy are cross-
covariance and auto-covariance matrices and σ 2 is variance. In frequency do-
main, channel estimation is,

h̄mmse = F ḡ = F QF H X H y, (9)

where Q=Γ [(F H X X XF )−1 σn2 + Γgg ], F is DFT orthonormal.

6 ADAPTIVE MODULATION and CODING

AMC scheme is included in a communication system in such a way that SNR

or BER from received packets computed in receiver is feedback to transmitter
by Channel Quality Indicator (CQI) to reconfigure itself for the next packet
to be send [24]. The selection of the proper modulation and coding mode to
be connected in the next transmission is made by the transmitter dependent
upon the estimation of channel conditions i.e. instantaneous SNR for the fol-
lowing time interval of packets. For this selection of modulation and coding
rate plans, SNR thresholds are defined from singular execution of every plan
in the channel upholding target BER of 0.01 or 0.001. The AMC execution
depends upon the precise channel estimation at the receiver and the authentic
feedback path between the collector and the transmitter on which the receiver
accounts Channel State Information (CSI) [25]. In order for AMC scheme to
have an exceptional execution, the postponement between the quality channel
estimation and the actual transmission in connection to the morel Doppler
Title Suppressed Due to Excessive Length 7

frequency of the channel ought to be as low as possible. Likewise, the trans-

mitter need to alert the collector which modulation and coding rate plan is
being used during the transmission. Any other way, it will lead demodulation
and decoding of appropriated packets which will compromise the execution.

7 SIMULATION RESULTS

In this paper, the OFDM parameters are focused on WiMAX environment and
noted in Table 1. Specifically the consideration will be basically focused on
BER execution of Adaptive modulation schemes with Target-BER of 0.01 and
0.001. For AMC, channel state information is given by the LSE, MMSE and
LMMSE channel estimators. Fig. 3 shows the magnitude and phase response
of QPSK modulation scheme with FFT size 1024 for WiMAX. BER vs. SNR
qualities have been compared in the Tables 2, 3 for BPSK, QPSK, 16QAM and
64QAM modulation schemes for distinctive FFT sizes i.e. 256, 1024 and 2048.
It might be concluded that as increment in FFT length results to diminishing in
BER value. As modulation scheme order increases BER value is also increased
at any considered SNR value. But when SNR value incremented, BER esteem
is reduced. Hence there is a trade-off around modulation order, BER and SNR
values. The LSE and MMSE channel estimators reaction for the actual channel
response has indicated in Figs 4, 5 and 6.
Table 1....
Table 2....
Table 3....
Figure 3....
Figure 4....
Figure 5....
Figure 6....
OFDM transmitter which might be incorporated in creating physical layer
of WiMAX is made utilizing Matlab and AWGN channel is added to account
for noise in channel. At that point channel estimation is carried out by trans-
mitting pilot subcarriers in receiver and correspondingly OFDM recipient is
made. Fig. 7 shows the singular BER vs. SNR characteristics of different mod-
ulation plans to be utilized within WiMAX without channel coding. QPSK
has preferable BER performance than 64QAM . Thus, it might be presumed
that as BER value is incremented with modulation order. Complete physi-
cal layer of WiMAX framework as outlined in section 2 is constructed using
Matlab. Different MCS schemes i.e. 1/2 QPSK, 3/4 QPSK, 1/2 16-QAM, 3/4
16QAM, 1/2 64QAM and 3/4 64QAM are employed within transmitting and
error rate is processed on bit by bit basis for each iteration and average BER
is computed. Channel coding is embedded in wireless communication system
to increase overall error rate performance as indicated in Fig. 8. 3/4 64QAM
has worse BER performance than remaining modulation schemes. Code rate
and modulation plot impacts BER execution of the physical layer framework.
Fig. 8 depicts the throughput characteristics of different modulation schemes
8 B. Siva Kumar Reddy, Dr. B. Lakshmi

(QP SK, 16QAM, 64QAM ) with code rate 1/2 and 3/4 in the AWGN chan-
nel evaluated by plotting the Capacity vs SNR characteristics. From Figs 7,8
and 9; it might be concluded that 1/2 QPSK has better performance of BER
but lesser throughput or spectral efficiency than other modulation and coding
rate.
Figure 7....
Figure 8....
Figure 9....
Table 4....
The Target BER method employed for AMC plan is carried out remember-
ing the error rate under a target limit say 0.01 or 0.001, upholding a fixed level
of quality of service in terms of error probability. Table 4 presents the SNR
threshold to change different modulation and coding rate dependent upon SNR
for Target BER of 0.01 and 0.001. The framework continues working with the
most minimal modulation and coding scheme, in particular QPSK modulation
with coding rate 1/2, until the signal-to-noise ratio permits to respect the error
rate constraint, then the system switches on higher modulation transmission
schemes to yield a better spectral efficiency while maintaining our desired BER
target. Figs. 10 & 11 depicts the BER and throughput characteristics as func-
tion of SNR for a Target BER of 0.01. If input data streams are transmitted
employing 64QAM with coding rate of 3/4, the throughput will be expanded
however there will be lower BER performance. So by compromising slightly
the throughput execution, the regulation and coding rate plans are change to
keep error rate below our desired BER level.
Figure 10....
Figure 11....
The BER aspects and throughput attributes of AMC technique for Target
BER of 0.001 are demonstrated in Figs. 12 and 13 separately. Throughput
execution has been looked at for Target BER 0.01 and 0.001 is demonstrated
in Fig 14. It could be concluded that to realize higher throughput, the system
need to compromise in terms of BER. For lower SNR values (Up to 9 dB),
throughput execution is same for acknowledged Target BERs.
Figure 12....
Figure 13....
Figure 14....
Figs 15, 16 present the BER execution over Rayleigh & Rician fading chan-
nels for 16QAM modulation scheme with 1024 number of subcarriers. It could
be concluded that, without channel estimation the BER has steady execution
for all modulation schemes in both the channels. Thus, without channel estima-
tion the determination of proper modulation scheme is challenging dependent
upon channel conditions. BER values for Rayleigh and Rician fading chan-
nels with different modulation schemes (BP SK, QP SK, 16QAM, 64QAM and
256QAM ) over LSE channel estimator are arranged in Table 5. As modula-
tion size increments, the BER value likewise increases at a given SNR value,
which is a trade-off. BER value is diminished with SNR increment but in-
creasing SNR is a trade-off in low power systems. Fig 17 depicts the variety of
Title Suppressed Due to Excessive Length 9

mean square error for LS and MMSE channel estimators. This is a simulation
based verification highlighting the superiority of the MMSE channel estimator
over the LSE estimator. The two parameters of evaluation turn out to be,
quite logically, the SER [Symbol Error Rate] and the Mean Square Error. It
ought to be noted that the MMSE estimators have been determined under the
assumption of known as channel correlation and noise variance. In practice
these quantities, Rgg ; and Vi , are either taken fixed or estimated possibly in
an adaptive way. This will increase the estimator complexity and reduce the
performance slightly.
Figure 15....
Figure 16....
Figure 17....
Table 5....
BER execution is analyzed by varying channel model for LMMSE estima-
tor are noted in Table 6 by utilizing plots demonstrated within Figs 18, 19 and
20. BER is low for AWGN channel compared with Rayleigh and Rician fade
channels. The BER is reduced gradually for AWGN and varies arbitrarily for
remaining channels. Henceforth, modulation and coding rates are changed ef-
fectively for AWGN as indicated in Fig 8. BER execution for AWGN, Rayleigh
and Rician channels are demonstrated in Figs 10, 21 and 22 under Target BER
0.01 and relating values are noted in Table 7. It could be concluded that BER
performance is good (BER is low) for AWGN channel compared to considered
remaining multipath fading channels.
Table 6....
Table 7....
Figure 18....
Figure 19....
Figure 20....
Figure 21....
Figure 22....

8 CONCLUSION

This paper has investigated the execution of OFDM frameworks with AMC
as far as BER, SER and throughput. It is connected several adaptive modes,
code-rate and constellation size relying upon channel conditions and the trans-
mit power. Code-rate of every subcarrier was adapted to subchannel state to
maximize data rate and to satisfy average BER. To get different code-rate
and to reduce complexity, RCPC code was utilized for every subcarrier. For
the channel state data, LSE, MMSE and LMMSE channel estimators have
been suggested and BER execution for different modulation schemes is anal-
ysed by the LSE channel estimator. The MMSE estimator expects from the
priori knowledge of noise variance and channel covariance. Also, its complex-
ity is extensive compared with the LSE estimator. For high SNRs the LSE
estimator is both simple and adequate. The outcomes presumed that higher
10 B. Siva Kumar Reddy, Dr. B. Lakshmi

modulation conspires give higher data rate however gives higher BER esteem.
Lower modulation level plans give better BER execution but worse through-
put execution. Confined to limitations, results derived a trade-off around the
modulation order, BER, FFT length and throughput. Target BER has been
recognized to compromise the trade off between BER Vs. SNR qualities.
Title Suppressed Due to Excessive Length 11

Fig. 2 Rayleigh and Rician fade channels occurrence with Rician Factor (K).

Table 1 Parameter Definition.

Parameters values
Number of sub channels (N) 1024
Number of pilots (p=N/8) 128
Number of data subcarriers (N-P) 896
Guard intervel length (GI=N/4) 256
Pilot position intervel 8
Channel length 16
Channel AWGN, Rayleigh fade
Digital Modulations BPSK, QPSK, 8PSK,
16QAM, 64QAM, 256QAM.

Table 2 BER vs. SNR for BPSK, QPSK modulation schemes over LSE channel estimator.

SNR BPSK BPSK BPSK QPSK QPSK QPSK

db fft=256 1024 2048 256 1024 2048
0 0.5213 0.5000 0.4063 0.5000 0.4297 0.4609
5 0.5156 0.4375 0.4063 0.5781 0.4688 0.4219
10 0.4063 0.4531 0.4688 0.3984 0.5234 0.5000
15 0.4688 0.4228 0.3750 0.3828 0.4688 0.4219
20 0.4844 0.5156 0.4531 0.4844 0.4531 0.4766
25 0.5938 0.5781 0.5000 0.4844 0.4922 0.4922
30 0.4531 0.5000 0.3750 0.4844 0.4766 0.4804
12 B. Siva Kumar Reddy, Dr. B. Lakshmi

Table 3 BER Vs. SNR for 16QAM, 64QAM modulation schemes over LSE channel esti-
mator.
SNR 16QAM 16QAM 16QAM 64QAM 64QAM 64QAM
db fft=256 1024 2048 256 1024 2048
0 0.5313 0.5430 0.4609 0.4922 0.5156 0.4922
5 0.4883 0.4766 0.4180 0.4818 0.5026 0.4695
10 0.4805 0.5156 0.5195 0.4974 0.4657 0.4922
15 0.4492 0.4453 0.4727 0.4714 0.4635 0.4714
20 0.5078 0.4961 0.4766 0.4844 0.5078 0.5651
25 0.5273 0.5391 0.5313 0.5339 0.4896 0.4948
30 0.4648 0.4648 0.4844 0.4896 0.5078 0.4661

Fig. 3 Magnitude and Phase response for LSE channel estimator with FFT size 1024.

Fig. 4 Actual channel response.

Title Suppressed Due to Excessive Length 13

Fig. 5 Channel response by LSE channel estimator.

Fig. 6 Channel response by MMSE channel estimator.

Fig. 7 BER performance without channel coding.

14 B. Siva Kumar Reddy, Dr. B. Lakshmi

Fig. 8 BER performance of AMC system with convolutional channel coding.

Fig. 9 Throughput performance of different MCS schemes.

Table 4 Switching thresholds for AMC system.

MCS scheme SNR thresholds(dB) SNR thresholds(dB)

Target-BER=0.01 Target-BER=0.001
1/2 QPSK <8 < 10
3/4 QPSK 8 − 10 10 − 11
1/2 QAM16 10 − 12 11 − 13
3/4 QAM16 12 − 16 13 − 18
1/2 QAM64 16 − 18 18 − 20
3/4 QAM64 > 18 > 20
Title Suppressed Due to Excessive Length 15

Fig. 10 BER performance of AMC system under Target-BER 0.01 over AWGN channel.

Fig. 11 Throughput performance of AMC system under Target-BER 0.01.

Fig. 12 BER performance of AMC system under Target-BER 0.001 over AWGN channel.
16 B. Siva Kumar Reddy, Dr. B. Lakshmi

Fig. 13 Throughput performance of AMC system under Target-BER 0.001.

Fig. 14 Throughput comparison of Target-BER 0.001 and BER 0.01.

Fig. 15 Rayleigh fading channel performance for 16-QAM modulation scheme.

Title Suppressed Due to Excessive Length 17

Fig. 16 Rician fading channel performance for 16-QAM modulation scheme.

Fig. 17 Mean square error performance for LS and MMSE channel estimators.

Table 5 BER Vs SNR performance for channels over LSE channel estimator.

Modulation Rayleigh Rayleigh Rician Rician

scheme SNR=30dB SNR=45dB SNR=30dB SNR=45dB
BPSK 0.0316 0.0010 0.0630 0.01
QPSK 0.0501 0.0015 0.0794 0.00154
16QAM 0.10115 0.00316 0.11748 0.01047
64QAM 0.10471 0.01584 0.13182 0.02691
256QAM 0.15848 0.06309 0.16595 0.1
18 B. Siva Kumar Reddy, Dr. B. Lakshmi

Fig. 18 BER performance of LMMSE channel estimator over AWGN channel.

Fig. 19 BER performance of LMMSE channel estimator over Rayleigh fade channel.

Table 6 BER Vs SNR performance for channels over LMMSE channel estimator.

Channel BER at BER at BER at BER at BER at BER at

SNR=5 dB 10 dB 15 dB 20 dB 25 dB 30 dB
AWGN 0.316 0.251 0.199 0.158 0.125 0.039
Rayleigh 0.506 0.497 0.501 0.490 0.510 0.504
Rician 0.505 0.496 0.506 0.500 0.495 0.484
Title Suppressed Due to Excessive Length 19

Table 7 BER Vs SNR performance for channels over LMMSE channel estimator under
Target BER=0.01.

Channel BER at Maximum Minimum

SNR=20 dB BER BER
AWGN 0.003 0.316 0.00003
Rayleigh 0.499 0.506 0.496
Rician 0.501 0.504 0.496

Fig. 20 BER performance of LMMSE channel estimator over Rician fade channel.

Fig. 21 BER performance of AMC system under Target-BER 0.01 over Rayleigh channel.
20 B. Siva Kumar Reddy, Dr. B. Lakshmi

Fig. 22 BER performance of AMC system under Target-BER 0.01 over Rician channel.

References

1. Ergen, Mustafa, ’Mobile broadband: including WiMAX and LTE’, Springer, 2009.
2. Rusek, Fredrik and Fertonani, Dario, ’Bounds on the information rate of intersym-
bol interference channels based on mismatched receivers’, Information Theory, IEEE
Transactions on, 2012, vol 58, num 3, pages 1470–1482.
3. Huang, Jianzhong and Zhou, Shengli and Huang, Jie and Berger, Christian R ’Progres-
sive inter-carrier interference equalization for OFDM transmission over time-varying
underwater acoustic channels’ Selected Topics in Signal Processing, IEEE Journal of,
2011, vol 5, num 8, pp.1524–1536.
4. Rabiei, Payam and Namgoong, Won and Al-Dhahir, Naofal, ’A non-iterative technique
for phase noise ICI mitigation in packet-based OFDM systems’, Signal Processing,
IEEE Transactions on, 2010, vol 58, num 11, pp.5945–5950.
5. Wu, Hsiao-Chun and Huang, Xiaozhou and Xu, Dongxin, ’Novel semi-blind ICI equal-
ization algorithm for wireless OFDM systems’, Broadcasting, IEEE Transactions on,
2006, vol 6, num 2, pp.211–218.
6. Salem, Mohamed and Adinoyi, Abdulkareem and Yanikomeroglu, Halim and Falconer,
David ’Opportunities and challenges in OFDMA-based cellular relay networks: A ra-
dio resource management perspective’, Vehicular Technology, IEEE Transactions on,
2010, vol 59, num 5, pp. 2496–2510.
7. Reddy, B Siva Kumar and Lakshmi, B ’Adaptive Modulation and Coding in COFDM
for WiMAX Using LMS Channel Estimator’, World Conference on Advanced Commu-
nications and Control systems, 2013.
8. Fantacci, Romano and Marabissi, Dania and Tarchi, Daniele and Habib, Ibrahim,
’Adaptive modulation and coding techniques for OFDMA systems’,Wireless Commu-
nications, IEEE Transactions on, 2009, vol 8, num 9, page 4876–4883.
9. Dai, Linglong and Wang, Zhaocheng and Yang, Zhixing, ’Time-frequency training
OFDM with high spectral efficiency and reliable performance in high speed environ-
ments’,Selected Areas in Communications, IEEE Journal on, 2012, vol 30, num 4,
page 695–707.
10. Choi, Dongwook and Lee, Dongwoo and Lee, Jae Hong, ’Resource allocation for CoMP
with multiuser MIMO-OFDMA’,Vehicular Technology, IEEE Transactions on, 2011,
vol 60, num 9, pp 4626–4632.
11. Ahn, Chang-Jun and Kamio, Takeshi and Fujisaka, Hisato and Haeiwa, Kazuhisa ’In-
fluence of laser line width and modulation level for coherent optical OFDM with punc-
tured LDPC codes’,WSEAS Transactions on Communications, 2011, vol 10, num 6,
pp 175–181.
Title Suppressed Due to Excessive Length 21

12. Qiao, Yantao and Yu, Songyu and Su, Pengcheng and Zhang, Lijun, ’Research on an
iterative algorithm of LS channel estimation in MIMO OFDM systems’,Broadcasting,
IEEE Transactions on, 2005, vol 51, page 149–153.
13. Minn, Hlaing and Bhargava, Vijay K, ’An investigation into time-domain approach for
OFDM channel estimation’,Broadcasting, IEEE Transactions on, 2000, vol 46, num 4,
page 240–248.
14. Chang, Ming-Xian and Su, Yu T, ’Model-based channel estimation for OFDM signals
in Rayleigh fading’,Communications, IEEE Transactions on, 2002, vol 50, num 4, page
540–544.
15. Petropulu, Athina and Zhang, Ruifeng and Lin, Rui, ’Blind OFDM channel estimation
through simple linear precoding’,Wireless Communications, IEEE Transactions on,
2004, vol 3, num 2, page 647–655.
16. Cai, Xiaodong and Giannakis, Georgios B, ’Adaptive PSAM accounting for channel
estimation and prediction errors’,Wireless Communications, IEEE Transactions on,
2005, vol 4, num 1, page 246–256.
17. Kyro, Mikko and Haneda, Katsuyuki and Simola, Jarno and Nakai, Kenji and Tak-
izawa, ’Measurement based path loss and delay spread modelling in hospital environ-
ments at 60 GHz’,Wireless Communications, IEEE Transactions on, 2011, vol 10,
num 8, pp 2423–2427.
18. Dardari, Davide and Conti, Andrea and Ferner, Ulric and Giorgetti, Andrea and
Win, Moe Z, ’Ranging with ultra wide bandwidth signals in multipath environ-
ments’,Proceedings of the IEEE, 2009, vol 97, num 2, pp 404–426.
19. Ghaznavi, Mahsa and Jamshidi, Azizollah, ’Interference impact on the outage capacity
of a frequency diversity paradigm in cognitive radio networks’,Communications, IET,
2012, vol 6, num 2, pp 179–186.
20. Taubock, Georg and Hampejs, Mario and Svac, Pavol and Matz, Gerald and Hlawatsch,
’Low-complexity ICI/ISI equalization in doubly dispersive multicarrier systems using
a decision-feedback LSQR algorithm’,Signal Processing, IEEE Transactions on, 2011,
vol 59, num 5, pp 2432–2436.
21. Ku, Ivan and Wang, Cheng-Xiang and Thompson, John, ’Spectral, energy and eco-
nomic efficiency of relay-aided cellular networks’, IET Communications, 2013, vol 7,
num 14, pp 1476–1486.
22. Marazin, Mélanie and Gautier, Roland and Burel, Gilles,’Algebraic method for blind
recovery of punctured convolutional encoders from an erroneous bitstream’,Signal Pro-
cessing, IET, 2012, vol 6, num 2, pp 122–131.
23. Reddy, Kumar and Siva, B and Lakshmi, B,’Channel estimation and equalization in
OFDM receiver for WiMAX with Rayleigh distribution’,Advanced Electronic Systems
(ICAES), 2013 International Conference on, 2013, pp 337–339.
24. Zhang, Jian A,’Response to” On Mathematical Equivalence Between Vector OFDM
and Quadrature OFDMA”’,Communications, IEEE Transactions on, 2013, vol 61,
num 2, pp 815–815.
25. Smida, Besma and Hanzo, Lajos and Affes, Sofiene,’Exact BER performance of asyn-
chronous MC-DS-CDMA over fading channels [Transactions Letters]’,Wireless Com-
munications, IEEE Transactions on, 2010, vol 9, num 4, pp 1249–1254.