Novel Channel Estimation Algorithm using Various Filter Design in

LTE-Advanced System

Saransh Malik, Sangmi Moon , Bora Kim, Cheolhong Kim and Intae Hwang

Department of Electronics and Computer Engineering, Chonnam National University,

300 Yongbong-Dong, Buk-Gu, Gwangju, Republic of Korea

Keywords: Channel Estimation, OFDM, LTE-Advanced.

Abstract: Channel estimation is a major issue in communication system. In this paper, we propose a new idea for

channel estimation that uses a Kalman Filter (KF) approach to predict the channel in OFDM symbols with

pilot subcarriers where channel affected is by high doppler spread. We design the algorithm considering the

lattice-type arrangement of pilot subcarriers in a LTE-Advanced system from 3GPP. In further advancement,

we use the filtering of channel impulse response and application of a Wiener Filter for the estimation of the

channel frequency response in the rest of the subcarriers.

1 INTRODUCTION

There are several techniques used to estimate the

channel in a wireless communication system. The

effectiveness of these techniques depends on key

factor such as the environment (indoor, outdoor,

rural, urban) and the mobility between transmitter

and receiver. Primarily, least square (LS) estimation

is the most simple of the conventional channel

estimations techniques, but it only works well for

time-invariant channels or with low Doppler spread.

Minimum mean squared error (MMSE) combined

with a linear interpolation method is more effective

than LS because it considers the channel correlation

values. However for time-variant channels, it is

necessary to implement techniques able to estimate

or predict the variations of the channel in time, such

as Wiener and Kalman filters. In this research, we

want to proposed a method to estimate and predict

the variations of channel in fast fading channels,

considering the especial arrangement of reference

signals in LTE, since other papers are usually based

on block-type or comb-type arrangements.

This paper is divided as follows, section II shows

the arrangement of reference signals in 3GPP-LTE

advanced, section III briefly describes conventional

channel estimation methods, section IV describes

our proposal, section V shows our simulation results

and finally section VI contains our conclusions.

2 REFERENCE SIGNAL IN LTE-

ADVANCED

LTE-Advanced uses signals known by both the

transmitted and the receiver, sent in predefined

locations. This signals are called downlink reference

signals (3 GPP TS 36.211, 2010). By processing the

received reference signals, the receiver can estimate

the whole channel response for each OFDM symbol.

In the time-domain, for one antenna port,

reference signals are transmitted during the first and

fifth OFDM symbols of each slot when the normal

cyclic prefix (CP) is used and during the first and

fourth OFDM symbols when the extended CP is

used. In the frequency-domain, they are inserted

every six subcarriers.

2.1 Relationship between Coherence

Bandwidth and Reference Signals

In 3GPP LTE

Table 1: Delay profile for 3GPP LTE channel model.

Channel

Number of

Chanel Taps

Delay Spread

(r.m.s)

Max. Excess

Tap delay

EPA 7 45ns 410 ns

EVA 9 357ns 2510 ns

ETU 9 991ns 5000 ns

120

Malik S., Moon S., Kim B., Kim C. and Hwang I..

Novel Channel Estimation Algorithm using Various Filter Design in LTE-Advanced System.

DOI: 10.5220/0004026901200125

In Proceedings of the International Conference on Signal Processing and Multimedia Applications and Wireless Information Networks and Systems

(SIGMAP-2012), pages 120-125

ISBN: 978-989-8565-25-9

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

Figure 1: Mapping of Downlink reference signals for one

antenna port in extended CP.

Table 2: Coherence bandwidth of 20 MHz LTE-Advanced

system.

Channel B

(0.9)

(0.5)

EPA 1.6 MHz 3.7 kHz

EVA 201.1 kHz 466.9 kHz

ETU 72.4 kHz 168.2 kHz

In LTE-Advanced, the subcarrier space frequency Δf

is 15 kHz. Since the reference signals in the

frequency domain are located every 6 subcarriers,

the reference signal bandwidth is 90 kHz. If the

channel frequency response remains constant, the

correlation in the frequency domain should have a

value close to 1; while we choose a value of 0.5 as

an indicator that the channel has changed. Based on

this assumptions, we can estimate the coherence

bandwidth for both situations using equation (1)

(Somasegeran, 2007):

)arccos(

πτ

≥

(1)

where c is the correlation in frequency domain, and

is the root-mean-square (RMS) delay spread given

for LTE extended channel models: Extended

Pedestrian A (EPA), Extended Vehicular A (EVA)

and extended urban (ETU) channels (Table 1).

The results in Table 2 demonstrate that, basically,

the channel is constant during the bandwidth

corresponding to the spacing of reference symbols in

frequency domain; and therefore, the reference

signals in LTE-Advanced are able to keep track of

the frequency selective fading.

2.2 Relationship between Coherence

Time and Reference Signals in

3GPP LTE

For the extended cyclic prefix case, there are 6

symbols in 1 slot (12 symbols per sub frame).

Therefore, we can calculate 1 OFDM symbol period

as:

symb

slot

symb

3.83

5.0

===

Reference signals are transmitted during the first and

fourth OFDM symbols in each slot. Therefore, the

spacing of reference signals in time domain is:

msTT

symbref

25.0.3 ==

Table 3: Maximum doppler frequency for 3 GPP LTE–

Advanced channel models.

Channel

Maximum Doppler

Frequency (Hz)

UE Maximum Speed (Km/h)

*carrier frequency 2 GHz

EPA 5 3

EVA 70 40

ETU 300 160

If the channel is constant in time, the correlation

in the time domain should be close to 1; as such we

choose a value of 0.5 as an indicator that the channel

has changed. Based on these assumptions, we can

estimate the coherence time for both situations,

using equation (2) and the maximum Doppler

frequency corresponding to the LTE-Advanced

channel models as shown in Table 3.

)arccos(

=Δ

The results shown in Table 4 demonstrate that the

spacing of reference symbols in time is less than the

coherence time; therefore, the reference signals in

LTE are able to keep track of the time-varying

fading channel.

Table 4: Coherence time 20 MHz 3GPP LTE-Advanced

system.

Maximum Doppler Frequency Δt

(0.9)

Δt

(0.5)

= 5 Hz 14.4 ms 33.3 ms

= 70 Hz 1.0 ms 2.4 ms

= 300 Hz 0.2 ms 0.6 ms

The results shown in Table 4 demonstrate that

the spacing of reference symbols in time is less than

the coherence time; therefore, the reference signals

in LTE are able to keep track of a time-varying

fading channel.

NovelChannelEstimationAlgorithmusingVariousFilterDesigninLTE-AdvancedSystem

121

3 CONVENTIONAL CHANNEL

ESTIMATION METHODS

In this section we briefly describe conventional techniques

to estimate and/or predict the channel frequency response

(CFR)

3.1 LS Channel Estimation

LS is the simplest method to estimate the channel.

First, we calculate the CFR only in the subcarriers

that contain reference symbols. We do so, by

dividing the received reference signals between their

transmitted value (Simko, 2009):

()

(

)

()

⎥

⎦

⎤

⎢

⎣

⎡

∧

LSp

H ...

(3)

The channel frequency response for the rest of the

subcarriers, is estimated with interpolation and

extrapolation in frequency and time dimensions.

Figure 2: Block Diagram of LS Channel Estimation

3.2 MMSE Channel Estimation

Another method to estimate the channel is the

MMSE algorithm which has better performance than

LS but is computationally

more complex. MMSE

calculates the channel impulse response that

minimizes the mean square error between the actual

and estimated channel impulse response (Hou and

Liu).

The channel in the frequency domain can be

estimated with equation (4):

[]

LSp

pphp

MMSEp

HXXRRH

)(

∧

−

∧

(4)

where is the noise variance, is the cross-

correlation matrix between all subcarriers and the

subcarriers with reference signals within the same

OFDM symbol,

is the autocorrelation matrix of

the subcarriers with reference signals within the

same OFDM symbol, and the superscript

denotes Hermitian transpose. By replacing

in (4) with its expectation , the

MMSE channel estimator in frequency domain can

be expressed as:

LSp

ppphp

MMSEp

SNR

RRH

∧

−

∧

⎥

⎦

⎤

⎢

⎣

⎡

(5)

where is a constant depending on the type of

modulation and β is the identity matrix. We can

estimate the channel for the other resource elements

using linear interpolation in time domain.

Figure 3: Block diagram of MMSE channel estimation.

3.3 Wiener Filter

The Wiener Filter (WF) uses the same principle than

MMSE method (Qin et al., 2007) and it also

eliminates noise effects in the channel estimation.

WF allows us to keep track of the variations of the

CIR in time-variant channels because it uses both

the time and frequency correlations.

To simplify the complexity, the 2-dimensional

WF is decompose into two separated WF's; one in

the frequency domain and one in the time domain.

First, we obtain directly the channel estimation in

frequency domain for the OFDM symbols with

reference signal as:

LSp

MMSEp

SNR

RRH

∧

−

∧

⎥

⎦

⎤

⎢

⎣

⎡

(6)

Then we estimate the total channel frequency for all

OFDM symbols using WF in time domain:

SNR

RRH

∧

−

∧

⎥

⎦

⎤

⎢

⎣

⎡

(7)

Figure 4: Block diagram 2x1 dimensional WF.

3.4 Kalman Filter

The purpose of the Kalman Filter (KF) is to use

measurements observed over time, containing noise,

and produce values that are closer to the true values

of the measurements and their associated calculated

SIGMAP2012-InternationalConferenceonSignalProcessingandMultimediaApplications

122

values. In this section, we study the KF that is used

to predict variations of the channel in time domain

and which can also be applicable to frequency

domain.

The principle behind KF applied to channel

estimation is that we can represent the channel

frequency response, at time, as an infinite order

autoregressive (AR) process (Karakaya et al., 2009).

Figure 5: System using KF and block type pilot

arrangement.

Where, k and A are the order and the coefficient

of the AR process, respectively. V is a white

Gaussian noise with zero mean and variance σ

. For

the case of first order AR process

and .

In order to reduce the computational complexity,

only the first order AR process, i.e., is considered.

Therefore, we can represent the vector form of the

channel frequency response at time n as:

Then, the channel estimate can be obtained by a

set of recursions (Ling and Ting, 2006):

The received symbol at time can be expressed in the

form of a linear regression model.

4 PROPOSED CHANNEL

ESTIMATION SCHEME

In this section introduce and explain the proposed

method to predict the channel in high Doppler

spread.

Fig. 6 shows the New Channel Estimator's block

diagram that was designed considering the lattice-

type reference signals arrangement of 3GPP LTE (so

far most of the studies focus on block-type or comb

type pilot arrangements).

Figure 6: Block diagram of the new channel estimator.

After FFT, the reference signals or pilots of the

first OFDM symbols are extracted and we can

estimate the channel frequency response (CFR)

using a simple method, such as LS. Using the

recursions of KF, we can estimate the variation of

CFR for the later OFDM symbols containing

reference signals. Then, we transform the CFR into

CIR and eliminate the taps with index larger than L;

this way we eliminate the noise contained in those

taps. Finally, we transform the CIR to CFR and

estimate the channel for the rest of the subcarriers

using WF in time dimension.

5 SIMULATION RESULTS

In this section we show the performance analysis of

channel estimation methods. The simulations are

performed in MATLAB using the simulation

parameters of 3GPP LTE-Advanced shown in Table

5. The time variant channel is modelled according to

the values given for LTE extended channel models

in Table 1 and the maximum Doppler Frequency

values for each channel given in Table 3. The

frequency power spectrum follows the Jakes model.

Following the results of section II, we assume the

channel to be constant for 1 OFDM symbol.

Fig. 7 shows the BER performance of different

channel estimation methods in EPA channel with

max. Doppler Frequency of 5Hz. For this case, the

motion speed is low; therefore, as shown in sections

II, the channel suffers little variation in time within

one subframe and techniques like LS or MMSE with

linear interpolation produce good results.

Fig. 8 shows the BER performance of different

channel estimation methods in EVA channel with

max. Doppler frequency of 70Hz. In this case, the

mobile user moves with medium speed; therefore, as

shown in section II, the channel suffers more

variation in time within one subframe compared to

the case of EPA. We can observe that the BER

obtained with LS and MMSE starts to separate from

the actual value, but WF and New CE produce more

accurate results.

Fig. 9 shows the BER performance of different

channel estimation methods in ETU channel with

max. Doppler frequency of 300Hz. In this case, the

NovelChannelEstimationAlgorithmusingVariousFilterDesigninLTE-AdvancedSystem

123

mobile user moves with very high speed; therefore,

as shown in section II, the channel suffers significant

variations in time, even within one subframe. We

can observe that the effectiveness of LS and MMSE

is affected by the high Doppler spread; therefore it is

necessary to employ techniques that consider the

time correlation of the channel. We demonstrate

through this simulation that our proposed technique,

New CE, produces the best results for high Doppler

spread environments.

Table 5: Simulation parameters.

Parameter Value

Bandwidth 20 MHz

Sample frequency 30.72 MHz

Subframe duration 1 ms

Subcarrier spacing 15 kHz

FFT size 2048

Occupied subcarriers

1200 + DC subcarrier =

1201

No. subcarriers/RB 12

No. of RB’s/subframe 100

CP size (samples) 512 (extended CP)

No. of OFDM symbols/subframe 12 (extended CP)

No. of reference signals per RB 8

Modulation scheme QPSK

Noise AWGN

No. of antennas 1x1

Channel estimation Techniques

LS with linear interpolation,

MMSE, Wiener Filter,

Creative CE

Channel models

3GPP LTE extended channel

models: EPA, EVA, ETU

Figure 7: BER Performance using different channel

estimation methods in EPA channel with maximum

Doppler frequency of 5 Hz.

Figure 8: BER performance using different channel

estimation methods in EVA channel with maximum

Doppler frequency of 70Hz

Figure 9: BER performance using different channel

estimation methods in Rayleigh ETU channel; with

maximum Doppler frequency of 300 Hz.

6 CONCLUSIONS

In this paper we proposed a novel channel

estimation method to improve transmission and

reception of data in high speed environments. We

designed our system considered the especial

arrangement of reference signals in 3GPP LTE-

Advanced and demonstrated through MATLAB

simulations that the BER performance result in high

Doppler spread is very close to the case of ideal

channel estimation.

SIGMAP2012-InternationalConferenceonSignalProcessingandMultimediaApplications

124

ACKNOWLEDGEMENTS

This research was supported by the MKE(The

Ministry of Knowledge Economy), Korea, under the

ITRC (Information Technology Research Center)

support program supervised by the NIPA(National

IT Industry Promotion Agency)" (NIPA-2012-

H0301-12-3005).This study was financially

supported by Chonnam National University, 2011.

REFERENCES

3GPP TS 36.211: "E-UTRA; Physical Channels and

Modulations", v 9.1.0, March 2010.

L. Somasegaran, "Channel Estimation and Prediction in

UMTS LTE", Master Thesis, Aalborg University,

Denamrk, Feb. 22nd - June 25th 2007, 2

edition.

M. Simko, "Channel Estimation for UMTS Term Long

Evolution", Master Thesis, Vienna University of

Technology, Austria, June 2009.

J. Hou, J. Liu, "A Novel Channel Estimation Algorithm

for 3GPP LTE Downlink System Using Joint Time-

Frequency Two-Dimensional Iterative Wiener Filter",

University of Science and Technology of China.

Y. Qin, B. Hui, K. Chang, "Novel Pilot-Assisted Channel

Estimation Techniques for 3GPP LTE Downlink with

Performance-Complexity Evaluation", Korean

Conference on Digital Communications, vol. 37, no. 7,

Oct. 2007.

B. Karakaya, H. Arslan, H. Cirpan, "An Adaptive Channel

Interpolator Based on Kalman Filter for LTE Uplink

in High Doppler Spread Environments.", EURASIP

Journal on WIreless Communications and

Networking, vol. 2009, article ID 893751.

W. Ling, L. Ting, "Kalman filter channel estimation based

on comb-type pilot in time-varying channel",

International Conference on Wireless

Communications, Networking and Mobile Computing

(WiCOM), Sept. 2006.

NovelChannelEstimationAlgorithmusingVariousFilterDesigninLTE-AdvancedSystem

125