WO2012044098A2

WO2012044098A2 - Method for estimating snr (signal to noise ratio) in wireless communication system

Info

Publication number: WO2012044098A2
Application number: PCT/KR2011/007207
Authority: WO
Inventors: Jeong Wook An; In Tae Hwang; In Sik Cho; Chang Woo Seo
Original assignee: Lg Innotek Co., Ltd.
Priority date: 2010-09-30
Filing date: 2011-09-29
Publication date: 2012-04-05
Also published as: WO2012044098A3; KR20120033761A

Abstract

Disclosed is a method for estimating an SRN in a wireless communication system. The SNR is exactly estimated without a channel estimation process by using preambles that are mutually informed to both sides of the transceiver in the wireless communication system. The power of a transmitted signal is estimated by using the fact that the size of a transmitted preamble is 1 through a predetermined algorithm (ρ_av,new)ofestimatinganSNR,andnoisepoweriscalculatedbyusingtheaverageofabsolutevaluesofthedifferencebetweentworeceivedpreambles,sothattheSNRisestimatedbasedontherelativevariationofthenoisepowerinfixedtransmitpower.TheSNRisexactlyestimatedwithoutachannelestimationprocessbyusingpreamblesthataremutuallyinformedtobothsidesofthetransceiverinthewirelesscommunicationsystem.

Description

METHOD FOR ESTIMATING SNR (SIGNAL TO NOISE RATIO) IN WIRELESS COMMUNICATION SYSTEM

The embodiment relates to a method for estimating an SNR (Signal to Noise Ratio) in a wireless communication system. In particular, the embodiment relates to a method for estimating an SRN in a wireless communication system, capable of exactly estimating the SNR without a channel estimation process by using preambles that are mutually informed to both sides of the transceiver in the wireless communication system.

To enhance the reliability and a throughput in a wireless communication environment, a link adaptation scheme, a subcarrier allocation scheme, a power allocation scheme, and the like are used. The schemes require feedback information of a channel state. In other words, to sufficiently operate and improve system performance through the schemes, the receiver must have an SNR estimator designed to exactly estimate the information of the channel state with low complexity.

Preamble-based SNR estimation algorithms according to the related art include a Boumard SNR estimation algorithm, a Ren SNR estimation algorithm, and a Milan SNR estimation algorithm.

According to the Boumard SNR estimation algorithm, on the assumption that a channel is rarely changed between adjacent sub-carriers in a 2×2 MIMO (Multi Input Multi Output)-OFDM (Orthogonal Frequency Division Multiplexing) system, an SNR is estimated, and two identical and consecutive preamble symbols are used. In this case, transmitted preambles may be expressed as following equation.

Equation 1

In this case, n represents the number of carriers (n=1, …, and N), "0"represents the first transfer preambles, and "1" represents the second transfer preambles.

SNR estimation is performed by estimating signal power and noise power as shown in following equation 2. In Equation 2,

represents estimated average power for the overall transmitted frames, and

represents average noise power. To find

,

is calculated as shown in Equation 3 by using two consecutive received preambles Y(0,n) and Y(1,n), and original preambles C(n), which have been transmitted, and the average of absolutely values of

is found as shown in FIG. 4.

Equation 2

Equation 3

Equation 4

According to the Boumard SNR estimation algorithm, the SNR can be estimated without channel estimation required in a conventional ML (Maximum Likelihood) estimator or a conventional MMSE (Minimum Mean Squared Error) SNR estimator. However, since noise power is estimated as shown in Equation 5 on the assumption that a channel is rarely changed between the adjacent sub-carriers, the Boumard SNR estimation algorithm is not suitable when the channel is greatly changed.

Equation 5

Next, according to the Ren SNR estimation algorithm, SNR is estimated as shown in following equation 6 based on two preambles having the same OFDM symbol size similarly to the Boumard SNR estimation algorithm, and noise power is estimated in the same sub-carrier as shown in FIG. 7 differently from the Boumard SNR estimation algorithm. Therefore, the Ren SNR estimation algorithm may be an algorithm improved from the Boumard SNR estimation algorithm sensitive to a frequency-selective characteristic of a channel. The signal power is estimated by removing the estimated noise power from the whole received signal power as shown in FIG. 8.

The

is calculated through the channel estimation similarly to the

of Equation 3.

Equation 6

Equation 7

Equation 8

Next, according to the Milan SNR estimation algorithm, several identical preambles are used in a time domain. If a preamble having N sub-carriers in total is converted into Q preamble structures, which are periodic and identical to each other in the time domain as shown in FIG. 1A, a signal periodically appear every subcarrier having Q intervals, and a Null (zero) signal appears between the intervals as shown in FIG. 1B. The Milan SNR estimation algorithm is to estimate an SNR using the characteristic. In other words, the S estimation after an FFT (Fast Fourier Transform) is performed every Q intervals in which a signal appears as shown in following equation 9, and the estimation of the W is performed in a Null carrier as shown in following Equation 10.

Equation 9

Equation 10

The Yp (m) represents a received signal appearing every Q intervals in the frequency domain and is expressed as following equation 11. In Equation 12, the Yz (mQ+q) represents a signal received in the position of a Null sub-carrier every m=0,1,..., and 63 and q =0.1. The η (k,n) represents AWGN (Additive White Gaussian Noise) having a size of 1 and an average of 0 at the Kth sample of an nth preamble.

Equation 11

Equation 12

Accordingly, an SNR estimation value

is calculated in Equation 13 based on Equations 11 and 12 through the Milan SNR estimation algorithm.

Equation 13

However, the above-described SNR estimation algorithms according to the related art represent complex operation procedures for SNR estimation and great feedback delay.

The present disclosure has been made keeping in mind the above problems occurring in the related art, and an object of the present disclosure is to provide a method for estimating an SRN in a wireless communication system, capable of exactly estimating the SNR without a channel estimation process by using preambles that are mutually informed to both sides of the transceiver in the wireless communication system.

In order to accomplish the object, an SRN estimation scheme in a wireless communication system according to the embodiment has the following features.

The power of a received signal is estimated by using the fact that the size of a transmitted preamble is 1 through a predetermined algorithm (ρav, new) of estimating an SNR, and noise power is calculated by using the average of absolute values of the difference between two received preambles, so that the SNR is estimated based on the relative variation of the noise power in fixed transmit power.

In this case, the algorithm (ρav, new) of estimating an SNR can be expressed as a following equation.

Equation

In the Equation, the Y(0,n) and Y(1,n) represent received signals obtained after FFT (Fast Fourier Transform) has been performed with respect to consecutively transmitted preambles which are formed through a BPSK scheme or a QPSK scheme, have a size of 1, and have a same pattern.

As described above, according to the embodiment, an SNR can be exactly estimated without a channel estimation process by using preambles that are mutually informed to both sides of the transceiver in a wireless communication system.

FIG. 1 is a view showing a preamble structure used in a Milan SNR estimation algorithm according to the related art;

FIG. 2 is a view showing a frame structure of a transmitted preamble applied for an SNR estimation scheme in a wireless communication system according to the embodiment;

FIG. 3 is a graph showing the comparison between real SNR values according to the embodiment and SNR values estimated through SNR estimation algorithms according to the related art under an environment of a Rayleigh flat fading channel;

FIG. 4 is a graph showing the performance comparison between an NMSE (Normalized Mean Square Error) according to the embodiment and an NMSE through the SNR estimation algorithms according to the related art under the environment of the Rayleigh flat fading channel;

FIG. 5 is a graph showing the comparison between real SNR values according to the embodiment and SNR values estimated through the SNR estimation algorithms according to the related art under an environment of a Rayleigh selective fading channel A;

FIG. 6 is a graph showing the performance comparison between an NMSE (Normalized Mean Square Error) according to the embodiment and an NMSE through the SNR estimation algorithms according to the related art under the environment of the Rayleigh selective fading channel A;

FIG. 7 is a graph showing the comparison between real SNR values according to the embodiment and SNR values estimated through the SNR estimation algorithms according to the related art under an environment of a Rayleigh selective fading channel B;

FIG. 8 is a graph showing the performance comparison between an NMSE (Normalized Mean Square Error) according to the embodiment and an NMSE through the SNR estimation algorithms according to the related art under the environment of the Rayleigh selective fading channel B;

FIG. 9 is a graph showing packet error rates of MCS (Modulation and Coding Scheme) levels under the environment of the Rayleigh flat fading channel;

FIG. 10 is a graph showing throughputs for MCS levels under the environment of the Rayleigh flat fading channel;

FIG. 11 is a graph showing throughputs according to the SNR estimation algorithm according to the embodiment and the SNR estimation algorithms according to the related art when an AMC (Adaptive Modulation and Coding) scheme is applied under the environment of the Rayleigh flat fading channel;

FIG. 12 is a graph showing packet error rates for the MCS levels under the environment of the Rayleigh selective fading channel A;

FIG. 12 is a graph showing throughputs for the MCS levels under the environment of the Rayleigh selective fading channel A;

FIG. 14 is a graph showing throughputs according to the SNR estimation algorithm according to the embodiment and the SNR estimation algorithms according to the related art when the AMC scheme is applied under the environment of the Rayleigh selective fading channel A;

FIG. 15 is a graph showing packet error rates for MCS levels under the environment of the Rayleigh selective fading channel B;

FIG. 16 is a graph showing throughputs for MCS levels under the environment of the Rayleigh selective fading channel B;

FIG. 17 is a graph showing throughputs according to the SNR estimation algorithm according to the embodiment and the SNR estimation algorithms according to the related art when the AMC scheme is applied under the environment of the Rayleigh selective fading channel B;

FIG. 18 is a block diagram showing a transmitter in a multiple antenna OFDM system according to the embodiment; and

FIG. 19 is a block diagram showing a receiver in the multiple antenna OFDM system according to the embodiment.

Hereinafter, the embodiment will be described in detail with reference to accompanying drawings.

FIG. 2 is a view showing a transmission frame structure applied for an SNR estimation scheme in a wireless communication system according to the embodiment.

Referring to FIG. 2, the transmission frame includes preamble signals and data signals.

The preamble signals may be used to obtain frequency synchronization and time synchronization in a receiver. For example, the preamble signal includes a short training field and a long training field in a wireless LAN system (see IEEE 802.11).

In addition, the preamble signal may be used in the receiver to estimate a channel and channel quality (SNR). Hereinafter, the SNR estimation scheme according to the embodiment can be performed by using the preamble signal.

FIG. 18 is a block diagram showing a transmitter 200 in a multiple antenna OFDM system according to the embodiment.

Referring to FIG. 18, the transmitter 200 includes a channel encoder 210, an interleaver 220, a serial/parallel converter 230, a mapper 240, an IFFT module 250, an AMC controller 260, a receiving circuit 270, and a multiple antenna 280. Hereinafter, the details of other components of the transmitter 200 that does not relate to the embodiment will be omitted.

The channel encoder 210 encodes input information stream through a predetermined encoding scheme to form coded words. The channel encoder 210 can insert error detection bits such as CRC (cyclic redundancy check) codes and redundancy codes for error correction into the information stream.

The channel encoder 210 may employ convolution codes, turbo codes, low-density parity-check codes, or rate compatible punctured convolution codes as error correction codes.

The interleaver 220 blends the coded data with each other to reduce noise from a channel.

The serial/parallel converter 230 converts serial signals output from the interleaver 220 to parallel signals. The mapper 240 modulates the coded words subject to the interleaving through a predetermined modulation scheme to provide modulated symbols. In other words, the coded data are mapped with the modulated symbols representing the positions according to amplitude constellation and phase constellation by the mapper 240.

The mapper 240 may use various modulation schemes such as an m-PSK (m-Phase Shift Keying) scheme or an m-QAM (m-Quadrature Amplitude Modulation) scheme. Meanwhile, according to the present embodiment, the mapper 240 may transmit signals through a BPSK scheme or a QPSK scheme.

The IFFT module 250 performs inverse FFT with respect to the modulated symbols output from the mapper 240 to transform the modulated symbols into time-domain samples. A CP (cyclic prefix) inserting module (not shown) inserts a CP, which serves as a guard interval, into the time-domain sample. The CP eliminates intersymbol interference to convert a frequency-selective channel into a flat fading channel. Signals output from the CP inserting module are converted into analogue signals and transmitted through the multiple antenna 280.

The receiving circuit 270 receives signals from a UE through the multiple antenna 280. The receiving circuit 270 digitalizes the received signals and outputs the signals to the AMC controller 260.

The AMC controller 260 determines an MCS (Modulation and Coding Scheme) level based on channel quality information provided from the UE. The channel quality information may include a signal-to-noise ratio (SNR) or an index of the MCS level.

Based on the determined MCS level, the AMC controller 260 provides an encoding scheme to the channel encoder 210 and provides a modulation scheme to the mapper 240.

The memory 290 may store a look-up table for indexes of the MCS levels.

The transmitter 200 generates a frame of FIG. 2 through the above components, and transmits the frame to the UE.

FIG. 19 is a block diagram showing a receiver 300 in the multiple antenna OFDM system according to the embodiment.

Referring to FIG. 19, the receiver 300 includes a multiple antenna 305, an FFT module 310, an equalizer 330, a de-mapper 340, a parallel/serial converter 350, a de-interleaver 360, a channel decoder 370, a controller 380, a channel quality estimator 390, and a transmitting circuit 395. Hereinafter, the details of other components of the receiver 300 that do not relate to the embodiment will be omitted.

The signals received in the receiver 300 through the multiple antenna 350 are digitalized, and the CP are eliminated from the signals by a CP eliminator (not shown). The samples without the CP are subject to Fast Fourier Transform in the FFT module 310 and transformed into frequency-domain symbols.

The equalizer 330 equalizes the frequency-domain symbols output from the FFT module 341.

The de-mapper 340 is controlled by a demodulation signal of the controller 380 to de-map the frequency-domain symbols with coded data. A demodulation scheme provided by the controller 380 corresponds to the modulation scheme provided to the mapper 240 by the AMC controller 260 of the transmitter 200.

The parallel/serial converter 350 converts parallel signals output from the de-mapper 340 into serial signals and then output the serial signals to the de-interleaver 360. The channel decoder 370 decodes the de-interleaved data under the control of the controller 380. The channel decoder 370 outputs estimated data bits. A decoding scheme provided by the controller 380 corresponds to the encoding scheme provided to the channel encoder 210 by the AMC controller 380 of the transmitter 200.

The controller 380 controls the overall operation of the receiver 300 and selects an MCS level to maximize the throughput of the OFDM system through channel quality estimated by the channel quality estimator 390.

The memory 385 may store a look-up table for MCS levels. The look-up table may be the same as the look-up table stored in the memory 290 of the transmitter 200. The controller 380 determines indexes of MCS levels based on the look-up table according to the determined MCS levels.

The transmitting circuit 395 receives channel quality information from the controller 380 and transmits the channel quality information to the counterparty through the multiple antenna 305. The channel quality information may include an SNR or the indexes of the MCS levels.

The channel quality estimator 390 estimates channel quality without a channel estimation process by using preambles that are informed to both sides of the transmitter 200/the receiver 300. In this case, the channel quality refers to an SNR.

The channel quality estimator 390 extracts preamble signals from outputs of the FFT module 310 to estimate the SNR. According to the SNR estimation scheme in the wireless communication system of the embodiment, the power of a received signal is estimated by using the fact that the size of the preamble of the transmitted frame shown in FIG. 2 is 1 through a predetermined algorithm (ρav, new) of estimating an SNR, and noise power is calculated by using the average of absolute values of the difference between two received preambles, so that the SNR is estimated based on the relative variation of the noise power in fixed transmit power.

In this case, the algorithm (ρav, new) of estimating the SNR can be expressed through following equations.

Equation 14

In Equation 14, the estimated power Snew of the received signal is 1. In other words, since the BPSK scheme or the QPSK scheme is used as the modulation scheme for the preamble in the transmitted frame, the size of the preamble is 1.

In Equation 14, the noise power Wnew is expressed as the average of absolute values of the difference between two received preambles, that is,

.

On the assumption that the first received preamble is Y1 and the second received preamble is Y2, the two received preambles are defined as following equations.

I) Y1 = X1 * H1 + N1

II) Y2 = X2 * H2 + N2

In addition, the difference (Y1- Y2) between the two received preambles is defined as the following equation.

III) Y1 - Y2 = (X1*H1 + N1) - (X2*H2 + N2)

In this case, the X1, and X2 represent consecutively transmitted preamble signals, and the H1 and H2 represent channel components. The N1 and N2 represent noise components.

On the assumption that the channel components H1 and H2 represent ideal channels, since the X1 and X2 represent transmitted preamble signals, the difference between the two preambles (Y1- Y2) has only noise components. Accordingly, the noise power Wnew is calculated as the average of absolute values of the difference between two received preambles.

In Equation 14, Y(0,n) and Y(1,n) represent received signals obtained after the FFT has been performed with respect to the consecutively transmitted preambles which have the same pattern and the size of 1, and have been modulated through a BPSK scheme.

In addition, the Y(0,n) and Y(1,n) may represent received signals after the FFT has been performed with respect to the consecutively transmitted preambles which have the same pattern, the size of 1, and have been modulated through a QPSK scheme.

Hereinafter, the comparison between the SNR estimation algorithm (ρav, new) applied for the SNR estimation scheme according to the embodiment and the SNR estimation algorithm according to the related art, and the analysis thereof will be described based on information obtained through performance simulations performed with respect to the SNR estimation algorithm (ρav, new) according to the embodiment and the SNR estimation algorithm according to the related art. Meanwhile, the SNR estimation algorithms according to the embodiment and the comparative example have been simulated with respect to an MIMO-OFDM-based wireless LAN system (see IEEE 802.11n).

Following tables 1 and 2 represent simulation parameters and channel parameters, respectively. Table 1 represents Simulation Parameter and Table 2 represents Channel Parameters.

Table 1

Parameters	Values
System Bandwidth (BW)	20MHz
1OFDM Symbol Time	4㎲(3.2㎲:FFS duration + 0.8㎲: CP length)
The number of data symbols per spatial stream (SSThe number of data sub-carriers)	468
The number of data sub-carriers	64
Sub-carrier interval	312.5KHz
MIMO	Layered 2*2
Noise	AWGN
FFT size	64 point
GI(CP) size	16 point
1 OFDM symbol sample	80
SNR estimation scheme	Boumard, Milan, Ren, New
Preamble	Two consecutive identical OFDM symbols of QPSK or BPSK
Modulation Scheme	BPSK, QPSK, 16QAM
Convolutional Coding	1/2, 3/4
Transmission packet	25000

Table 2

Channel	Delay Path (samples)	Reyleigh Power
Reyleigh Selective Channel A	3 Path:[0 12 15]	[-1.92 -5.92 -9.92]
Reyleigh Selective Channel B	4 Path:[0 12 15 18]	[-1.92 -5.92 -9.92 -12.92]
Reyleigh Flat Channel	No Delay	-

The system parameters have been formed based on the IEEE 802.11n MIMO-OFDM system, and performance analysis has been performed with respect to channels for two transmit/receive antennas.

Each SNR estimation algorithm employs two consecutive preambles that have been modulated through the BPSK or the QPSK and have an OFDM symbol size (64 point). The simulation has been performed by constructing an environment of a Rayleigh flat fading channel, in which the channel state is rarely changed, a Rayleigh selective fading A, in which the maximum delay sample is smaller than the length of the CP, and a Rayleigh selective channel B in which the maximum delay sample is greater than the length of the CP.

For the purpose of performance evaluation, SNR values estimated through the algorithms are roughly compared with a real SNR value to determine the approximation degree of the estimated SNR values to the real SNR. In order to accurately perform performance comparison, an NMSE (Normalized Mean Square Error) is found as shown in Equation 15, and performance analysis is performed based on the NMSE.

Equation 15

In Equation 15, the Nt represents the number of transmitted packets, in which 25, 000 transmitted packets are used, and the

represents an SNR value that is estimated from the received preamble of an ith packet. The

represents the real SNR value.

Meanwhile, regarding the comparison between SNR values estimated through the above algorithms with the real SNR value under the environment of the Rayleigh flat channel through the graph of FIG. 3, the Ren algorithm shows a SNR estimation error higher than that of another algorithm within a low SNR range. In contrast, in the case of another algorithm, an estimated SNR value is substantially equal to the real SNR value.

In addition, hereinafter, the more accurate performance difference between the algorithms will be described based on the NMSE performance results. In the case of a novel algorithm (marked as "NEW" in FIG. 4) according to the embodiment and the Boumard algorithm, the NMSE approximates 0. Performance of the OFDM system is improved in order of the Milan algorithm and the Ren algorithm. As described above, according to the performance results, the Boumard algorithm is suitable for an environment in which channels are rarely changed. The novel SNR estimation algorithm according to the embodiment shows the performance of the Boumard algorithm.

FIGS. 5 and 6 show performance graphs and NMSE performance results obtained by comparing the real SNR and SNRs estimated through the above algorithms under the environment of the Rayleigh selective channel A.

Referring to FIG. 5, in the case of the performance of the Boumard algorithm, the estimated SNR value approximates the real SNR value up to -4dB. After -4dB, estimation errors are increased due to the frequency-selective characteristics of the channel. In contrast, as shown in FIG. 6, in the case of the Milan algorithm or Ren algorithm, the estimated errors are gradually reduced. In detail, as the SNR value is increased, the NMSE is gradually reduced and then constantly maintained at about 0.3. In the case of the algorithm according to the embodiment, the NMSE approximates about 10-3, and the SNR can be estimated close to the real SNR without the influence of multiple paths exerted on the frequency-selective characteristic of the channel.

FIGS. 7 and 8 are graphs showing performance results under the environment of the Rayleigh selective channel B having 4 multiple paths, in which the maximum delay of the channel is greater than the CP. The performance analysis is performed under an environment inferior to that of the channel A.

Therefore, when comparing with the channel A, the performance for the SNR estimation algorithms represents higher estimation errors. As shown in FIG. 7, the Boumard algorithm does not estimate the real SNR value from -2dB. In the case of the algorithm according to the embodiment and the Milan and Ren algorithms, the estimated SNR value approximates the real SNR value up to about 20dB. Among the above algorithms, the algorithm according to the embodiment has a lower estimation error in terms of the NMSE. As shown in FIG. 8, even the algorithm according to the embodiment shows performance degradation from 28dB due to the interference of the frequency-selective characteristic. However, when comparing with the other algorithms, the algorithm according to the embodiment shows superior performance with respect to all channels.

In order to evaluate the performance for throughputs according to estimated SNR performance when the receiver 300 estimates SNR values and feeds the information about the estimated SNR values back, an AMC scheme is applied for the SNR estimation algorithms, and the performance for the throughputs is analyzed.

Following table 3 represents MCS levels selected by suitably taking into an MCS table of IEEE 802.11n consideration so that the AMC is applied to the SNR estimation algorithms. Table 3 represents MCS levels selected to apply AMC.

Table 3

MCS	modulation		CodeRate	MCP	Throuput[Mbps](Data Subscriber=64,CP=16)	Nss	Nsts
MCS	stream1	stream2	CodeRate	MCP	Throuput[Mbps](Data Subscriber=64,CP=16)	Nss	Nsts
1	BPSK	BPSK	1/2	1	16.0	2	2
2	QPSK	BPSK	1/2	1.5	24.0	2	2
3	QPSK	QPSK	1/2	2	32.0	2	2
4	QPSK	QPSK		3/4	3	48.0	2	2
5	16QAM	16QAM	1/2	4	64.0	2	2

* Nss : The number of spatial streams, * Nsts : The number of temporal and spatial streams

* MCP : Modulation and Coding Product

FIG. 9 is a graph showing packet error rates for MCS levels shown in table 3 under the environment of the Reyleigh flat fading channel, and FIG. 10 is a graph showing the performance for the throughputs for the MCS levels under the environment of the Reyleigh flat fading channel.

Referring to FIGS. 9 and 10, as the MCP, that is, the MCS level is increased, the performance for the packet error rate is degraded and the maximum throughput is increased. In FIG. 10, in order to properly select an MCS level allowing the optimum operation within the reference range set about an SNR corresponding to the intersection between throughputs of the MCS levels, SNR estimation must be exactly achieved. Therefore, as the exact SNR estimation is achieved, the best throughput can be obtained when the AMC is applied. As shown in FIG. 11, the algorithm according to the embodiment and the Boumard algorithm to exactly achieve SNR estimation represent the throughputs superior to that of the Ren algorithm or the Milan algorithm on the whole.

Therefore, under the environment of the flat fading channel environment, the suggested algorithm according to the embodiment and the Boumard algorithm, which represent NMSE performance superior to that of the Ren algorithm or the Milan algorithm, represent superior throughput.

FIGS 12 and 13 are graphs showing packet error rates and throughput performance for MCS levels shown in table 3 under the environment of the Reyleigh selective fading channel A.

Although performance aspects for the MCS levels are similar to those of the environment of the flat fading channel, the overall performance for packet error rates is degraded due to the multi-path fading. In particular, the environment of the Reyleigh selective fading channel A makes a greater difference in the performance for MCS4 and MCS5 from the environment of the flat fading channel. The maximum throughput can be represented at an SNR higher than that under the environment of the flat fading channel.

FIG. 14 is a graph showing the performance for the throughput in each SNR estimation algorithm when the AMC is applied for the environment of the Rayleigh selective fading channel A.

As shown in FIG. 14, under the environment of the Rayleigh selective fading channel A, the Boumard algorithm does not sufficiently achieve SNR estimation from about 10dB and satisfies only the maximum throughput corresponding to the MCS 1. In contrast, the algorithm according to the embodiment, the Ren algorithm, and the Milan algorithm sufficiently achieve the SNR estimation on the whole and satisfy the maximum throughput corresponding to the MCS 5. In addition, as shown in NMSE performance results of FIG. 6, under the environment of the channel A, since the algorithm according to the embodiment represents SNR estimation error rate lower than that of the Ren algorithm or the Milan algorithm, the algorithm according to the embodiment can represent the best throughput on average.

FIGS. 15 and 16 are graphs the package error rates for MCS levels and the performance for the throughputs under the environment of the Rayleight selective channel B.

Referring to FIGS. 15 and 16, when comparing with the environment of the Rayleight selective channel A, since the maximum delay of the channel is greater than the guide interval (CP), ISI (intersymbol interference) occurs. Therefore, in the case of the packet error rate, even if the SNR is increased, performance degradation may occur, and an original maximum throughput is not satisfied.

FIG. 17 is a graph showing the performance results for throughputs when SNR values estimated through the above algorithms are provided as feedback information, and the AMC applied under the environment of the Rayleigh selective fading channel B.

Referring to FIG. 17, the Boumard algorithm under the environment of the Rayleigh selective channel B represents inferior AMC throughputs due to the high SNR estimation errors similarly to the channel A. The remaining three algorithms satisfy the maximum throughput while achieving exact SNR estimation. Among them, the algorithm according to the embodiment represents superior throughputs on average.

As described above, according to the SNR estimation scheme in the wireless communication system of the embodiment, the SNR can be exactly estimated without a channel estimation process by using preambles that are mutually informed to both sides of the transceiver in an IEEE 802.11n system.

Any reference in this specification to "one embodiment", "an embodiment", "example embodiment", etc., means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with any embodiment, it is submitted that it is within the purview of one skilled in the art to effect such feature, structure, or characteristic in connection with other ones of the embodiments.

Although embodiments have been described with reference to a number of illustrative embodiments thereof, it should be understood that numerous other modifications and embodiments can be devised by those skilled in the art that will fall within the spirit and scope of the principles of this disclosure. More particularly, various variations and modifications are possible in the component parts and/or arrangements of the subject combination arrangement within the scope of the disclosure, the drawings and the appended claims. In addition to variations and modifications in the component parts and/or arrangements, alternative uses will also be apparent to those skilled in the art.

Claims

A method for estimating channel quality in a wireless communication system, the method comprising:

estimating power of a received signal by using a size of a transmitted preamble;

calculating noise power by using a difference between two preamble signals received from a transmitter; and

estimating the channel quality based on the power of the received signal and the noise power.
The method of claim 1, wherein the channel quality is an SNR (Signal to Noise Ratio).
The method of claim 1, wherein the two preamble signals are transmitted through a BPSK (Binary Phase Shift Keying) scheme or a QPSK (Quadrature Phase Shift Keying) scheme by the transmitter.
The method of claim 3, wherein the size of the transmitted preamble is 1.
The method of claim 3, wherein the power of the received signal is estimated as 1.
The method of claim 1, wherein the transmitted preamble is pre-informed to both of the transmitter and a receiver.
The method of claim 1, wherein the noise power is an average of an absolute value of a difference between the two preamble signals.
The method of claim 7, wherein the two preamble signals are received signals of consecutively transmitted preambles.
The method of claim 1, wherein the channel quality estimation (ρ_av,new)iscalculatedaccordingtoafollowingequation,

,

where the Y(0,n) and Y(1,n) represent received signals obtained after FFT (Fast Fourier Transform) has been performed with respect to consecutively transmitted preambles which are formed through a BPSK scheme or a QPSK scheme with a size of 1 and a same pattern.
An apparatus for estimating channel quality in a wireless communication system, the apparatus comprising:

an antenna configured to receive a preamble signal from a transmitter;

a channel quality estimator configured to estimate power of a received signal by using a size of a transmitted preamble, to calculate noise power by using a difference between two preamble signals received from the transmitter, and to estimate the channel quality based on the power of the received signal and the noise power; and

a controller configured to feedback information about the channel quality received from the channel quality estimator to the transmitter.
The apparatus of claim 10, wherein the channel quality is an SNR.
The apparatus of claim 10, wherein the two preamble signals are transmitted through a BPSK scheme or a QPSK scheme by the transmitter.
The apparatus of claim 12, wherein the size of the transmitted preamble is 1.
The apparatus of claim 12, wherein the power of the received signal is estimated as 1.
The apparatus of claim 10, wherein the transmitted preamble is pre-informed to both the transmitter and a receiver.
The apparatus of claim 10, wherein the noise power is an average of an absolute value of a difference between the two received preamble signals.
The apparatus of claim 16, wherein the two received preamble signals are received signals of consecutively transmitted preambles.
The apparatus of claim 10, wherein the channel quality estimation (ρ_av,new)iscalculatedaccordingtoafollowingequation,

,

where the Y(0,n) and Y(1,n) represent received signals obtained after FFT (Fast Fourier Transform) has been performed with respect to consecutively transmitted preambles which are formed through a BPSK scheme or a QPSK scheme with a size of 1 and a same pattern.