KR20040070499A - A Receiver and a Decoding Method for a Space-time Coded CDMA System in Fading Channels with Arrival Time Difference - Google Patents

Abstract

PURPOSE: A receiver for a space-time coded CDMA system in a fading channel having an arrival time difference and a decoding method are provided to simultaneously decode a received signal and update a weight vector. CONSTITUTION: An error(em(i)) between a received signal(ym(i)) of an ith symbol interval and a received signal corresponding to a value determined as a symbol in the ith symbol interval is obtained(S10). A tap weight vector for the (i+1)th symbol is updated from the error(em(i)) by an LMS(Least Mean Square) algorithm(S20). A received signal is despreaded by using the updated tap weight vector(S30).

Description

A receiver and a decoding method for a space-time coded CDMA system in Fading channels with Arrival Time Difference}

The present invention relates to a receiver for a CDMA communication system space-time encoded in a fading channel having a time difference of arrival between a plurality of transmit antennas and a plurality of receive antennas, the maximum likelihood decoder having an error correction decoder, and The present invention relates to a receiver for a CDMA system and a method for decoding a received signal, comprising an adaptive interference canceller (AIC) corresponding to a channel of.

Space-time coding technology is proposed to simultaneously obtain diversity gain and coding gain in a wireless communication system having a plurality of transmit and receive antennas. By using a space-time code that satisfies the rank condition and the determinant condition, maximum diversity gain and encoding gain can be obtained, and high data rate can be obtained by dividing the encoded data into a plurality of transmitting antennas.

Most studies on space-time coding assume that the channels between transmit and receive antennas are independent of each other, and that signals transmitted from transmit antennas arrive at the receive antenna simultaneously. However, in order to maintain independence between fadings in order to obtain sufficient diversity gain, the distance between the antennas must be increased, but the time for each signal from the transmitting antennas to the receiving antenna is different from each other. If such a time difference of arrival exists, the performance of the system is degraded because signals transmitted simultaneously from multiple transmit antennas interfere with each other.

On the other hand, the space-time encoding technique is also applied to a code division multiple access (CDMA) system in order to meet the demand for high-speed wireless transmission. In the space-time coded DS-CDMA system, the interference between the transmitting antennas due to the time difference of arrival has the same effect as the presence of multipath. Thus, even in the forward link, multiple access interference occurs due to loss of orthogonality between signature sequences. Since such interference increases in proportion to the product of the number of users and the number of transmit antennas, the performance degradation is more severe than that of a single transmit antenna system.

Therefore, when there is a time difference of arrival from the transmitting antenna, an appropriate interference cancellation technique should also be used in the forward link to improve performance degradation.

Among various interference cancellation techniques, Adaptive Interference Canceller (AIC) is known to provide better performance than matched filter without increasing the complexity in the forward link. The adaptive algorithm used in the conventional adaptive interference canceller is to estimate the transmitted symbol per symbol period in order to update the filter coefficient, and calculate an error signal therefrom.

However, the AIC is theoretically proposed and studied to be implemented in an uncoded system. Therefore, in a space-time coded DS-CDMA system, since symbols encoded over a plurality of transmit antennas are added and received in units of frames, an additional complexity is required to implement a conventional adaptive algorithm for estimating a transmitted symbol every symbol period. do.

In order to solve the above problems, the present invention provides a maximum likelihood of a CDMA communication system receiver having an error correction decoder in space-time coded in a fading channel having a time difference of arrival between a plurality of transmit antennas and a plurality of receive antennas. A receiver and received signal decoding method for a CDMA system including a decoder (Maximum Likelihood Decoder), and an adaptive interference canceller (AIC) corresponding to each channel.

It is therefore an object of the present invention to provide a receiver for a CDMA communication system that is space-time encoded in a fading channel in which a time difference of arrival exists between a plurality of transmit antennas and a plurality of receive antennas.

It is also an object of the present invention to provide a method of decoding a received signal in a CDMA communication system which is space-time encoded in a fading channel having a time difference of arrival between a plurality of transmit antennas and a plurality of receive antennas.

1 illustrates a space-time coded DS-CDMA system model,

2 shows an example of a receiver according to the invention,

3 shows a receiver for an mth receiving antenna among the receivers according to the present invention,

4 illustrates a decoding algorithm of the receiver of FIG. 3 in detail.

5 is a flowchart illustrating a received signal decoding process according to the present invention;

6 illustrates a maximum likelihood decoder in a BPSK space-time coded DS-CDMA system having L = 2, M = 1, and state number 8, and a bit error rate of a receiver according to the present invention;

7 shows the bit error rate when M = 2,

FIG. 8 shows the frame error rate of a QPSK space-time coded DS-CDMA system with L = 2, M = 2, E _b / N _o = 20 dB, number of states 8 and 16. FIG.

<Explanation of symbols for the main parts of the drawings>

110: transmit antenna 120: receive antenna

130: channel 210: chip matching filter

320: adaptive interference canceller (AIC) bank 322: adaptive interference canceller (AIC)

330: likelihood decoder (ML decoder) 331: Viterbi comparator

The present invention relates to a receiver for a CDMA communication system space-time encoded in a fading channel having a time difference of arrival between a plurality of transmit antennas and a plurality of receive antennas, the maximum likelihood decoder having an error correction decoder, and A receiver for a CDMA system comprising an adaptive interference canceller (AIC) corresponding to a channel of.

The present invention can be applied not only to a direct sequence spread spectrum DS-CDMA system but also to a general CDMA system using a sequence for distinguishing users.

In the present invention, the adaptive interference canceller updates the tap weight vector every symbol period, and despreads the received signal received from each channel into the updated tap weight vector.

The tap weight vector is a reception symbol that is determined from the received signal y _m of the previous symbol interval and the previous symbol interval for every symbol interval. Received signal corresponding to It is preferable to update by the LMS (least mean square) algorithm from the error value of the liver in terms of complexity reduction. The initial value of the tap weight vector is the signature sequence of a particular user.

Receive symbol determined in the received signal y _m of the previous symbol period and the previous symbol period Received signal corresponding to The error value is preferably an optimum value of an Euclidean distance value used for decoding the received signal y _m of the previous symbol period in the error correction decoder.

More preferably, an optimal value of the Euclidean distance value is a reception symbol determined in the previous symbol period in the error correction decoder. It is preferable that the Euclidean distance in the state transition minimizes the path estimates for.

Therefore, the receiver for the CDMA system according to the present invention decodes the received signal by the algorithm of the error correction decoder, and simultaneously uses the Euclidean distance value used in the algorithm as the error value for the AIC, thereby decoding and weighting the received signal. The vector can be updated at the same time.

The error correction decoder may be, for example, a Viterbi decoder or a turbo decoder.

In addition, the present invention provides a method for decoding a received signal in a CDMA communication system in which time-space encoding is performed on a fading channel having a time difference of arrival between a plurality of transmit antennas and a plurality of receive antennas.

Received as the received symbol of the i-th symbol section in the m-th receive antenna Received signal corresponding to Received signal between symbol and i-th symbol section Calculating an error value of the liver;

(B) updating the tap weight vector by the LMS algorithm from the error obtained in step (a); And

(C) despreading the signal received in the i + 1 th symbol interval using the tap weight vector obtained in step (b) instead of the signature sequence for each user. And a decoding method.

In the interference cancellation and decoding method of a received signal according to the present invention, in the step (a) And above The error value is preferably an optimum value of the Euclidean distance value used in the Viterbi decoding algorithm for decoding the received signal y _m of the i-th symbol section.

More preferably, the optimum value of the Euclidean distance value is the Euclidean distance in the state transition to minimize the path estimate for the i-th symbol by the Viterbi decoding algorithm.

Here, the tap weight vector means that the tap weight vector is a vector for despreading the received signal synchronized to the l-th transmit antenna.

Hereinafter, an embodiment of a receiver and a decoding method according to the present invention will be described in detail with reference to the drawings. However, the present invention is not limited by the following examples.

First, a space-time coded CDMA communication system will be described.

1 shows a space-time coded DS-CDMA system model with K users, with L transmit antennas 110 and M receive antennas 120.

the k th user in the i-th data bit b _k (i) is the length L of the codeword _{x k (i) = (x} 1 k (i), ..., x L k (i)) and the space-time encoding to, A signature sequence a _k = [a _{k, 0} , a _{k, 1} ,..., A _{k, N-1} ] of length N is mapped to each transmit antenna 110. The baseband signal transmitted from the l &lt; th &gt; transmit antenna is given by Equation 1:

In the above formula,

F represents the frame length,

a _k (t) represents the signature waveform,

Φ (t) represents a square chip pulse having unit energy,

T _S represents a symbol period,

T _C represents the chip period.

The channel 130 between each transmit / receive antenna is an iid Rayleigh fading channel, and it is assumed that a relative arrival delay time from the l th transmit antenna to the m th receive antenna is τ ^l _m . The baseband signal received at the m th receive antenna through this channel is represented by Equation 2:

In the above formula,

E _S represents the symbol energy (E _b / L) divided by the bit energy (E _b ) divided by the number of transmit antennas (L),

α ^l _m (t) and θ ^l _m (t) represent the magnitude and phase of Rayleigh fading, respectively.

n _m (t) represents Gaussian noise with a power density of N ₀ .

Signals extracted at chip periods through a chip matched filter (210 in FIG. 2) are received symbol vectors r _m (i) = [r _{m, 0} (i), r _{m, 1} (for a symbol period). i), ..., r _{m, N-1} (i)]. The fading of the channel is slow so as not to change for one frame period, and the arrival delay time τ ^l _m is assumed to be d ^l _m T _C for the integer d ^l _m ∈ [0, N-1].

Is a modified sequence of signature sequence a _k and defines the n th component as in Equation 3:

In the above formula, I _A is an indicator function for event A.

From the above definition, the signature sequence a _k is It can be seen that the same as.

The received symbol vector is expressed as in Equation 4:

Wherein, n _m (i) is the noise vector is distributed real part and the imaginary part consists of N _0/2 of the N independent complex Gaussian random variables.

The following describes the decoding process when there is no arrival time difference.

Assuming that the arrival delay times d ^l _m are all zeros, the signal despread with the signature sequence for the first user is given by:

In the above formula,

χ _m (x ₁ (i)) is a received signal corresponding to the transmitted signal x ₁ (i) of the first user, as shown in Equation 6a:

;

η _m (i) is equal to Equation 6b:

;

γ _{k, 1} is the correlation-correlation between the signature sequence of the k-th user and the first user.

Since the multiple access interference when the cross-correlation is 0 or very small using an orthogonal signature sequence can be regarded as noise, the decoding of the space-time coded CDMA system can use the maximum likelihood decoding of the conventional space-time code system as it is. have.

Hereinafter, a reception signal decoding process according to the present invention in a CDMA system space-time coded in a fading channel having an arrival time difference will be described.

2 shows an example of a receiver according to the present invention, and FIG. 3 shows a receiver for an mth receive antenna among the receivers.

The receiver according to the present invention decodes the received signal by the algorithm of the error correction decoder, and simultaneously updates the tap weight vector by the LMS algorithm using the Euclidean distance value used in the algorithm as an error value for the AIC. can do. As described above, in the receiver according to the present invention, the decoder and the AIC are organically coupled, but in the figure, the decoder 330 and the AIC bank 320 are separately shown for convenience of description.

The received signal vector is synchronized to the arrival delay time 321 from each transmit antenna and then despread by the tap weight vector w _{m, l} (i) in the adaptive interference canceller bank 320 instead of the desired user's signature sequence. The despread signal received in the i th symbol period is expressed by Equation 7:

In the above equation, I _m (i) represents an interference term including inter-transmission antenna interference, inter-symbol interference, and multiple access interference.

The tap weight vector is adaptively selected to minimize the mean squared error every symbol period, where a low complexity LMS (Least Mean Square) algorithm is used that is applicable to the mobile station.

The tap weight vector multiplied by the received signal vector synchronized with the l th transmit antenna in the (i + 1) th symbol period is expressed by Equation 8:

In the above formula,

e _m (i) is determined by the received signal [y _m (i)] of the i th symbol period and the i th symbol period. Received signal corresponding to Error Indicates,

μ represents the step size of the LMS algorithm.

The initial value of the tap weight vector is selected as the signature sequence of the desired user, and is limited to | w ^H _{m, l} (i + 1) a ₁ | = 1 to ensure convergence.

On the other hand, if there are enough users, assuming that η _m (i) including interference and noise follows a Gaussian distribution, maximizing the likelihood function is given by Equation 9:

In the above formula,

x ₁ is the frame sequence of the desired user ( x ₁ (0), x ₁ (1), ..., x ₁ (F-1)).

In this embodiment, a maximum likelihood decoder 330 using a Viterbi algorithm is used for error correction. The Viterbi algorithm selects a path having a minimum state transition path from the Viterbi comparison selector 331 through a variable of branch metric and path metric and outputs the decoding result.

In the Viterbi algorithm, when the transition from the state s _i to s _{i + 1} , the branch and path measurements are respectively ΔSi _{+ 1} (s _i | s _i-1 ) and Γ _{Si + 1} (s _i | s _{i- 1} ), the branch weighing is as shown in Equation 10:

In the above formula,

ε (s _{i + 1} | s _i ) is the Euclidean distance [χ (x ₁ (i))-y (i)] for the state transition associated with the i th symbol, which is the error signal of the adaptive interference canceller 322 use.

In the present invention, an algorithm for simultaneously performing interference cancellation by obtaining the error signal in the decoding process is as follows:

For state s _{i + 1} All states that transition to s _i

All state transitions to the state s _i s _i-1 For

Calculation

End

Selection

Calculation

Save

End

Optimal Euclidean distance Feedback to the error signal e _m (i)

Update tap weights for (i + 1) th symbol intervals by LMS algorithm

From here, Means the variable selected to minimize weighing.

As described above, the Euclidean distance [χ (x ₁ (i))-y (i)] for the state transition associated with the i-th symbol is obtained and used as an error signal of the adaptive interference canceller.

Figure 4 shows in more detail the algorithm, namely the Viterbi algorithm for decoding, the algorithm for obtaining the optimal Euclidean distance, and the adaptation algorithm (LMS algorithm) for tap weight update. As described above, in the receiver according to the present invention, since the decoder and the AIC are organically connected, in the drawing, the algorithm is represented in the ML decoder 330 to facilitate the understanding of the drawing.

5 is a flowchart illustrating a decoding process according to the present invention. As described above, the received signal [y _m (i)] of the i th symbol period is determined as a symbol in the i th symbol period. Received signal corresponding to Obtain the error e _m (i) between (S10), and then tap weight vector w _{m, l} (i + 1) for the (i + 1) th symbol by the LMS algorithm of Equation 8 from the value of e _m (i). i + 1) is updated (S20). Thereafter, by despreading the received signal using the updated tap weight vector (S30), the received signal may be decoded while reducing the interference.

The decoding method according to the present invention obtains an error signal in units of symbols in the process of decoding in units of frames, and simultaneously performs adaptive interference cancellation and maximum likelihood decoding. Thus, no additional symbol determination apparatus is required, and the configuration is simplified.

Hereinafter, the simulation results using the receiver according to the present invention will be described.

Assuming that the estimation of channel fading and arrival delay time is accurate in the receiver, simulation is performed in a DS-CDMA system with the number of transmit antennas L = 2 and the number of receive antennas M = 1, 2.

As the signature sequence, a gold sequence of length 31 was used. As the BPSK space-time code, 1/2 convolutional code (64, 74) with state number 8 is used, and the space-time code proposed by Tarokh et al. With state number 8 and 16 is used as the QPSK space-time code.

One frame consists of 260 symbols, and the normalized fading rate j _D T _{f, which} is the product of the maximum Doppler frequency j _D and the frame period T _f , is assumed to be 0.001 and the arrival delay time τ ^l _m ∈ [0, 10Tc]. .

FIG. 6 shows the bit error rate (BER) of a BPSK space-time coded DS-CDMA system with L = 2, M = 1, and number of states 8. FIG. The degree of performance degradation of about 3dB occurred if the arrival time difference exists compared with the standard performance in the case of a single user without a (K = 1), time difference of arrival indicated by the solid line channel (τ ^l _m = 0). This degradation decreases rapidly as the number of users (K) increases, so the maximum likelihood decoding (ML) receiver does not operate properly due to severe performance degradation. However, it can be seen that the receiver (AIC + ML) according to the present invention alleviates performance degradation.

7 illustrates bit error rate when M = 2 in the same channel environment. As the number of reception antennas increases, the error signal returned from the maximum likelihood decoder to the adaptive interference canceller becomes accurate. Therefore, the receiver (AIC + ML) according to the present invention performs its performance by correctly inputting the signal whose interference has been removed to the maximum likelihood decoder. This is improved.

8 shows the frame error rate (FER) of a QPSK space-time coded DS-CDMA system with L = 2, M = 2, E _b / N _o = 20 dB, number of states 8 and 16, in a Rayleigh fading channel with arrival time differences. It is shown. In the case of using the QPSK space-time code, the data rate is improved, but the performance improvement is slightly reduced than in the BPSK space-time code. Nevertheless, even in a QPSK space-time coded DS-CDMA system, the receiver (AIC + ML) according to the present invention can accommodate more users than the conventional receiver (ML). In addition, it can be seen that the degree of performance increase increases as the number of states increases.

The receiver for a CDMA system according to the present invention decodes the received signal by the algorithm of the error correction decoder, and simultaneously uses the Euclidean distance value used in the algorithm as the error value for the AIC. Updates can be made simultaneously.

In addition, even by the simulation, the conventional maximum likelihood decoding receiver is difficult to apply due to severe performance degradation in a fading channel having arrival time difference, but the adaptive interference cancellation and the maximum likelihood decoding composite receiver according to the present invention alleviate the performance degradation. It can be seen that the excellent performance. The degree of performance improvement increases as the number of receive antennas and the number of space-time codes increase.

Claims (11)

A receiver for a CDMA communication system that is space-time encoded in a fading channel having a time difference of arrival between a plurality of transmit antennas and a plurality of receive antennas, A receiver for a CDMA system, comprising: a maximum likelihood decoder with an error correction decoder, and an adaptive interference canceller (AIC) corresponding to each channel. 2. The receiver of claim 1, wherein the adaptive interference canceller updates the tap weight vector every symbol period and despreads the received signal received from each channel into the updated tap weight vector. The reception symbol of claim 2, wherein the tap weight vector is determined by the received signal y _m of the previous symbol period and the previous symbol period every symbol period. Received signal corresponding to A receiver for a space-time encoded CDMA system, wherein the receiver is updated by an LMS algorithm. The reception symbol of claim 3, wherein the reception signal y _m of the previous symbol period and the reception symbol determined in the previous symbol period are determined. Received signal corresponding to And an error value is an optimal value of a Euclidean distance value used for decoding the received signal y _m of the previous symbol period in the error correction decoder. The reception symbol of claim 4, wherein an optimal value of the Euclidean distance value is determined in the previous symbol period in the error correction decoder. Receiver for a space-time encoded CDMA system, characterized in that it is a Euclidean distance in a state transition that minimizes the pathometrics for. 6. The receiver of claim 5, wherein the error correction decoder is a Viterbi decoder or a turbo decoder. 4. The receiver of claim 3 wherein the initial value of the tap weight vector is a signature sequence of a particular user. A received signal decoding method in a CDMA communication system, which is space-time encoded in a fading channel having a time difference of arrival between a plurality of transmit antennas and a plurality of receive antennas, Received as the received symbol of the i-th symbol interval in the m-th receive antenna Received signal corresponding to Received signal between symbol and i-th symbol section Calculating an error value of the liver; (B) updating the tap weight vector by the LMS algorithm from the error obtained in step (a); And (C) despreading the signal received in the i + 1 th symbol interval using the tap weight vector obtained in step (b) instead of the signature sequence for each user. And decoding method. The method of claim 8, Said step (a) And above The error value between And an optimum Euclidean distance value used in a Viterbi decoding algorithm for decoding the received signal y _m (i) of the i-th symbol section. 10. The reception symbol of claim 9, wherein an optimal value of the Euclidean distance value is determined in the i-th symbol interval in the Viterbi decoding algorithm. The Euclidean distance in the state transition to minimize the path measurement for the interference cancellation and decoding method of the received signal. 9. The method of claim 8, wherein the tap weight vector is a vector for despreading a received signal synchronized to an lth transmit antenna.