Channel estimation for time division duplex communication systems

ES 2 255 010 T3

DESCRIPTION

Channel estimation for time division duplex communication system.

This application claims priority over US Provisional Patent Application No. 60 / 175,167, filed on January 7, 2000.

Background

The invention relates generally to wireless communication systems. In particular, the invention relates to channel estimation in a wireless communication system.

Figure 1 is an illustration of a wireless communication system 10. Communication system 10 has base stations 12₁ to 12₅ communicating with user equipment (UEs) 14₁ to 14₃. Each base station 12₁has an associated operational area in which it communicates with the UEs 14₁ to 14₃ in your operational area.

In some communication systems, such as code division multiple access (CDMA) and time division duplex using code division multiple access (TDD / CDMA), multiple communications are sent on the same frequency spectrum. These communications typically differ in the code sequences of their chips. To use the frequency spectrum more efficiently, TDD / CDMA communication systems use repetitive frames divided into time slots for communication. A communication sent in such a system will have one or more chip codes associated with it and time slots assigned to it depending on the bandwidth of the communication.

Because multiple communications can be sent on the same frequency spectrum and at the same time, a receiver in such a system must distinguish between the multiple communications. One method for detecting such signals is simple user detection. In simple user detection, a receiver detects only communications from a desired transmitter using a code associated with the desired transmitter, and treats signals from other transmitters as interference. Another procedure is called co-screening. In joint detection, multiple communications are detected simultaneously.

To use these detection techniques, it is desirable to have an estimate of the wireless channel through which each communication travels. In a typical TDD system, channel estimation is done using intermediate sequences in communication bursts.

A typical communication burst 16 has an intermediate 20, a guard period 18, and two data bursts 22, 24, as shown in Figure 2. The intermediate 20 separates the two data bursts 22, 24 and the guard period. 18 separates the communication bursts 16 to account for the difference in arrival times of the bursts 16 sent from different transmitters. The two data bursts 22, 24 contain the data of the communication burst. Intermediate 20 contains a training sequence that is used to estimate the channel.

After a receiver receives a communication burst 16, it estimates the channel using the received intermediate sequence. When a receiver receives multiple bursts 16 in a time slot, it typically estimates the channel for each burst 16. One method for this channel estimation for communication bursts 16 sent over multiple channels is a Steiner Channel Estimator. The Steiner Channel Estimator is typically used for uplink communications from multiple UEs, 14₁ to 14₃, where the channel estimator has to estimate multiple channels.

Steiner and Jung's "Optimal and Sub-Optimal Channel Estimation for Uplink CDMA Mobile Radio Systems with Junction and Detection" describes a procedure for channel estimation. One procedure uses a simple cyclic correlator. Using the known transmitted intermediate sequences, an M matrix is ​​constructed. The received intermediate vector e is multiplied by the first column of M. The multiplication is carried out by means of the cyclic correlator on P values ​​and by shifting the values ​​2P - 1 times. P is a period of the intermediate codes.

In some situations, multiple bursts 16 use the same wireless channel. One case is a high data rate service, such as a 2 megabits per second (Mbps) service. In a system like this, a transmitter can transmit multiple bursts in a single time slot. The Steiner estimation can be applied in such a case by averaging the estimated responses of the channels for all bursts 16. In any case, this procedure is highly complex. Accordingly, it is desirable to have alternative procedures for channel estimation. Compendium

A single transmitter transmits K communication bursts in a shared spectrum in a time slot in a time division duplex communication system. The system is associated with N intermediate sequences. Each burst has an associated breakthrough sequence. A receiver receives a vector corresponding to the transmitted intermediate sequences of the K communication bursts. A matrix with N matrix blocks

Identical circulating ES 2 255 010 T3 is constructed based in part on the N known intermediate sequences. The wireless channel between the transmitter and the receiver is estimated based in part on one of the N blocks and the received vector.

Brief description of the drawing (s)

Figure 1 is a wireless communication system.

Figure 2 is an illustration of a communication burst.

Figure 3 is a simplified multi-burst transmitter and receiver.

Figure 4 is a flow chart of a multi-burst channel estimate.

Detailed description of the preferred embodiment (s)

Figure 3 illustrates a simplified multi-code transmitter 26 and receiver 28 in a TDD / CDMA communication system. In a preferred application, such as the 2 Mbs downlink service, the receiver 28 is in a UE 14₁ and the transmitter 26 is in a base station 12₁although receiver 28 and transmitter 26 can be used in other applications.

Transmitter 26 sends data over radio channel 30 wirelessly. The data is sent in communication Kráfagas. Data generators 32₁ to 32_K at the transmitter 26 generate data to be communicated to the receiver 28. The modulation / expanding and training sequence inserters 34₁ to 34_K they expand the data and time multiplex the expanded reference data with a training intermission sequence in the appropriate assigned timeslot and codes to expand the data, producing the K communication bursts. Typical values ​​of K for a base station 12i transmitting downlink bursts are 1 to 16. The communication bursts are combined by a combiner 48 and modulated by a modulator 36 to radio frequency (RF). An antenna 38 sends the RF signal through the wireless radio channel 30 to an antenna 40 of the receiver 28. The type of modulation used for transmitted communication can be any of those known to those skilled in the art, such as shift modulation. binary phase (BPSK) or quadrature phase shift keying (QPSK).

The antenna 40 of the receiver 28 receives various radio frequency signals. The received signals are demodulated by a demodulator 42 to produce a baseband signal. The baseband signal is processed, for example by a channel estimating device 44 and a data detection device 46, in the time slot and with the appropriate codes assigned to the transmitted communication bursts. The data detection device 46 can be a multi-user detector or a single user detector. Channel estimator 44 uses the intermediate training sequence component in the baseband signal to provide channel information, such as pulsed channel responses. The channel information is used by the data detection device 46 to estimate the transmitted data from the received communication bursts as binary symbols.

To illustrate an implementation of a multi-burst channel estimate, the following type of intermediate is used, although the multi-burst channel estimate is applicable to other types of intermediate. The K intermediate codes, m^(K), where k = 1 ... K, are obtained as time-shifted versions of a single basic periodic intermediate code, m_p, period P chips. The length of each intermediate code is L_m = P + W - 1. W is the length of the impulse response of the user channel. Typical values ​​of L_m they are 256 and 512 chips. W is the length of the user channel's impulse response. Although the following reasoning is based on each burst having a different intermediate code, some intermediates may have the same code. As a result the analysis is based on N intermediate codes, N <K. Additionally, the system can have a maximum number of acceptable N intermediate codes. The receiver 28 in such a system can estimate the channel for the maximum number N of codes, even if less than N codes are transmitted.

The elements of m „take values ​​from the set of integers {1, -1}. The sequence m „first becomes a complex sequence iñ_P[i] = j · m_P[i], where i = 1 ... P. The m^(k) they are obtained by choosing K sub-sequences of length Lm from a sequence of length 2P formed by concatenating two periods of mn. The i-th element of m^(k) is related to mn by Equation 1.

m^(k) = m_P[(K - k) W + i], for 1 <i <P - (K - k) W = m_P [i - P + (K - k) W] for P - (K - k) W <i <P + W - 1 Equation 1

Therefore, the starting point of m^(k), k = 1 ... K shifts W chips to the right as k increases from 1 to K.

The combined received intermediate frequencies are a superposition of the K convolutions. The k-th convolution represents the convolution of m^(k) with h^(k) . h^(k) is the k-th user's channel response. The preceding data field in the burst corrupts the first (W-1) chips of the received intermediate. Therefore, for the purpose of channel estimation, only the last P of Lm chips are used to estimate the channel.

ES 2 255 010 T3

The multi-burst channel estimation will be explained in conjunction with the flow diagram in Figure 4. To solve the individual channel responses h^(k), Equation 2 is used.

<td>m_pZHj</td><td> '' ^(^-1)^+2 ’' ^m(K - \) W + 3</td><td>^m(£ -lX + l ' ^τη(κ-ήψ + 2 '</td><td>^m(K-2) ir + i '· ' ^m(K-2) lT + 2 ^: ’ ^m(K-2) go + 3 · '</td><td> ’ ^mw^{: m}w + \ ' : tn_{w + 2} ·</td><td>-j * 3 J __1</td><td>X</td><td>~ h^m'</td><td> =</td><td>^rw^rtr + i^rgo + 2</td>

<td></td><td></td><td></td><td> ^(^-2)^ ^:</td><td></td><td>• m_p</td><td></td><td>1- S '</td><td></td><td> _^rG.</td>

Equation 2 r_W · · · Are the combined received chips of the intermission sequences. The m values ​​are the elements of m_P.

Equation 2 can also be rewritten in abbreviated form using Equation 3.

Σ M^(k)h^(k) = r k - 1

Equation 3

Every M^(k) is a matrix of KW times W. r is the response of the intermediate chip received. When all the bursts travel through the same channel, h⁽¹⁾ · · · H^(k) they can be substituted for h as seen in Equation 4, 50.

Σ M^(k)k = 1

Equation 4

G is defined by Equation 5.

G = [M⁽¹⁾, ..., M^(k), ..., M^(K)]

Equation 5

As a result, G is a KW times KW matrix. Since G is a right circulating matrix, Equation 4 can be rewritten using K identical circulating right matrix blocks B, as shown in Equation 6.52.

<img file="ES2255010T3_D0001.tif" />

B is a right-hand circulating matrix of W by W. The number of blocks B is K. Using Equation 6, Equation 4 can be rewritten by Equation 7.

Dh = r

Equation 7

Equation 7 describes an overdetermined system of dimensions KW by W. A procedure to solve Equation 7 is a least squares solution, 54. The least squares solution of Equation 7 is given in Equation 8.

h = (D^HD)^-1D^Hr

Equation 8

D^H is the Hermitic of D.

ES 2 255 010 T3

Applying Equation 6 to Equation 8 results in Equation 9.

(D^HD)^-1 = - (B^HB)^-1K

Equation 9

The received vector r of dimension KW can be decomposed using Equation 10.

n

Γ2

Tk

Equation 10

The dimension of r_k is W. Substituting Equations 9 and 10 in Equation 8, the least squares solution for the channel coefficients by Equation 11 results.

h = (B^HB)^-1B^H k \ _

-ΣΑ = (B^HB)^-1B^H rk

K k = 1 /

Equation 11 r_k represents the mean of the segments of r. Since B is a square matrix, Equation 11 becomes Equation 12.

h = B^-1 fk

Equation 12

Since B is a right circulating matrix and the inverse of a right circulating matrix is ​​also a right circulating matrix, the channel estimator can be implemented by a simple cyclic correlator of dimension 57, or by a discrete Fourier transform (DFT) solution .

A DFT method at point Wes as follows. Since B is right-hand circulating, Equation 13 can be used.

^B - ^DW ' ^TOC ^DW

D_w is the point W of the DFT matrix of Equation 14.

Equation 13

<td>~ w °</td><td>w °</td><td>w^to</td><td> ··</td><td> W ° '</td>

<td>iv °</td><td></td><td>w²</td><td>jy³ ·</td><td></td>

<td>w °</td><td>w²</td><td>w *</td><td>jy⁶ .</td><td>. jy2 (tr-t)</td>

<td>w °</td><td>w³</td><td>w⁶</td><td>jy⁹ ·</td><td>. jy3 (f-t)</td>

<td>w °</td><td></td><td></td><td></td><td></td>

Equation 14

TO_C is a diagonal matrix whose main diagonal is the DFT of the first column of B, as follows from Equation 15.

^toc = ^{diag (D}w^{(B (}:,¹⁾⁾⁾

Equation 15

ES 2 255 010 T3

W = e fw. Therefore, D_W is the DFT operator such that D_Wx represents the W DFT point of vector x. Substituting the

Equation 13 in Equation 12 and using DW¹ = -, Equation 16 is obtained.

^WW h = I DW · Éz · A^-1 DW r

Equation 16

D^*W is the conjugate of DW element-by-element.

Alternatively, an equivalent way of expressing h as a function of A can be obtained_R instead of A_C. Ar is a diagonal matrix whose main diagonal is the DFT of the first row of B, as follows from Equation 17.

Ar = diag (D_w(B (1, :)))

Equation 17

As the transpose of B, B^T, is also circulating right and since its first column is the first row of B, B^TIt can be expressed by Equation 18.

B^T = DW¹ ArD_w Equation 18

Using Equation 18 and that DW = D_W, AR = A_R and that for any invertible matrix A, (A^T)^-1 = (A^-1)^T, B can be expressed as shown in Equation 19.

^B = ^DW ^TOR ^{D 1}W Equation 19

Substituting Equation 19 in Equation 12 and knowing that DW = Equation 20 is obtained.

^h = ^dw ^tor¹ —DWI · r _WW

Equation 20

Equations 16 or 20 can be used to obtain h. As all the DFTs are of length W, the complexity of solving the equations is substantially reduced.

A procedure using a simple loop correlator is as follows. Like B^-1 is the inverse of a right circulating matrix, it can be written as shown in Equation 21.

T_r - T₃ T₂T₂ T¡ - Γ, T₃

T_w_y T¡r-2 ^ 1. T_w ^ wi ^ 2 ^ 1.

Equation 21

The first row of the matrix T is equal to the inverse DFT of the main diagonal of A_R¹. Therefore, the matrix T is completely determined by A_R¹.

The elements of the response of the channel h are obtained successively by means of an internal product of the successive rows of T with the average of the segments of length W of the received vector r. Successive rows of T are versions circularly shifted to the right of the previous row. Using registers to generate the inner product, the first register stores the averaged r segments, and the second register is a shift register that stores the first row of the matrix T. The second register is circularly shifted at a certain clock rate. At each clock cycle, a new element of h is determined by the inner product of the vectors stored in the two registers. It is advantageous to shift the first row of the matrix T instead of the received intermediates. As a result, no extra storage capacity is needed for intermediates. Intermediates continue to reside in the received buffer that stores the entire burst. Since the length of the correlator is only W, a significant reduction in the complexity of the channel estimation is achieved.