PILOT SIGNAL MANIPULATION IN TRANSMIT DIVERSITY COMMUNICATIONS
The invention relates to transmit diversity wireless communications schemes.
In a transmit diversity scheme, several transmitting antennas are used simultaneously to convey a signal to a receiver. In simple cases, the receiver will have a single reception antenna.
The current standards for UMTS require user equipment (UE), such as mobile telephones, to receive and process correctly signals that have been sent to the UE with transmit diversity. A discussion of the transmit diversity schemes that are currently included in UMTS can be found in "Introduction to 3G Mobile Communications" by Ju a Korhonen, published by Artech House, MA, USA, at pages 89-93. These transmit diversity schemes can be of the open-loop, space time block coding transmit diversity (STTD) variety or the closed loop type. These forms of transmit diversity encoding will now be described by reference to a data signal, comprising a series of symbols, So, Si, S2 Sn, that is destined for transmission from a node B (a basestation) via a pair of antennae to a single-antenna receiver.
In STTD encoding, consecutive pairs of symbols are treated as blocks. Consider the block comprising symbols S0 and Si. At a time to, one of the transmit antennae transmits S0 and the other transmit antenna sends - Si* , the asterisk denoting the complex conjugate throughout this text. At a time ti, one symbol period later than t0, the antenna that sent S0 now sends Si and the other antenna now sends S . At the receiver, the symbols R0 and Ri are acquired during the symbol periods that correspond, respectively, to t0 and i. In mathematical terms:
where N
0 and Ni are additional noise terms and pj
. and p
2 are complex numbers describing the propagation environment between the receiver and, respectively, the transmit antenna that sent So and Si and the transmit antenna that sent - S,
* and S
* 0.
The receiver therefore acquires a stream of received symbols, Ro, Ri, R2,....Rn, and directs the stream to an STTD decoder, where the received symbols are handled in pairs corresponding to the transmit blocks. Consider the transmit block S0 and Si and the corresponding receive block Ro and Ri. The receive block is manipulated by the STTD decoder to produce two consecutive output symbols D0 and Oι using:
A> = Pl
R +
= (P Pi + (2) = PΛ - Pi = i iPi +
which consist of the data symbols So and Si plus additional noise. Terms p^nd /32 are estimates of the complex numbers pi and p2 obtained by measurements performed on pilot signals sent from the transmit antennae to the receiver. These pilot signals will be discussed in more detail later.
In the closed-loop transmit diversity scheme, the stream of symbols S0 Sn is sent from one of the transmit antennae and the same stream multiplied by a complex gain w is sent from the other transmit antenna. The value of w is known by the UE and, for closed loop mode 1, has a modulus of 1. Considering a given symbol, S, the corresponding symbol acquired by the receive antenna at the UE is R, where:
R = pl - S + p2 - w - S + N = (p[ + p2 - w)- S + N = p - S + N (3)
where N is the additional noise term, p = pi + p2-w and i and p2 are complex numbers describing the propagation environment between the receiver and, respectively, the transmit antenna that sends symbol S and the transmit antenna that sends symbol w.S.
The UE passes the receive sample R to a decoder, where a corresponding output symbol D is produced using:
which consists of the data symbol S plus additional noise. Terms px and p2 are estimates of the complex numbers pi and p2 obtained by measurements performed on pilot signals sent from the transmit antennae to the receiver. These pilot signals will be discussed in more detail later.
A 3 GPP - compliant UE receiver needs to measure the signal to interference ratio (SIR) of the channels that it receives. This information can be returned to the node B that is transmitting to the UE through these channels to allow the node B to perform power control on these channels. Pilot bits are time-multiplexed into 3 GPP Release 99 DPCH channels and a UE provided with knowledge of these pilot bits can directly estimate the SIR of such DPCH channels. However, some channels, for example HSDPA data channels, do not contain multiplexed pilot bits and their SIRs need to be estimated by adjusting an SIR measured for a pilot channel to take account of the difference in spreading factors between the pilot channel and the channels whose SIRs are to be estimated. However, where one of the two transmit diversity schemes described above is employed, there are problems associated with attempting to use SIR information from the pair of pilot signals as a basis for estimating the SIR of a transmit diversity encoded data signal, as will now be discussed.
In both of the two transmit diversity schemes described above, the pilot signals, one from each of the transmit antennae, are sent in the same pilot channel but are arranged to be orthogonal to one another to facilitate the derivation of the estimates px and p2 or, as the case may be, the estimate p . It can be said that the pilot signals use their orthogonality as a form of transmit diversity encoding to protect against interference between them, whereas the data channel transmissions use encoding according to equation (2) or, as the case may be, (4). This difference in the transmit diversity encoding employed in the pilot and data channels means that SIR information from the pilot signal cannot be adapted in the same way as in the absence of transmit diversity.
One aim of the invention is to provide a way of manipulating information received through the pilot channel in a scheme of one the two types discussed above to render that information suitable for making calculations such as the estimation of the SIR of the accompanying data channel.
According to one aspect, the invention provides a method of producing a pseudo pilot signal from a pair of orthogonal pilot signals sent in a pilot channel from a pair of antennae that are also transmitting a transmit diversity encoded data signal to a single-antenna receiver, the method comprising acquiring a pair of samples on the pilot channel tlirough the receiver antenna and combining the pilot channel samples with channel estimate values derived for the pilot signals to produce one or more values of the pseudo pilot signal.
The invention also consists in apparatus for producing a pseudo pilot signal from a pair of orthogonal pilot signals sent in a pilot channel from a pair of antennae that are also transmitting a transmit diversity encoded data signal to a single-antenna receiver, the apparatus comprising sampling means for acquiring a pair of samples on the pilot channel through the receiver antenna and processing means for combining the pilot channel samples with channel estimate values derived for the pilot signals to produce one or more values of the pseudo pilot signal.
In certain embodiments the pilot signals comprise a first pilot signal comprising a repeated symbol and a second pilot signal comprising the same symbol alternated with its complex conjugate.
If one regards the acquired pilot channel samples as R0 and Ri and the channel estimates for the first and second pilot signals as ,and p2 respectively, then, in preferred embodiments, the process of forming the products of the pilot channel samples and the channel estimates follows one of the following calculations:
(p
2 ■ R
0 + p ■ R,)+ j • (p ■ R
0 + p • R, J
(p2 • R1 + t31 * - R0 )+ y - (/ 2 - Ri + Λ* - ^o )
(Pl +P XRo +Ri +HRo -Ri))
where w is a complex value known to the receiver.
In some embodiments, an open-loop STTD scheme is used to encode the data signal. In such embodiments, the STTD scheme may transmit two symbols a and b from one of the antennae and, at the same time, send symbols -b and a from the other antenna.
In some embodiments, a closed-loop transmit diversity scheme is used to encode the data signal. In such embodiments, the transmit diversity scheme may be arranged to send, for each symbol and simultaneously from the antennae, a symbol of the data signal from one antenna and the same symbol scaled by a complex gain value from the other antenna.
The pseudo pilot signal developed by the invention can be used to estimate the SIR of a transmit diversity encoded data signal that accompanies the pilot signals.
The invention also relates to a radio unit, such as a mobile telephone, that is arranged to derive a pseudo pilot signal according to the invention.
From another perspective, the invention also relates to a program for causing data processing apparatus to perform the process of pseudo pilot signal generation according to the invention. Such a program can be conveyed by a suitable carrier, e.g. a type of ROM.
By way of example only, certain embodiments of the invention will now be described with reference to the accompanying drawing, Figure 1, which is a block diagram illustrating a radio link between a node B and a UE in a UMTS network. The drawing shows only the key elements involved in generating and manipulating a pseudo pilot signal.
A node B 10 is arranged to transmit a transmit diversity encoded data signal via two antennae 1 and 2 to a UE 12. The UE 12 has a single antenna 14 through which it acquires signals from both antennae 1 and 2. Each of antennae 1 and 2 sends its own pilot signal and a part of the encoded data signal to the UE 12. The pilot signals are mutually orthogonal.
The pilot signal from antenna 1 has the form:
A A A A A A A A A A ...
and the pilot signal from antenna 2 has the form:
A -A -A A A -A -A A A -A ...
The value of the symbol A is 1+j.
The signal acquired by the UE 12 on antenna 14 is demodulated at RF section 16 and converted into a train of digital samples by analogue to digital conversion (ADC) unit 18. This digital signal is then processed by a suite of processing and memory resources to recover and exploit the data contained in the transmit diversity encoded data signal. These resources implement, inter alia, a transmit diversity decoder 20, a channel estimator 22, a pseudo pilot creator 24 and an SIR estimator 26.
The transmit diversity decoder 20 decodes the transmit diversity encoded data signal using an appropriate algorithm. For example, if STTD encoding is used, then decoder 20 applies equation (2). The decoded data signal is then put to its intended use within the UE.
The channel estimator 22 isolates the orthogonal pilot signals and uses them to estimate the properties of the propagation environment between antennae 1 and 2 and antenna 14. The channel estimator produces complex values j^and p2 which are complex numbers describing the propagation environment between antennae 1 and 2 respectively and antenna 14. The values jjj-tnd p2 are updated periodically using the relation R„ = pn .Sn where pn is the complex number describing the propagation environment between antenna n and antenna 14 when antenna n sent symbol Sn and antenna 14 correspondingly received symbol Rn.
The pseudo pilot creator 24, as will be described in more detail later creates a pseudo pilot signal from the symbols acquired from antenna 14.
The SIR calculator 26 deduces from the pseudo pilot signal an SIR value for the decoded data signal.
Examples of pseudo pilot signal creation will now be given for both of the STTD and closed-loop transmit diversity schemes that were described in the introduction.
Open-loop STTD Coding
If Ro and Ri are two consecutive symbols received on the pilot channel, then either:
RQ = p A + p2 - A + N0 (5) R, Px ■ p2 - A + Nl
or: R0 = Pl - A - p2 - A + NQ (6) R, = px - A + p2 - A + N
where N0 and Ni are additional noise terms.
If the timing is such that equation (5) is appropriate, then two consecutive output samples D
0 and Di can be calculated from:
D
0 (?)
This calculation is performed by the pseudo pilot creator 24 which is so-called because Do and Di can be said to be the consecutive symbols that would be received if a pilot signal of the form AAAA....A were sent using the STTD encoding scheme that is applied to the data signal. Using the relation j. A = A, equation (7) can be rearranged into:
A + A #ι + J
■ PiK + J
■ Pi
(8) A
- AX + j
■ P N
Q - j
■ p
2N[
It will be apparent that the transmitted pseudo pilot signal symbol A has been scaled in equation (8) by a coefficient 2(pl *.pl + p2.p2) . By contrast, it will be seen that the transmitted data symbols Si and S appear in equation (2) scaled by a coefficient (A* -A + A. -Pi ) • That is to say, the scaling in amplitude of the wanted signal in equations (1) and (2) is a factor of 2 lower than the scaling of the amplitude of the pseudo pilot signal in equations (7) and (8).
One can perform a similar comparison of the noise terms of equation (2) with these of equation (8) and demonstrate that the noise power in equation (2) is a factor of 2 lower than the noise power in the pseudo pilot signal of equation (8). In short, the noise terms in each line of equation (8) comprise four product values, each product value being the product of a noise value and a channel estimate. In equation (2), on the other hand, only two such product values appear in each line. As these product values are incoherent, their total power is given by the sum of the power values of the individual product values. Also, the power contained in each of these product values is broadly the same, so the noise power is equation (8) is approximately double the noise power in equation (2).
Thus, the SIR calculator 26 can use the scaling factors discussed above to calculate the SIR of the STTD encoded data signal. That is to say, the signal power from equation (8) is reduced by a factor of 4 and the noise power by a factor of 2 when using the signal and
noise power values of the pilot signal to deduce an SIR for the data signal. Alternatively stated, the SIR of equation (8) can be halved to yield an SIR for the data signal.
It should be noted that the pseudo pilot signal creator 24 can use the following equation instead of equation (7):
D = p;{R0 + Rl)+ p;(R0 - Rl) (9)
This calculation produces output symbols of a pseudo pilot signal that has the same signal amplitude and noise power relationships with the decoded data signal as equation (7). However, equation (9) only produces one output pseudo pilot symbols per pair of consecutive symbols acquired from the pilot channel. The symbols produced by the pseudo pilot creator 24 can be used as described above to produce an SIR value for the STTD encoded data signal.
If the timing is such that equation (6) is appropriate then equation (7) is replaced by:
A = (A • ^i + A • ^o)+ j ■ (pi ■ Rι + A ' o) (10) A = (A - Rι ~ P • ^,)+ j - (pι* - Rι ~ P - R )
equation (8) is replaced by:
A =
2(A A + )A + P N
x + AX +
■ P
2N + j
■ ANo ,
) A =
2(A A +
+ AX - AX + J
■ PΛ -
• PiK
and equation (9) is replaced by:
D = p {R0 + Ri) + p2(Rl - R0) (12)
Again, it is seen that the pseudo pilot symbol A in equations (10) and (12) has been scaled by a coefficient that is twice the size of the coefficient that scales the STTD decoded data signal symbols S0 and Si in equation (2) and that the noise power in equations (10) and (12) is twice as great as the noise power in equation (2). As before, the pseudo pilot output
samples that are produced can be used by the SIR calculator 26 to estimate an SIR for the STTD encoded data signal.
Closed-loop Coding
It is assumed that closed loop mode 1 is used such that |w| = 1.
If R0 and Ri are two consecutive symbols received on the pilot channel, then we have the same possibilities as before. That is to say, either:
R0=p A + p2-A + N0 (5) R1=p A-p2-A + Nl
or:
Ro=Pι-A-ρ2-A + N0 (6) Rl=p1-A + p2-A + Nl
If the timing is such that equation (5) is appropriate, then one can calculate one output sample D using:
D = p (R0+Rl+w(R0-Rl)) (13)
This calculation is performed by the pseudo pilot creator 24 which is so-called because D can be said to be the symbol that would be received if a pilot signal of the form AAAA...A were sent using the closed-loop encoding scheme of equation (3). It should be noted that equation (13) can be rearranged as:
D = 2- p p- A + p (l + w)N0 + p (l-w)Nl (14)
It will be apparent that the transmitted pseudo pilot symbol A has been scaled in equation (14) by a coefficient 2p* p . By contrast, it will be seen that the transmitted data symbol S
appears in equation (4) scaled by a coefficient p*p . That is to say, the scaling of the amplitude of the wanted signal in equations (3) and (4) is a factor of 2 lower than the scaling of the amplitude of the wanted signal in equations (13) and (14).
One can perform a similar comparison of the noise terms of equation (4) with those of equation (14). In equation (4), the noise term is p* N and in equation (14) the noise term is p* {(\ + w)N0 + (l + w)N, } so the power of the equation (14) noise term is a factor of w\ ) layer than the power of the noise term in equation (4). Since w has a modulus of 1 for close loop mode 1 transmission, this factor is 4 here.
Thus, the SIR calculator 26 can use the scaling factors described above to estimate the SIR of the closed-loop encoded data signal.
If the timing is such that equation (6) is appropriate, then equation (13) is replaced by:
D = p (R0 + Rl + w(Rl - Rϋ)) (15)
and equation (14) is replaced by:
D = 2 - p* p ■ A + p*(l - w)N0 + p* (l + w)Nr (16)
Once more, it will be seen that the pseudo pilot signal symbol A in equation (16) has been scaled by a coefficient that is twice the size of the coefficient that scales the data signal symbol S in equation (4) and that the noise power in equation (16) is a factor of 4 greater than the noise power in equation (4). The pseudo pilot signal output symbols that are produced can therefore be used in the manner previously described to deduce an SIR value for the closed-loop encoded data signal.