WO2004075434A1

WO2004075434A1 - Method for processing antenna verification.

Info

Publication number: WO2004075434A1
Application number: PCT/IB2004/000325
Authority: WO
Inventors: Eric Alliot
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2003-02-19
Filing date: 2004-02-03
Publication date: 2004-09-02

Abstract

The present invention relates to a method for processing antenna verification by means of a first channel (CPICH) and of pilot symbols of a second channel (DPCH) at a receiver, said receiver being associated to a base station (BS) transmitter comprising two antennas (ANT1, ANT2) with associated weight (w1, w2), said pilot symbols being sent during time slots in said second channel (DPCH). The method comprising the steps of: - performing directly a first estimate of the second antenna's weight (w2) using the orthogonality of the pilot symbols of the first and second antennas (ANT1, ANT2) and using known channel estimates ( , ) of said first channel (CPICH), - making a hard decision on the sign of said first estimate, - evaluating the reliability of said first estimate using a sign property between the complex possible values of said second antenna's weight (w2) of successive slots, and - if the reliability is not verified, performing a correction of said first estimate using feedback information (FBI) that has been sent in an uplink and applied to a preceding downlink slot of said second channel (DPCH). Use: Receiver in a UMTS communication system.

Description

Method for processing antenna verification

Field of the invention

The present invention relates to a method for processing antenna verification on a receiver. The invention further relates to an associated receiver using said method.

Such a method may be used in particular in any UMTS communication system.

Background of the invention

A UMTS communication system, defined in the UMTS standard referenced "3GPP Technical Specifications Rel.99 of March 2002 referenced 25.211 v3.10.0 http://www.3gpp.org/specs/specs.html", comprises at least a base station (BS) and a mobile phone terminal, referred to as user equipment (UE). When the BS sends signals to the UE receiver, the communication is called downlink, and when the UE sends signals to the BS receiver, the communication is called uplink. In both types of communications the UE and BS map the data to be transmitted on physical channels. In particular for the downlink communication two physical channels are provided, a first channel denoted common pilot channel (CPICH) and a second channel denoted dedicated physical cham el (DPCH). Said second channel DPCH comprises a first dedicated physical sub-channel denoted dedicated physical data channel (DPDCH), that carries some information data symbols intended for the user of interest, time multiplexed with a second physical sub-channel denoted dedicated physical control channel (DPCCH), that carries known user dedicated pilot symbols.

In UMTS, when the downlink DPCH is transmitted from two antennas in the BS, the DPCH is weighted differently on both antennas. The UE receiver estimates the weights applied at the BS in order to recover correctly the signal sent from the two antennas. This process of estimating the weights is called antenna verification. These weights correspond to some phase settings applied to the two antennas.

The 3 GPP specification describes a method of performing antenna verification at the UE in the presence of two antennas at the BS, which is as follows. - If the channel estimate is made on the CPICH, use antenna verification to correct the phase settings. In case of orthogonal pilot patterns on Downlink DPCCH, apply a beam former verification (SBV) for which the formulas of tests are described in the document TS 25.214 in annex A of March 2002 referenced 25.214 v3.10.0. The method makes it possible to obtain a rough estimate of the phase adjustment φ(i) by performing the following steps:

Making two different tests by means of two equations. These two equations have to be applied depending on the number of the current slot. One equation is made on the real part of symbols received, demodulated by the known pilots patterns, corrected by noise plus interference power, by DPCH SNIR/CPICH

SNIR and summed on different fingers; the other equation concerns the imaginary part of symbols received. Thus, these equations permit to calculate a rough estimate called φ-rx of the phase adjustment φ used at the base station BS. - Defining a phase intermediate variable x that takes a value depending on the number of the current slot (from 0 to 14) and on the result of the equation verified or not by φ-rx (formula given in annex A de 25.214). Calculating an estimate of the weight using the current and preceding value of n n

∑cos ) ∑sin ) x and the formula w₂ = ^■fe≤zl + jⁱ≡^ , n being a slot.

2 2 One problem of this solution is that it gives only a rough estimate of the phase adjustment φ at each slot. Consequently, the estimate of the weights has to be recalculated at each slot and no improvement is made on this rough estimate. The antenna verification, which is based on this rough estimate, is not totally reliable.

Summary of the invention

Accordingly, it is an object of the invention to provide a method and an associated receiver, which permits to process antenna verification with a more reliable result.

In addition, according to a first object of the invention, there is provided a method as claimed in claim 1. To this end, according to a second object of the invention, there is provided a receiver as claimed in claim 7.

As we will see in detail further on, the method according to the invention to perform antenna verification for closed loop mode 1 directly gives an estimate of antenna weights using a sign property noticed between the complex possible values of the second antenna weight of odd and even slots. The estimate of the antenna weights is accomplished by processing the received DPCCH pilot symbols, which are distorted by fading and AWGN channel. Such a method gives an affined estimate of the second antenna weight useable for recovering user information symbols, which are the useful information, without too much complexity and increasing performance. It uses fewer steps for the estimating process and economizes hardware.

In a first non-limitative embodiment of the invention, the method is characterized as claimed in claim 2.

In a second non-limitative embodiment of the invention, the method is characterized as claimed in claim 3.

In a second non-limitative embodiment of the invention, the method is characterized as claimed in claim 4. In a second non-limitative embodiment of the invention, the method is characterized as claimed in claim 5.

In a second non-limitative embodiment of the invention, the method is characterized as claimed in claim 6.

Brief description of the drawings

Additional objects, features and advantages of the invention will become apparent upon reading the following detailed description and upon reference to the accompanying drawings in which:

- Fig.1 illustrates an example of a UMTS communication system with a receiver according to the invention,

- Fig.2 shows the frame structure of a downlink physical channel DPCH used by the UMTS communication system of Fig.1, - Fig.3 illustrates a diagram of a method for antenna verification according to the invention,

- Fig.4 is a detailed step of the method for antenna verification of Fig.3, and

- Fig.5 is a graph illustrating some errors of phase corresponding to some errors of weights that may be corrected by the method of Fig.3.

Corresponding reference numerals for corresponding elements will be used throughout the description.

Detailed description of the invention In the following description, functions or constructions well known to the person skilled in the art are not described in detail since they would obscure the invention in unnecessary detail.

The present invention relates to a method for processing antenna verification. Said method is used in particular in UMTS communication systems SYS, of which an example is illustrated in Fig.1.

The UMTS communication system SYS comprises: - At least a base station BS comprising a first ANT1 and a second ANT2 antenna, and - user equipment UE, a mobile phone terminal for example.

Both the BS and the UE include a transmitter and a receiver. In this example, the BS sends signals R by means of the BS transmitter to the UE receiver in the downlink. The UE sends signals R by means of the UE transmitter to the BS receiver in the uplink. In uplink and downlink communications, both the UE and BS map the data to be transmitted into dedicated physical channels, said channels being transmitted through a propagation channel such as the air.

In particular the UMTS standard provides two physical channels for the downlink communication, a first common physical channel denoted common pilot channel CPICH consisting of a continuous sequence of known pilot symbols and a second dedicated physical channel denoted dedicated physical channel DPCH. Said second physical channel DPCH comprises a first physical dedicated sub-channel denoted dedicated physical data channel DPDCH, that carries the information data intended for the user, and a second physical dedicated sub-channel denoted dedicated physical control channel DPCCH, that carries known user-dedicated pilot symbols in one pilot field.

Fig.2 shows the frame structure of the downlink dedicated physical channel DPCH. Each frame of length 10 ms is split into 15 slots. Each slot has a length of 2560 chips and being split into the first sub-channel DPDCH comprising N data bits datal and data2 and the second sub-channel DPCCH comprising N tpc "transmit power control", giving command to the UE on the power of transmission to be used in uplink, N tfci "transport format combination indicator", giving information on the format of physical channels, N pilot bits. It may be noted that the signal sent from an antenna ANT is subjected to multiple reflections, refractions and diffraction in the air which correspond to multiple paths.

The method for processing antenna verification is based on the following hypothesis.

"TxD" (Transmit Diversity) means that the downlink DPCH channel is transmitted from two antennas in the BS,

- "Mode 1" means that the downlink DPCH channel is weighted differently on both antennas ANT1 and ANT2 with different complex weights applied on each antem a at BS, wl and w2 respectively, and that the UE knows the possible set of weight values.

- "Closed loop" means that a feedback-signaling message FSM sent from the UE to the BS is taken into account to weight the downlink DPCH channel at BS. Said message FSM comprises in particular for model phase settings φ to be applied on each antemia ANT corresponding to the different complex weights applied on each antenna, wl and w2. Said phase settings or complex weights wl, w2 can be deduced by the values coded in a feedback information field FBI of the FSM of the uplink DPCCH.

• The CPICH is not subjected to these weights but can give good estimates of channel responses (gl and g2) for signals coming from antenna 1 and antenna2 at BS, gl, g2 being some complex coefficients permitting the model of the propagation channel. • The DPCCH pilots used on each antenna are orthogonal for mode 1 (except for N = 2 pilot bits as will be further explained).

It may be noted that once every slot, the UE computes the phase adjustment φ that should be applied at the base station BS in order to maximize the UE received power. The UE feeds back to the base station BS the information on which phase settings to use by means of the FSM via uplink. Since the FSM can be distorted in uplink because of channel perturbations, the recommended weights can be distorted. Hence the recommended weights and the weights used at the base station BS can be different, the base station adjusting said recommended weights. Therefore, at the UE receiver, one estimate the weights really used at the base station BS, said estimates being used to recover later on the signal received from both antennas ANT1 and ANT2. The process of estimating the weights used by the base station BS and estimated at the UE is called antemia verification.

The antenna verification takes place in the receiver of the mobile UE after despreading of both the CPICH and the DPCH, more particularly at each finger of the receiver, one finger being associated to a path. A good antenna verification permits to recover correctly the user information symbols DATA sent from the base station BS, which are issued from a combination of the information (user information and pilot symbols) sent from the two antennas ANT1 and ANT2 of the base station BS.

Since in mode 1 the first weight wl is constant and equal to —_f= , the aim of the

V2 antenna verification for mode 1 is to estimate the second weight w2 on each slot of a finger.

It may also be noted that in mode 1 the real and imaginary parts of the second weight w2 can take the value 0.5 or -0.5. The problem is then to estimate which of these values is to be attributed to the real and the imaginary part.

Note that at the base station BS, the second weight w2 is calculated using the following formula:

and n is the number of the current slot.

Regarding the above hypothesis, the antenna verification is performed as follows as illustrated in the Fig.3.

In a first step EST_1), a first estimate of the second antenna weight w2 is performed, using the orthogonality of the pilot symbols sent on the first and second antennas ANTl, ANT2 in the second dedicated physical sub-channel DPCCH. This first estimate is a raw one.

It comprises the sub-steps of:

(a) Multiplication of the pilot symbols of the pilot field of the sub-channel DPCCH received on each finger by the complex conjugate of the channel estimate of the second antenna ANT2,

(b) Averaging of the physical channel estimate for the second antenna ANT2 on one symbol pilot field of a slot,

(c) Demodulation of the pilots of the second anteima ANT2 received in one slot DPCCH; and

(d) Combination of the activated fingers.

It is first to be noted that the signal R(n) received at the UE during one slot #n pilot field can be expressed as:

R(n) = G_u (n) i^> («) w, + G_v(n) P₂(n) w₂ (n) + v(n) [1]

with: ^m G, ,. (n) and G₂ ,. (n) , the propagation chamiel estimates of the signal coming from the first and second antennas ANTl and ANT2 at symbol rate for slot n and finger i corresponding to a pilot field of a received slot. These estimates are made on the physical channel CPICH transmitted on the first and the second antennas ANTl, ANT2 respectively. Practically, they model the propagation chamiel.

• f; (n) and P₂( ) , the channel pilot symbols of the DPCCH slot received.

• w, (n) and w₂ (n) , the weights applied on the first and second antennas ANTl, ANT2 at the base station BS for a second channel's DPCH slot received. • v(n) , the noise AWGN remaining after despreading of data chips in symbols well known to those skilled in the art with an average equal to zero.

• n, a slot, a frame being composed of 15 slots.

The first sub-step a) is performed as follows. The aim of the multiplication by the complex conjugate of the channel estimate of the second antenna ANT2 is to suppress the variations due to this channel. To this end, one multiplies the expression [1] with the complex conjugate (G_{2 l} ) of the propagation channel estimate of the second antenna ANT2.

One obtains:

The second sub-step b) is performed as follows. Firstly, the realistic hypothesis is made.

The terms: (<?./ ( ) G₂» ) and (G₂ (») G (n) ) [3]. can be considered almost constant over the symbol pilot field of the second subchannel DPCCH slot received. Indeed, a channel estimate coefficient can be considered almost constant over the pilot field of a DPCCH slot received if it verifies the fact that the Fading Coherence time, well known by those skilled in the art, is longer than the DPCCH slot pilot field period.

Note that the fading is one type of perturbation due to the diffraction and refraction of the waves of the signals sent over the propagation channel, which may affect the propagation channel (another one being the noise AWGN which can be considered negligible after despreading, the despreading being equivalent to an averaging, and the noise having an average equal to zero) in such a way that the power of said signals can decrease considerably. The relation to be verified by the DPCCH slot pilot field period Tpilot with/_d the Doppler frequency is: at -3dB: _Tc =-^- > Tpilot [4]

2π/_rf

For a fading of 60 km/h, we have: (2.1_g9)*60 _{= U6M} ^ H ._{002 j =} c 300000*3600 2U "

Tpilot = ¹⁰⁰⁰° *N_max = 0.26*1024 = 266.64 y, 15*2560 with:

- N_max equal to maximum size of a pilot field in the DPCCH per slot, expressed in chips, - F_p, is the carrier frequency for the downlink communication in UMTS, which is about 2.1*10⁹ Hz, one frame of information comprising 15 slots and is of lOOOOμs, each slot comprising 2560 chips,

T_c is the time coherence of the Fading regarding the Jakes model, in particular it is the transmit time during which the Fading is considered constant, - c is the light celerity (300000km/h), and v the speed of the mobile also called Fading speed.

The relation [4] is verified.

For a fading of 500 km/h, we have:

Tpilot = ^100QQ *N_max = 0.26*1024 = 266.64_/s 15*2560

the relation [4] is almost verified. Thus until 500km/h, the hypothesis of the fading being constant is verified, at 500km/h, it is still a reasonable approximation and hypothesis.

Thus, it means that even if the fading is high, the coefficients G of the physical channel is almost constant, therefore, its estimate, which is the estimate of the propagation channel, is also almost constant. It is to be noted that there is always white noise (which means that the average equal to zero) associated to an estimate. Therefore, in order to avoid the effect of this noise, the best approximation of the estimate of the channel is to take the average said pilot field.

In conclusion on reception, the term G₂ (n) can be replaced by the term (G₂ ")_av , which can be evaluated by a linear averaging of G_{2 l}" ( ) over the pilots field of one slot of the sub-channel DPCCH.

Thanks to the above hypothesis, the expression [2] can be simplified as: [2] <?./ (n) R(n) « (G₂ )_avg R(n)

(G₂, ")_avg ^R(n) = r₅-ι

((G_2, " )_avg G,„ («)) , (n) w, + ((G₂ )_mg G_2,, («)) P₂ (n) w₂ (n) + (G₂/ )_avg v(n)

The third sub-step c) is performed as follows. The aim of the demodulation of the pilots of the second antenna ANT2 received in one slot DPCCH, is to supress the variations due to the first antenna ANTl and to make the variations of the second antenna ANT2 constant. To this end, the expression [5] is multiplied by the complex conjugate P₂ ^H (n) of the pilot symbols of the second antenna ANT2.

One obtains:

((G₂ )_avg G (n)) P₂ ^H (n) P_l (n) w_l t⁶J

+ ((G₂ )_mg G_2>, (»)) P₂ ^H («) P₂ («) w₂ («) + P₂ ^H (n) (G₂ )_mg v(n)

As the pilots of the two antennas are orthogonal, p₂ (_n)^H p_l(n) = 0 nd as P₂ (n)^H P₂(n) = N , with N being the number of pilot bits in a pilot field, one obtains:

(G_2, ")_mg P₂" (n) R(n) = ((G₂ )_mg G_2<l («)) N w₂ (n) + P₂ ^H ( ) (G₂ )_avg v(n) [7]

Note that the noise v(n) , which is the noise of the propagation channel, is considered negligeable in comparison to the other terms, as the AWGN in the channel has been summed during despreading and has an average equal to zero. In Practice, this noise represents the difference remaining between the model used for the propagation channel and reality.

The fourth sub-step d) is performed as follows. It is to be noted that the steps described in the foregoing are implemented for each finger of the receiver. For each finger i of a receiver and a slot n, such a receiver being also called RAKE receiver in a UMTS application, the estimate of the weight w2 of the second antenna ANT2 is performed. Therefore, the expression [5] is written as follows.

[5] P₂ ^H (n) G₂ (n) R_t (n) « (G₂ )_avg Pξ (n) R_t (ή)

Hence:

(G₂ )_avg P₂ ^H (n)R_t(n) = ((G₂ )_avg G_u („)) P₂ ^H (n) P_y (^ή) w, + (( )_avg G_2J («)) P₂" (n) P₂ (n) w₂ (n) + P₂ ^H (^ή) (G₂ )_avg v(n)

The estimate of w2 obtained on a fmger i is then as described in [7]:

P₂ £^Hin) (β ^■ ₂,ι ) ^; _aa_vv_gg R-_iι(vn) w₂(n ^ w₂ n ' N -\τ(t G n₂ ^H H)^ _avgG n₂ [8]

An average first estimate of the weight w2 for the signal received R(n) on the Rake receiver can be obtained as below. The average improves the estimate because it averages also the perturbations.

Note that in the example of the Rake receiver, such receiver treats one physical channel with a maximum of K multipaths corresponding to K fingers, with M activated fingers of the same chamiel, M being less than or equal to K. For example, K=6 fingers.

Thus, in a first non-limitative embodiment, the estimate is:

M W₂(«) = ∑ («)/ i=0

In a preferred non-limitative embodiment, the different estimates of w2 of each finger i are just combined, and the value of the result of the combination is shifted by approximatively the number of activated fingers (to simplify hardware implementation) as only the sign of the estimate is interesting (a hard-decision is taken as will be further described with respect to the sign). For example, if there is only one finger, no shift is made, if there are 2 fingers combined, right shift by 1, and if there are 3 fingers right shift by 1, and 5 fingers right shift by 2.

In another non-limitative embodiment, the different estimates of w2 of each fmger I are combined, and they are not averaged (to simplify hardware implementation) as only the sign of the estimate is interesting (a hard-decision is taken as will be further described with respect to the sign).

5

Thus in this case, the estimate is Σ^w2, («) z=0

In a second step HARD_DEC), a hard decision on the sign of this first estimate w2 of w2 is made.

One knows that theoretical value of w2 has a constant module of 0.5 and that its real R and its imaginary I parts differ only by their sign. Therefore, getting the sign of the real and imaginary parts of this first estimate w2 remains. The hard decision is made as follows. If the sign of the real part R of w2(n) is greater than or equal to 0, then w2(n) =+0.5+j*I, otherwise, w2(n) =-0.5+j*I.

Then, if the sign of the imaginary part I of w2(n) is greater than or equal to 0, then w2(n) =R+ ^:0.5, otherwise, w2(w) =R-j*0.5.

In a third step RΕLIA), an evaluation of the reliability of said first estimate w2(n) is performed, using a sign property noticed between the complex possible values of the second antenna weight w2 of successive slots, and especially of two immediate successive slots.

cos(^^"') + cos(^) , , sin(^^"') + sin(^)

² 2 2

Note that at a frame border, for the slot #0 of a next frame j, the slot #13 of the preceding frame j-1 is taken instead of the slot #14 because at the base station BS, the second weight w2 has been calculated using the phase from the slot #13 of the preceding frame.

The current slot #n and its immediately preceding slot #n-l are taken by way of example.

The evaluation is performed according to the odd or even slot number. Indeed, the phase adjustment ψ values depend on the odd or even slot number as described in the table below:

So the following table may be deduced for the second antenna ANT2 weight w2:

With R( w2(n) ) = (cos Φ(n - 1) +cos Φ(n) )/2 and I( w2(n) )=(sin Φ(n - 1) +sin Φ(n) )/2 for a current slot #n.

Note that in the following, the FBI of the received current slot #n is used, whenever it is sent to the preceding slot #n-l or #n-2.

In more detail, if one goes from a preceding odd slot to an even current slot, the following expression will be obtained:

FBI received for an odd preceding slot and a current even slot:

[a] 0->0 R(w2(«) )=(cos(π/2)+cos(0))/2=0.5 _; I( w2( ) )=(sin(π/2)+sin(0))/2= 0.5

[b] 0->\ -»R(w2(n) )=(cos(π/2)+cos(π))/2=-0.5 ; I(w2(n) )=(sin(π/2)+sin(π))/2= ft5

[c] 7->0 -»R(w2(rc) )=(cos(-π/2)+cos(0))/2=0.5 ; I(w2(ra) )=(sin(-π/2)+sin(0))/2= -0.5

[d] l->\ -^R(w2(n) )=(cos(-π/2)+cos(π))/2=-0.5 ;I(w2(rc) )=(sin(-π/2)+sin(π))/2= -0.5 And in the same manner, if one goes from a preceding even slot to an odd current slot, the following expression will be obtained:

FBI received for a even preceding slot and a current odd slot: [e] 0->0 R(w2(n) )=(cos(0)+cos(π/2))/2=0.5 ; I( w2(n) )=(sin(0)+sin(π/2))/2= 0.5 [fj 0->l -»R( w2(n) )=(cos(0)+cos(-π/2))/2=0.5 ; I( w2(n) )=(sin(0)+sin(-π/2))/2= -0.5 [g] i->0-»R( w2(n) )=(cos(π)+cos(π/2))/2=-0.5 ; Ϊ(w2(n) )=(sin(pi)+sin(π/2))/2= 0.5 [h] l->\ -»R( w2(n) )=(cos(π)+cos(-π/2))/2=-0.5 ; I( w2(n) )=(sin(pi)+sin(-π/2))/2= -0.5

One can remark that for an even current slot #n, the imaginary part I of the estimate w2(n) of the current slot #n is the same as the one of the estimate w2(n-\) of the preceding slot #n-l, and that for an odd current slot #n, the real part R of the estimate w2(n) of the current slot #n is the same as the one of the estimate w2(«-l) of the preceding slot #n-l.

Some examples are illustrated in the table below:

Thus, according to the first example of the table above, one may see that for a current even slot #n, the imaginary part I of the estimate w2(n) is equal to: [a] 0->0 I( w2(n) )=(sin(π/2)+sin(0))/2= 0.5 and is equal to the imaginary part of the estimate w2(n-ϊ) calculated for the preceding slot #n-l and which is equal to

[e] 0->0 -» I(w2(«) )=(sin(0)+sin(π/2))/2= 0.5. Note that the UE knows exactly if the current received by slot #n is odd or even.

Therefore, based on this remark, the reliability test is performed as follows.

For an even current slot n:

One compares the sign of the imaginary part I of the first estimate w2(n) of the second antenna ANT2 of the currently received slot n with the sign of the imaginary part I of the first estimate w2(n - 1) of the second antenna ANT2 of the precedent slot n-1. If these signs are equal, then the first estimate w2(n) is a good estimate and the imaginary part I of this first estimate w2(ή) of the current slot n is kept. If the signs are not equal, then the estimate is wrong w2(n) . In summary, if sign(I( w2(n) ))==sign(I( w2(n - 1) ), then w2(n) = R( w2(n) ) + j *I( w2(n) ) For an odd current slot n:

The sign of the real part R of the first estimate w2(n) of the second antenna ANT2 of the currently received slot n is compared with the sign of the real part R of the first estimate w2(n - 1) of the second antenna ANT2 of the precedent slot n-1. If these signs are equal, then the first estimate is a good estimate and the real part R of this first estimate w2(n) of the current slot n is kept. If the signs are unequal, then the first estimate w2(n) is wrong. In summary, if sign(R ( w2(n) )=sign(R ( w2(n - 1) ), then w2(n) = R ( w2(n) ) + j *I ( w2(n) )

Note that of course, the estimate w2(n - 1) of the preceding slot n-1 is kept in memory for the verification described above.

In a third step EST_2), depending on the above evaluation, a correction of said first estimate w2( ) of the weight w2 is performed using the feedback information FBI of the FSM that had been sent in an uplink and applied to the preceding downlink slot of the sub-channel DPCCH. This step is illustrated in Fig. 4. If the first estimate w2(^ή) is wrong, the value of the weight w2 on the second antenna ANT2 of a preceding slot, n-1 for example, is used. The real part R or imaginary part I of said value w2 can be deduced from the FBI of the uplink FSM corresponding to the slot received according to an odd or even slot number, as already described in the table above.

This evaluation permits to improve the estimate w2(n) and is performed as follows.

For an even current slot n:

If the sign of the imaginary part I of the first estimate w2(n) of the second antenna ANT2 of the currently received slot n and the sign of the imaginary part I of the first estimate w2(n - 1) of the second antenna ANT2 of the precedent slot n-1 are unequal, the imaginary part I (w2_UE_tx (n-1)) of the weight of the second antenna ANT2 that has been specified in the FBI via the uplink FSM during the preceding slot n-1 will be taken.

In summary, if sign (I ( w2(n) ))≠sign (I ( w2(n - 1) ), then w2(n) = R ( w2(n) )+ j*sign (I (w2_UE_tx (n-l)))/2; For an odd current slot n:

If the sign of the real part R of the first estimate w2(n) of the second antenna ANT2 of the currently received slot n and the sign of the real part R of the first estimate w2(n - 1) of the second antenna ANT2 of the precedent slot n-1 are unequal, then the real part R (w2__UE_tx (n-1)) of the weight of the second antenna ANT2 that has been specified in the FBI via the uplinlc FSM during the preceding slot n- 1 will be taken. In summary, if sign(R ( w2(n) )≠sign(R ( w2(n - 1) ), then w2(n) = sign(R (w2_UE_tx (n-l)))/2+j*I (w2(n) )

Note that in another embodiment the base station BS can also apply the weight w2_UE_tx (n-2) of the second antenna ANT2 specified at the preceding slot (n-2) to the received slot n. Note that of course, the estimate of the preceding slot is kept in memory for the verification described above.

Note that w2J E_tx (n-1) or w2_UE_tx (n-2) is the value of the FBI. Said value is coded as follows: 1 for the value 0 and-1 for the value 1. This coding is called antipodal.

Note that what is interesting is that the weights really used at BS are estimated even if they are different due to uplink errors from the ones that had been advised by the UE via FSM in an uplink. The tests made do not reveal this type of uplink errors, they only permit to find errors in the estimation of the weight really used at BS. If the test made on the first estimate at the UE is false, the FBI will be used to correct said first estimate of w2. This choice is confirmed by simulations and can be explained by the fact that 4 per cent of errors in an uplink in normal operation may be considered a maximum, which percentage is specified in the 3GPP specifications 25.214 annex A. Note that the change of value is preferably made either on the real part R of the estimate w2(n) , or on its imaginary part I, depending on the even or odd slot, but not on the two together.

The first interest of not changing the two parts is to economize hardware resources. Moreover, simulations confirm no big differences between the two solutions (change the real or imaginary part, or to change the two). This result can be explained by statistic and trigonometric considerations.

In the case of a downlink, if there are perturbations, the channel estimation (G₂ ^H) should correct the major ones. The smaller phase errors, that should impact the two angles used in real and imaginary parts remain. For example, if we take the case of a current even slot #n, with an FBI(#n)=0 and FBI(#n-l)=0, if there are no errors, one should have: 0->0 -*R(w2(n) )=(cos(π/2)+cos(0))/2=0.5 and I( w2(n) )=(sin(π/2)+sin(0))/2= 0.5

A phase error means that: l(w2(r ) )=(sin(π/2+θ)+sin(0+θ))/2= -0.5, and R(w2(n) )=(cos(π/2+θ)+cos(0+θ))/2, θ being the phase error. In the case of an even current slot, only the imaginary part I counts and is tested. Therefore, if there is a phase error, it means that angle φ is in quadrant 3 or 4 in the graph of Fig.5.

If the angle φ is in quadrant 3, R(w2(n) )=0.5, there is no change from what was expected, and so no need to change the real part, only the imaginary part of w2 needs to be changed. This case is supported by what we do (changing just imaginary part or real part in case of current odd slot).

If the angle φ is in quadrant 4, R(w2(n) )= -0.5, there is a big phase error, and thus a need to change the imaginary and the real parts of w2. To support this case, real and imaginary parts of w2 would have to be changed.

In practise, we have more chance of being in the first case (quadrant 3), as channel estimation normally should correct the major perturbations coming from a big angle θ as observed before.

Of course, one can change both the real and imaginary parts, if needed or if one wants to.

In conclusion, the aim is reached because the estimate w2(n) has been affined accordingly and the antenna verification is performed.

The following step), which is a recovering of the user information data of the first sub- physical channel DPDCH, is performed, using the weight w2 thus estimated. This kind of recovering is well known to those skilled in the art and will not be described here.

It is to be noted that there is a special case N = 2 pilot bits. Indeed, for this case, the pilot bits on the two antennas ANTl, ANT2 are not orthogonal in closed loop model as described in the standard. However, as the channel estimates Gl, G2 for the first and second antennas ANTl and ANT2, the pilot patterns used and the constant first weight wl for the first antenna ANTl are known or can be easily computed, the hypothesis [3] is still verified. Therefore, the term G, (n) can be replaced by an approximation (G, )_avg , which is evaluated by a linear averaging of G_x(n) over the pilot field of one slot DPCCH. So, if one takes the formula [6], .^H) P )Λ(B) = 2, i avg 2

( _{η i} ^H) G. . (n))P^H(n)P(n)w + 2,ι avg l,ι 2 1 1 l, i avg l,ι I 2 2 2 l,ι avg the disturbing term ((G₂ i^H)avg^G t n))P^H() P(n) can be subtracted and one obtains:

([P₂ ^H(n)]*R(n)-(P₂ ^H(n) (<?,)„, P,(«) w,)) * (G₂" )_avg = ((G")_ms G₂(n))P₂ ^H(n)P₂(n)w₂(n) + P₂"(n)v(n)

Finally, we come back to the previous expression [8] and get an estimate of w2.

Hence, the receiver according to the invention has the advantage of resolving the problems of antenna verification by using a method not too complex (no phase to compute), which gives an affined estimate of the antenna weight, thus improving said antenna verification. One obtains directly a rough estimate of w2, without a need to estimate the phases φ. And after this, a simple test is applied in order to improve the estimate. Thus, fewer steps in the antenna verification of the invention are needed than the one of the prior art, and less hardware is also needed.

It is to be understood that the present invention is not limited to the aforementioned embodiments and variations and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims. In this respect, the following closing remarks are made.

It is to be understood that the present invention can also be applied to other receivers than the Rake receiver where antenna verification is needed.

It is to be understood that the method according to the present invention is not limited to the aforementioned implementation. There are numerous ways of implementing functions of the method according to the invention by means of items of hardware or software, or both, provided that a single item of hardware or software can carry out several functions. It does not exclude that an assembly of items of hardware or software or both carry out a function, thus forming a single function without modifying the method for processing antenna verification in accordance with the invention.

Said hardware or software items can be implemented in several manners, such as by means of wired electronic circuits or by means of an integrated circuit that is suitably programmed.

The integrated circuit can be contained in a computer or in a receiver. In the latter case, the receiver comprises first means for performing directly a first estimate of the second antenna's weight using the orthogonality of the pilot symbols sent on the first and second antennas and using chamiel estimates made on CPICH, second means for making a hard decision on the sign of said first estimate, third means for evaluating the reliability of said first estimate using a sign property between the complex possible values of said second antenna's weight of successive slots, fourth means for performing a correction of said first estimate of said second antenna's weight using feedback information that has been sent in an uplink and applied to a preceding downlink slot of said channel, depending on the above evaluation, as described previously, said means being hardware or software items as stated above.

The integrated circuit comprises a set of instructions. Thus, said set of instructions contained, for example, in a computer programming memory or in a receiver memory may cause the computer or the receiver to carry out the different steps of the antenna verification method.

The set of instructions may be loaded into the programming memory by reading a data carrier such as, for example, a disk. A service provider can also make the set of instructions available via a communication network such as, for example, the Internet.

Any reference sign in the following claims should not be construed as limiting the claim. It will be obvious that the use of the verb "to comprise" and its conjugations do not exclude the presence of any other steps or elements besides those defined in any claim. The article "a" or "an" preceding an element or step does not exclude the presence of a plurality of such elements or steps.

Claims

1. Method for processing antenna verification by means of a first channel (CPICH) and a second channel (DPCH) of a receiver, said second channel (DPCH) comprising pilot symbols, said receiver being associated to a base station (BS) comprising a first and second antennas (ANTl, ANT2) with associated weight (wl, w2), said pilot symbols being sent during time slots (#n) in said second channel (DPCH), said method comprising the steps of:

- performing directly a first estimate ( w2(ή) ) of the second antenna's weight (w2) using the orthogonality of the pilot symbols sent on the first and second antennas (ANTl, ANT2) in the second channel (DPCH) and using channel estimates ( i , G₂) of said first channel (CPICH),

- making a hard decision on the sign of said first estimate ( w2(n) ),

- evaluating the reliability of said first estimate ( w2(n) ) using a sign property between the complex possible values of said second antenna's weight (w2) of successive slots (#n-l, #n), - depending on the above evaluation, correcting said first estimate ( w2(n) ) using feedback information (FBI) that has been sent in an uplink and applied to a preceding downlink slot of said second channel (DPCH).

2. Method for processing antenna verification as claimed in claim 1, wherein one channel estimate ( Gi , G₂ ) of the first channel (CPICH) is evaluated by a linear averaging over the pilot field of one slot of the second (DPCH).

3. Method for processing antenna verification as claimed in claim 1, wherein the hard decision comprises the sub-steps of: - testing if the sign of the real part (R) of the firs estimate ( w2(n) ) is greater than or equal to 0, and applying a first value if the test is positive, otherwise applying a second value if the test is negative, and then,

- testing if the sign of the imaginary part (I) of the firs estimate ( w2(n) ) is greater than or equal to 0, and applying a first value if the test is positive, otherwise applying a second value if the test is negative.

4. Method for processing antenna verification as claimed in claim 1, wherein the sign property is as follows: - for an even current slot (#n), the imaginary part (I) of the estimate ( w2(n) ) of said current slot (#n) is the same as the one of the estimate ( w2(»-l) ) of a chosen preceding slot (#n-l), and

- for an odd current slot (#n), the real part (R) of the estimate ( w2(n) ) of said current slot (#n) is the same as the one of the estimate (w2(n-\) ) of a chosen preceding slot (#n-l).

5. Method for processing antenna verification as claimed in claim 1, wherein the reliability evaluation is verified if the sign property is verified.

6. Method for processing an antenna verification as claimed in claim 1, wherein the correction of said first estimate (w2(n) )is performed if the reliability evaluation is verified and comprises the sub-steps of:

- for an even current slot received of the second channel (DPCH), if the imaginary parts (I) of the weight's estimate ( w2(n) ) of a current slot (#n) and of a preceding slot (#n-l) are different, the imaginary part (I) of said current weight's estimate

( w2(n) ) is replaced by the one (I(w2_UE_tx (n-1))) deduced from the feedback information (FBI) uplink specified for said preceding slot (#n-l), and

- for an odd current slot received of the second channel (DPCH), if the real parts (R) of the weight's estimate ( w2(n) ) of said current slot (#n) and of said preceding slot (#n-l) are different, the real part (R) of said current weight's estimate ( w2(n) ) is replaced by the one (R(w2_UE_tx (n-1))) deduced from the feedback information (FBI) uplink specified for said preceding slot (#n-l).

7. Receiver for processing antenna verification by means of a first channel (CPICH) and of a second channel (DPCH), said second channel comprising pilot symbols, said pilot symbols being sent during time slots (#n) in said second channel (DPCH), said receiver comprising a first and second antennas (ANTl, ANT2) with associated weights (wl, w2) and: first means for performing directly a first estimate ( w2(n) ) of the second antenna's weight (w2) using the orthogonality of the pilot symbols sent on the first and second antennas (ANTl, ANT2) in the second channel (DPCH) and using channel estimates (G_{ , G2) of the first channel (CPICH),

- second means for making a hard decision on the sign of said first estimate ( w2(n) ),

- third means for evaluating the reliability of said first estimate ( w2(n) ) using a sign property between the complex possible values of said second antenna's weight (w2) of successive slots (#n-l, #n),

- fourth means for performing a correction of said first estimate ( w2(n) ) of said second antenna's weight (w2) using a feedback information (FBI) that has been sent in an uplink and applied to a preceding downlink slot of said second channel (DPCH), depending on the above evaluation.

8. Mobile phone comprising the receiver as claimed in claim 7.