GB2344720A - Receiving methods and apparatus using digital beam steering - Google Patents

Abstract

Receiving apparatus, for use for example in a base station of a mobile communications network, carries out a series of cycles and includes a processing unit (8, 10<SB>1</SB>, 10<SB>2</SB>, 12, 14, 16, 18) and a determining unit (20). In each cycle the processing unit processes receive signals (Y<SB>1</SB> to Y<SB>4</SB>), derived from different locations in space (2<SB>1</SB> to 2<SB>4</SB>), in accordance with three different trial beam patterns. The processing unit produces for each trial beam pattern a measure (SQ<SB>1</SB> to SQ<SB>3</SB>) of the quality of a wanted signal included in the receive signal. The determining unit (20) then, in each cycle, compares the respective quality measures produced for the trial beam patterns and employs the comparison results to determine the trial beam patterns for use in the next cycle of the series. Such a receiving apparatus implements a digital beam steering technique capable of tracking a wanted signal, whilst eliminating one or more of the problems of slow convergence, instability, significantly increased hardware complexity, and the requirement for a very high digital signal processor (DSP) speed found in previously-considered receiving apparatus.

Description

RECEIVING METHODS AND APPARATS USING DIGITAL BEAM STEERING The present invention relates to receiving methods and apparatus using digital beam steering. In particular, but not exclusively, the present invention relates to receiving methods and apparatus for use in an adaptive array antenna of a base station of a cellular mobile communication system.

In a cellular mobile communications system, it is necessary for the base station to detect the signal of each wanted user (i. e. each active mobile station) in a multi-user and multi-path environment. In order to achieve satisfactory signal detection at low bit error rates, two conditions must be satisfied. Firstly, the power level of the signal received by the base station from the mobile station must be greater than a certain threshold value. Secondly, the multi-user interference (MUI), sometimes referred to also as multiple access interference (MAI), must be reduced to an acceptable level.

To satisfy the two conditions identified above, it is effective to use adaptive beamformers in general and digital beamformers in particular. The principle underlying a digital beamformer is to form a spatial beam pattern in such a way that the angles of arrival of wanted signals fall well within a main lobe of the beam pattern whereas the interfering signals are located as much as possible in the nulls, low side lobes or boundary regions of the main lobe.

Figure 1 of the accompanying drawings shows parts of a previously-considered digital beamformer. Such a beamformer is provided, for example, in receiving apparatus at a base station of a mobile telecommunications network. At the base station, a plurality of independent sensors (an array of antenna elements 21 to 24) are provided to detect, at different points in space, a transmission signal sent to the base station by a mobile unit (not shown in Figure 1). The antenna elements 21 to 24 permit sampling of the received signal in space. The respective receive signals produced by the antenna elements 21 to 24 are filtered by matched filters and then applied to a digital beamformer 6 which is employed as a spatial filter.

The digital beamformer 6 includes a set of complex-conjugate multipliers 8 connected respectively for receiving the different receive signals. Each complex multiplier multiplies the receive signal applied thereto by a weight value set by a weight setting unit 12 of the beamformer 6. The resulting outputs of the multipliers 8 are then combined by a combiner 10 to produce an output signal y (t) of the digital beamformer. The object of the spatial filtering carried out by the digital beamformer 6 is to optimise the beamformer response with respect to some prescribed criterion so that noise and interference are minimised in the output signal y (t).

Assuming that the mobile unit of interest is at a large distance from the base station, the respective nominal paths between the mobile unit, on the one hand, and the different antenna elements 2 :, to 24, on the other hand, along which the transmission signal generated by the mobile unit propagates will be fractionally different in length, so that there is a phase difference 6, corresponding to the difference in path length, in the receive signals produced by adjacent antenna elements, for example the elements 23 and 24. The path-length differences, and hence the phase differences 6, depend upon the angle of arrival 9 of the transmission signal and the spacing t between adjacent antenna elements, as shown in Figure 1 (6 e sin i).

Sources of interference, on the other hand, such as other mobile units operating in the same area, will generally have a different angle of arrival from the angle of arrival 6 of the wanted signal produced by the mobile unit of interest. In this case, it is possible to set the weights W, to W4 applied to the complex multipliers in such a way that the wanted signal, having the angle of arrival B, is received satisfactorily whilst the interfering signals, not having the angle of arrival 6, are filtered out.

The output signal y (t) of the beamformer is compared with a reference signal d (t) using a subtraction element 16, and the resulting difference signal e (t), representing the difference between the actual output signal of the beamformer and the reference signal, is applied to the weight setting unit 12 which uses that error signal to adjust the weights W to W4 applied to the multipliers 8.

In a steady-state condition, in which the wanted and interfering signals each have a fixed angle of arrival at the receiving apparatus, there will be a fixed optimum set of beamformer weight values Wi to W4 which satisfies the prescribed criterion for minimising noise and interference at the output of the beamformer.

An adaptive algorithm is employed in the weight setting unit 12 which, in the above steady-state condition, would cause the weight values to converge to their optimum steady-state values and, thereafter, the noise and interference at the output of the beamformer (or, equivalently, the difference signal e (t) representing the difference between the desired output signal d (t) and the actual output signal y (t) of the beamformer) would remain at a minimum level related to the number of weights. However, in the real world, multipath propagation means that the transmission channel between the subject mobile unit and the base station is timevariant and, furthermore, the positions of the interfering signal sources (for example other mobile units) will change, with respect to one another and the base station, over time. Accordingly, the weight setting unit 12 is required to update the beamformer weight values continuously in accordance with the changing operating parameters.

Further information about digital beamforming techniques and related adaptive algorithms can be found, for example, in"Digital Beamforming in Wireless Communications", John Litra & Titus Kwok-Yeung Lo, Artech House Publishers, 1996, ISBN: 0-89006-712-0, the content of which is incorporated herein by reference.

One previously-considered beamformer uses an adaptation algorithm which is based on a least mean square (LMS)-type algorithm. However, with such an algorithm, the convergence speed and the stability depend on the step size and the system parameters.

Also, convergence to the optimum is not always guaranteed.

Another previously-considered beamformer uses an adaptation algorithm which is non-iterative. However, the use of a non-iterative method requires knowledge of the data and the angles of arrival of the multi-path signals. This demands both a significant increase in hardware complexity and very high digital signal processor (DSP) speed.

According to a first aspect of the present invention, there is provided a receiving method, for use in a mobile communications network, in which a series of cycles is carried out, each said cycle comprising the steps of: processing receive signals, derived from different locations in space, in accordance with at least two different trial beam patterns to produce for each trial beam pattern a measure of the quality of a wanted signal included in the said receive signals; and comparing the respective quality measures produced for the trial beam patterns and employing the comparison results to determine the trial beam patterns for use in the next cycle of the series.

According to a second aspect of the present invention, there is provided receiving apparatus, for use in a mobile communications network, operative to carry out a series of cycles and including: processing means operable in each said cycle to process receive signals, derived from different locations in space, in accordance with at least two different trial beam patterns to produce for each trial beam pattern a measure of the quality of a wanted signal included in the said receive signal; and determining means operable in each said cycle to compare the respective quality measures produced for the trial beam patterns and to employ the comparison results to determine the trial beam patterns for use in the next cycle of the said series.

According to a third aspect of the present invention, there is provided a base station for use in a mobile communications network, including receiving apparatus embodying the aforesaid second aspect of the present invention.

Such a receiving method and apparatus can provide a digital beam steering technique capable of tracking a wanted signal, whilst eliminating one or more of the problems of slow convergence, instability, significantly increased hardware complexity, and the requirement for a very high DSP speed.

Such a receiving method and apparatus can also provide fast convergence if the intended mobile station is in the initial beam. If the intended mobile station is not in the initial beam an initial scanning operation may be performed in which a scanning beam having a relatively large angular spread (e. g. 10 ) is moved in steps across the sector of operation of the receiving apparatus.

Reference will now be made, by way of example, to the accompanying drawings, in which: Figure 1, discussed hereinbefore, is a block diagram showing parts of a previously-considered digital beamformer; Figure 2 is a schematic diagram for use in explaining a principle underlying a receiving method embodying the present invention; Figure 3 shows parts of receiving apparatus according to a first embodiment of the present invention ; Figure 4 shows parts of receiving apparatus according to a second embodiment of the present invention ; Figure 5 shows parts of receiving apparatus according to a third embodiment of the present invention; and Figure 6 shows parts of receiving apparatus according to a fourth embodiment of the present invention.

A receiving method embodying the present invention involves a series of cycles, and in each cycle a plurality of different beam patterns are used to process a set of receive signals produced respectively by different antenna elements.

For example, as shown in Figure 2, assume that in the current cycle n three beam patterns are produced: a middle beam pattern B having an angular direction (look-direction) e. ; a lower beam pattern BW having an angular direction e that is less than the middle-beampattern direction 80 by a first predetermined amount A6l ; and an upper beam pattern BU, having an angular direction #+ that is greater than the middle-beampattern direction 6o by a second predetermined amount ti2&commat; The first and second predetermined amounts may have the same value Ad, in which case 9 = 00-Ad and e o + Ai For each beam pattern Bt, BLn and BUn a measure of the quality of the particular wanted signal is produced and then, based on a comparison of the quality measures, a decision is made as to the beam patterns to be used in the next cycle of the series. The underlying aim is to update the beam patterns (or lookdirection) of the antenna in such a way that the quality of the wanted signal is improved in every adaptation cycle, which consequently leads to an improved bit error ratio (BER).

For example, in a simple decision-making case, in each cycle one of three options may be taken for steering the beam: 1) Increase the angle of the look-direction from Oc to 0+ where #+ = #0 + ##.

2) Decrease the angle of the look-direction from Go to 6~ where 6~-80-06.

3) Leave the look-direction unchanged, i. e equal to 60.

As shown in Figure 2, if the upper beam pattern BU, pointed at 6, = 60 + Ae in cycle n has the best quality measure, that beam pattern will be used as the middle beam pattern B, l in the next adaptation cycle n+1 (the new look-direction 80'= o 0 + A6). The middle beam pattern B pointed at 80 in cycle n will be used in cycle n+1 as the lower beam pattern sLn+l (#@' = #0) and the remaining beam pattern IBL, pointed at e = eO-AE in cycle n will be repointed so as to form the upper beam pattern BUn+1 (#+' = #@ + 2##) in cycle n+1. Thus, in this example, although conceptually the angular direction of each beam pattern is increased by A&commat;, in practice two out of the three beam patterns used in cycle n are reused in cycle n+1, and only one beam pattern is changed. Since in this version of the method two of the three beam patterns are preserved in every adaptation cycle, stability is improved.

Once the beam patterns for the next cycle have been decided the next cycle begins.

Next, embodiments of receiving apparatus capable of implementing a method as described above will be described with reference to Figures 3 to 6.

In order to facilitate a simple implementation of the above method, the incremental angle change (A8) per cycle is preferably made small. Preferably, D0 satisfies the following equation: ## < < (M-1)) where M is the number of antenna elements in the array.

Typically, M is less than 8 in practice. In a preferred embodiment M is 4 and Ae is about 2 (0.035 radians).

Since the incremental angle change is small, it is possible to use a first-order Taylor series in the signal processing as explained below.

Further (optional) features permitting simplification of the apparatus are: (i) the antenna elements of the array are equally spaced; and (ii) the set of weights used in the beamformer have the same magnitude and linear phase distribution.

Assume that at the start of a cycle n+1 the set of weights (weight vector) applied to the M antenna elements of the array is given by

In the cycle n+1 the weight vector can be changed to =. where S is a steering change matrix

and h is an incremental phase shift leading to the steering of the look direction (middle-beam-pattern angular direction). The use of such a steering change matrix leaves the beam shape unchanged from one cycle to the next, which leads to greater stability. ## can be calculated using A # = ###(sin#n+1 - sin#n) where # is the antenna spacing, X is the operating wavelength and En and #n+1 are the present and new look directions of the array respectively (see Figure 1).

The output signal sa in cycle n is given by S. Yn (4) where =(y.....y is the signal at the array input in cycle n.

Similarly, the output signal Sn+1 in cycle n+1 is given by =V (6) Sn+1=Wn, lYn+1 where Y., is the array input signal in cycle n+1.

Substituting Eq. (2) into Eq. (6) gives = (7) Sn+1=WnT SYn+1 Assuming the phase increment tf is small the steering charge matrix S can be replaced by a first-order Taylor series approximation, which gives S=I+j##M (8) where

Then one has Sn+1 = Sn+1(1) + j##Sn+1(2) (10) where Sn+1(1) = WnT Yn+1 (11) and Sn+1(2) = WnT # Yn+1 (12) Similarly, nl 1 = Wn + j###Wn (13) Assuming the absolute value of ## is ##o, the following beam steering algorithm can be used.

1) Using two beamforming circuits to produce (1) and (2) Sn+l e 2) Generating an upper-beam-pattern output signal

and a lower-beam-pattern output signal

and a middle-beam-pattern output signal

Sn+1 = Sn+1(1) (14.c) 3) Estimating the quality of the above output signals. Examples of possible techniques include signal-to-interference-and-noise ratio (SINR) estimation and power level estimation. For the latter, one has

where the brackets denote averaging. A typical averaging window is 1 or 2 timeslots.

4) Making a decision on A : In one embodiment, the decision-making strategy is:

O if Pn+, otherwise A (P =-A90'f Pn+l""Pn+l (16) & (PO if Pn+l > Pn+l If desired the"larger than"and"smaller than" comparisons can be modified to"larger than by at least a threshold value"and"smaller than by at least a threshold value". In this case, Eq. (16) is modified to the following:

-pn, 'E 17 F (Po if (Pn+l Pn+l) > e s Aq) O if (Pn+l-Pn+l) > e (17) 0 otherwise where e can be, say, 0.25dB.

Other decision making strategies, for example interpolation techniques, can also be used. One possible technique is:

Else

Else

End End 5) Updating the weight vector Wn to Wn+1 using Eq. (13) and go to step 1.

Figure 3 shows parts of receiving apparatus according to a first embodiment of the present invention. Components of the Figure 3 apparatus which correspond to components already described with reference to Figure 1 will be denoted by the same reference numerals as in Figure 1.

The Figure 3 apparatus comprises an array of antenna elements 2 ; tao 24 connected via respective downconverters (not shown) to a first set of complexconjugate multipliers 8. l to 8, 4. The multipliers 811 to 814 each receive a corresponding receive signal Y1 to Y4 derived from one of the antenna elements.

Each multiplier of the first set is also connected to a corresponding weight setting unit 12l to 124 for receiving one of a set of weight values W, to W4 therefrom.

Respective multiplier signals Y1W1, Y2W2, Y3W3 and Y4W4 produced at the outputs of the first-set multipliers 811 to 814 are applied to a first combiner 101 which provides an output signal S1.

The Figure 3 apparatus further comprises a second set of complex-conjugate multipliers 821 and 822. The multiplier 821 has an input connected to the output of the multiplier 813 in the first set for receiving the multiplier signal Y3W3. The multiplier 822 has an input connected to the output of the multiplier 814 for receiving therefrom the multiplier signal Y4W4.

Each multiplier of the second set also has an input connected to a corresponding constant value supplying unit 14l or 142. The unit 14., applies a constant value 2 to the multiplier 821, whereas the unit 142 supplies the constant value 3 to the multiplier 822.

The Figure 3 receiving apparatus further comprises a second combiner 102 which has a first input connected to the output of the multiplier 812 for receiving the output signal Y2W2 thereof, a second input connected to the output of the multiplier 821 for receiving an output signal 2Y3W3 thereof, and a third input connected to the output of the multiplier 822 for receiving an output signal 3Y4W4 thereof. The second combiner 102 has an output at which a second output signal S2 is produced.

The Figure 3 receiving apparatus further comprises a multiplier 831 having an input connected to the output of the second combiner 102 for receiving the second output signal S2, and a further input connected to a further constant value supplying unit 143 which supplies a constant value jaoo to the multiplier 83l.

A first adder 16l has a first input connected to the output of the first combiner 10l for receiving the output signal S1, and a second input connected to the output of the multiplier 831 for receiving an output signal jA 0S2 produced thereby.

A second adder 162 has a first input connected to the output of the first combiner 10l for receiving the output signal S1, and a second input, which is a negative input, connected to the output of the multiplier 831 for receiving the output signal jAOcpoS2 thereof.

The apparatus further comprises respective first, second and third signal quality indicator (SQI) units 181 to 183. The first SQI unit 181 has an input connected for receiving the output signal S1, the second SQI unit 182 has an input connected to the output of the second adder 162 for receiving a difference signal S-thereof, and the third SQI unit 183 has an input connected to the output of the first adder 161 for receiving a summation signal S+ thereof.

A decision unit 20 has three inputs connected respectively to outputs of the SQI units 181 to 183 for receiving respective signal quality measures SQI to SQ3 therefrom. An output signal produced by the decision unit 20 is applied to each of the weight setting units except for the weight setting unit 121.

A detector 22 is coupled to the output of the first combiner 101 for receiving the output signal S1.

Operation of the Figure 3 circuitry will now be described. In use, the first-set multipliers 8ll to 814, the first combiner 101 and the weight setting units 12 to 124 together serve as a first beamformer. The respective receive signals Y1 to Y4 produced by the antenna elements 21 to 24 of the array are multiplied by the relevant weight values W1 to W4 supplied by the weight setting units 121 to 124 to produce the multiplier signals Y1W1, Y2W2, Y3W3, Y4W4 applied respectively to the first-combiner inputs. These multiplier signals are combined by the combiner 101 to produce the output signal S1, i. e.

S1 = Y1W1 + Y2W2 + Y3W3 + Y4W4.

The first-set multipliers 812 to 814, the weight setting units 122 to 124, the second-set multipliers 821 and 822 and the second combiner 102 together serve as a second beamformer. The second-combiner inputs are respectively the multiplier signals Y2W2,2Y3W3 and 3Y4W4. Thus, the output signal S2 of the second beamformer is given by S2 = Y2W2 + 2Y3W3 + 3Y4W4.

It will be appreciated that the output signals S1 and S2 correspond respectively to the terms (1) in equation 11 and (2) in equation 12.

The output signal S2 of the second beamformer is multiplied by jA < 0 in the multiplier 831 and the resulting multiplied signal jA0S2 is both added (in the first adder 16J to, and subtracted (in the second adder 162) from, the output signal S1 of the first beamformer.

The resulting summation signal S+ produced by the first adder 161 corresponds to the upper-beam-pattern output signal S ++1 of equation 14a above. The resulting difference signal S-produced by the second adder 162 corresponds to the lower-beam-pattern output signal S +l of equation 14b above. The signal Si itself corresponds to the middle-beam-pattern output signal Sn+l of equation 14c above.

The SQI units 181 to 183 produce respective signal quality measures SQI. to SQ3 corresponding respectively to the signals S1, S-and S+.

For example, each SQI unit may produce a measure of the average power level of its corresponding signal S1, S-and S+ over a predetermined averaging window (e. g. one or two timeslots), as indicated in equations 15a to 15c above.

The decision unit 20 compares the signal quality measures SQ1 to SQ3 and decides whether to make the steering-change parameter A 0,-Af0 or +h 0, as indicated in equation 16 or 17 above.

The determined steering-change parameter A is applied to the weight setting units 122 to 124 which update their respective weight values W2 to W4 in accordance with equation 13 above. Thus, the weight value W2 is increased by jW2, the weight value W3 is increased by 2jA < W3, and the weight value W4 is increased by 3jAçW4. The weight value W1 is constant and is not changed in dependence A < .

The detector 22, which processes the output signal S1 of the first beamformer, detects the wanted signal based on the current look-direction of the antenna array which corresponds to the middle-beam pattern.

Instead of using the decision making strategy of equation 16 or 17 it is also possible to base the decision on just the output signals for the lower and upper beam patterns. Since the middle-beam-pattern output signal has no effect on this decision, the SQI unit 18, is not required and may be removed.

For example,

Alternatively, the following decision making strategy may be used:

< Po (i-) if t. f Pn. l-Pn+i)'E 19 0 otherwise Figure 4 shows parts of a receiving apparatus according to a second embodiment of the present invention. This embodiment is suitable for use in a code-division-multiple-access (CDMA) cellular mobile communications system, and uses a"finger beamforming" configuration.

In mobile communications environments, reflectors are inevitably present which lead to multi-path propagation of the transmission signal (wanted signal) from the mobile station to the base station. In a receiver which uses a"finger configuration", instead of using one circuit section to detect the wanted signal, a plurality of"fingers" (circuit sections) are provided which are intended to enable the receiving apparatus to lock on to plural best paths between the mobile station and the base station, i. e. the strongest versions of the transmission signal. One finger is used for each path to be processed, each producing an output signal for its particular path. The signals from all the individual fingers are then combined in a so-called Rake receiver. In such a receiver, for example, four fingers may be used.

Figure 4 shows the circuitry used in one of the fingers F in this embodiment. All the other fingers have the same circuitry.

As Figure 4 shows, each finger includes circuitry generally similar to that used in the first embodiment.

In particular, each finger has its own first and second beamformers, operable independently of the beamformers in the other fingers (a"finger beamforming configuration"). Thus, the three beam patterns formed in each cycle can be different for the different fingers, which means that the look-directions of the fingers can be different. Also, the decisions made in each cycle as to the look-direction for the next cycle can be different in each finger. This is useful since, if the beam patterns are narrow, the different paths of the wanted signal will probably not all be confined within a single beamformer.

In Figure 4, the circuitry in each finger further comprises a set of despreaders (matched filters) 32l to 324. Each despreader is connected between a corresponding one of the antenna elements 21 to 24 and its corresponding first-set multiplier 8ll to 8l4. Each despreader despreads the spectrum of the receive signal of its corresponding antenna element in accordance with a spreading code of the wanted signal. The dominant paths in the receive signals are identified by a path searcher (not shown), and each identified path is assigned to one of the fingers.

The Figure 4 embodiment also includes a Rake receiver 34 which receives the respective middle-beampattern output signals Si of all the fingers and combines them to provide a combined output signal. A detector 22 is connected to the output of the Rake receiver 34 for detecting the wanted signal based on the combined output signal.

It should be appreciated that another embodiment of the invention may employ what is called a"common beamformer configuration", as opposed to the finger beamformer configuration described above. In such an embodiment a Rake receiver is placed immediately before each SQI unit 181 to 183 and all the matched filters 32 to 324 between the antenna elements and first-set multipliers are removed. In this case, each finger includes a despreader with appropriate delay. The fingers all use the same set of beam patterns, and a path searcher identifies the dominant paths for each finger. An output signal per path is produced for each beam pattern. It is not necessary to despread the output signals S, S-and S+ for the beam patterns for signal quality measurement purposes, but the middlebeam output signals will require despreading for detection of the wanted signal.

Figure 5 shows parts of a receiving apparatus according to a third embodiment of the present invention.

This embodiment is a simplified version of the second embodiment shown in Figure 3. In this embodiment the decision made in the decision unit 20 of each finger F'is based exclusively on the output signals S-and S+ for the lower-and upper-beam patterns, as formed in the finger concerned. The decision making strategy may be as in equation 18 or 19 above.

Figure 6 shows parts of a receiving apparatus according to a fourth embodiment of the present invention. Components of the Figure 6 apparatus which correspond to components already described with reference to Figure 3 will be denoted by the same reference numerals as in Figure 3.

The Figure 6 apparatus comprises a set of weight value supplying units 40l to 404 which replace the weight setting units 121 to 124 of the Figure 3 apparatus. The apparatus also comprises timemultiplexing adders 362 to 364 and a fourth set of complex-conjugate multipliers 842 to 844.

Each weight value supplying unit 40l to 404 has an output at which its corresponding weight value W1 to W4 is produced, and the units 4 2 to 404 have respective updating inputs for receiving corresponding updated weight values W2'to W4'.

The output of the weight value supplying unit 40 is connected directly to the complex-conjugate multiplier 811. The output from each weight value supplying unit 4 2 to 404 is connected to a first input of a corresponding one of time-multiplexing adder 362 to 364.

Each time-multiplexing adder 362 to 364 has two outputs. The first of these outputs is applied to the corresponding weight value supplying unit 402 to 404 and the second is applied to a corresponding one of the first-set multipliers 812 to 8l4.

Each time-multiplexing adder 362 to 364 also has a second input which is connected to a corresponding one of the complex-conjugate multipliers 842 to 844 of the fourth set.

All the complex-conjugate multipliers 842 to 844 receive an output signal jazz of the decision unit 20 at a first input.

The second input to the complex-conjugate multiplier 842 receives an output signal M2 of the complex-conjugate multiplier 812. The second input to the complex-conjugate multiplier 843 is connected to receive an output signal 2M3 of the complex-conjugate multiplier 821. Similarly, the second input to the complex-conjugate multiplier 844 is connected to receive an output signal 3M4 of the complex-conjugate multiplier 822- Operation of the Figure 6 apparatus will now be described.

In use, signal reception and weight updating are carried out in a time-multiplexed fashion. Hence, each cycle comprises a signal reception phase and a weight updating phase.

During the signal reception phase, the first input of each time-multiplexing adder 362 to 364 is connected to the second output thereof. The second inputs of the adders 362 to 364 are deactivated (isolated).

Accordingly, the time-multiplexing adders 362 to 364 function so as to pass the weight values W2 to W4, received from the corresponding weight value supplying units 402 to 404, directly to the complex-conjugate multipliers 842 to 844. Hence, during the signal reception phase, the inputs to the complex-conjugate multipliers 81l to 814 correspond to those in the embodiment described with reference to Figure 2, and the multiplier signals M2 to M4 are respectively Y2W2, Y3W3 and Y4W4. Operation of the remainder of the circuit is the same as in the embodiment of Figure 2 and the steering change parameter tç is chosen as in that embodiment.

During the weight-updating phase, which commences when the steering-change parameter Ap has been chosen, the weight value supplying units 401 to 404, the timemultiplexing adders 362 to 364, the constant value supplying units 14l and 142 and the first, second and fourth sets of complex-conjugate multipliers 81, to 814, 821 to 822 and 842 to 844, cooperate to update the weight values, using the formula Wil [1 + j (i-l) A] Wi, where i = 2 to 4.

In particular, during the weight-updating phase, each time-multiplexing adder 362 to 364 connects its first input to its second output (as in the signal reception phase) and also adds together the signals received at its first and second inputs to produce the updated weight value W2'to W4'at its first output.

Also, each complex-conjugate multiplier 812 to 814 functions so as to make its multiplier signal M2 to M4 equal to the weight value W2 to W4 received from its corresponding time-multiplexing adder 362 to 364, irrespective of its receive-signal input Y2 to Y4 from its corresponding antenna element 22 to 24. The multiplier 821 multiplies the signal M3 by 2 and the resulting signal 2M3 is delivered to the multiplexer 843. The multiplier 822 multiplies the signal M4 by 3 and the resulting signal 3M4 is delivered by the multiplier 844.

The complex-conjugate multipliers 842 to 844 multiply the signals M2,2M3 and 3M4 by jazz p has the value chosen in the decision unit 20 during the preceding signal reception phase).

The output signal jås, oM2, jAso2M3 or jAf3M4 of each complex-conjugate multiplier 842 to 844 is then added to the corresponding current weight value W2, W3 or W4 in the corresponding time-multiplexing adder 362 to 364 to produce the updated weight value W2', W3'or W4'.

These updated weight values are then stored in the weight value supplying units 402 to 404 for use in the next cycle.

It will be appreciated that in the Figure 6 apparatus the second-set multipliers are used to carry out different operations in the two phases. This leads to hardware savings. Signal reception and weight updating can also be carried out in a time-multiplexed fashion in the CDMA embodiments of Figures 4 and 5.

It should be appreciated that another embodiment may implement, in the decision unit 20, a decision making strategy that has only 2 outcomes. In such an embodiment the angular direction of the middle beam is either increased or decreased in each cycle.

It should be appreciated that instead of using a constant magnitude as o of steering change parameter in all cycles (i. e. a constant angular-direction change) it is possible to reduce the angular-direction change in each cycle as the series of cycles progresses. For example, the steering change parameter can be halved from one cycle to the next. The angular offset between the beam patterns used in each cycle may in this case also be scaled from one cycle to the next. This can provide course initial steering followed by fine steering.

The present invention is not limited to using power as the measure of signal quality. Other possible measures of quality include signal-to interference ratio (SIR), signal-to-noise ratio (SNR), signal-to- interference-and-noise ratio (SINR), bit-error-ratio (BER) and signal strength (SS). It is also possible to use a combination of these. For example, initially a comparison of powers may be made but if this comparison is inconclusive a further comparison, of say SINR, can be made.

It should also be appreciated that Figures 3 to 6 are only schematic. In practice, a digital signal processor could be used to carry out all the functions.

It is also unnecessary for processing to be carried out in parallel for all the beam patterns. In practice, the receive signals could be buffered and then processed for the different beam patterns sequentially, so as to save hardware.

Claims (36)

1. CLAIMS 1. A receiving method, for use in a mobile communications network, in which a series of cycles is carried out, each said cycle comprising the steps of: processing receive signals, derived from different locations in space, in accordance with at least two different trial beam patterns to produce for each trial beam pattern a measure of the quality of a wanted signal included in the said receive signals; and comparing the respective quality measures produced for the trial beam patterns and employing the comparison results to determine the trial beam patterns for use in the next cycle of the series.
2. 2. A method as claimed in claim 1, wherein in each said cycle the receive signals are processed in accordance with a first such trial beam pattern, having an angular direction that is less than a trial angular direction determined for the cycle concerned by a first predetermined amount, and with a second such trial beam pattern having an angular direction that is greater than the said trial angular direction by a second predetermined amount.
3. 3. A method as claimed in claim 2, wherein the said first and second predetermined amounts are equal in any particular cycle of the said series.
4. 4. A method as claimed in claim 2 or 3, wherein the said first predetermined amounts are equal in all cycles of the said series, and the second predetermined amounts are equal in all cycles of the series.
5. 5. A method as claimed in any one of claims 2 to 4, wherein the trial angular direction for the next cycle is made greater than the trial angular direction for the present cycle by a predetermined upward adjustment amount, or is made lower than the trial angular direction for the present cycle by a predetermined downward adjustment amount, or is made equal to the trial angular direction for the present cycle, in dependence upon the said comparison results.
6. 6. A method as claimed in claim 5, wherein the said predetermined upward and downward adjustment amounts are equal in any particular cycle of the said series.
7. 7. A method as claimed in claim 5, wherein the said predetermined upward adjustment amounts are equal in all cycles of the said series, and the said predetermined downward adjustment amounts are equal in all cycles of the said series.
8. 8. A method as claimed in claim 5,6 or 7, wherein, for each said cycle of the said series, the said predetermined downward adjustment amount is equal to the said first predetermined amount, and the said predetermined upward adjustment amount is equal to the said second predetermined amount.
9. 9. A method as claimed in any one of claims 5 to 8, wherein the said trial angular direction for the next cycle is made greater than the present-cycle trial angular direction by the predetermined upward adjustment amount when the second-beam-pattern quality measure exceeds the first-beam-pattern quality measure by at least a predetermined first threshold value, and is made lower than the present-cycle angular direction by the predetermined downward adjustment amount when the first-beam-pattern quality measure exceeds the second-beam-pattern quality measure by at least a predetermined second threshold value, and is otherwise made equal to the present-cycle trial angular direction.
10. 10. A method as claimed in any one of claims 2 to 9, wherein in each said cycle the said receive signals are processed in accordance with a third such trial beam pattern having the said trial angular direction determined for the cycle concerned.
11. 11. A method as claimed in any one of claims 2 to 10, wherein the trial angular direction for the next cycle is made equal to the angular direction of that one of the trial beam patterns in the present cycle having the highest quality measure.
12. 12. A method as claimed in any one of claims 2 to 10, wherein the trial angular direction for the next cycle is derived by an interpolation method from the respective angular directions and quality measures of two or more of the trial beam patterns in the present cycle.
13. 13. A method as claimed in any one of claims 2 to 12, further comprising the step of detecting the wanted signal by processing the said receive signals in accordance with a beam pattern having the said trial angular direction.
14. 14. A method as claimed in any one of claims 2 to 13, wherein the said trial angular direction is defined by a set of weights.
15. 15. A method as claimed in claim 14, wherein the weights in the said set have a uniform amplitude distribution and a linear phase distribution.
16. 16. A method as claimed in any preceding claim, wherein the said trial beam patterns in the next cycle include at least one of the trial beam patterns used in the present cycle.
17. 17. A method as claimed in any preceding claim, wherein the said quality measure for each trial beam pattern includes a measure of wanted-signal power for the trial beam pattern concerned.
18. 18. A method as claimed in any preceding claim, wherein the said quality measure for each trial beam pattern includes a measure of wanted-signal signal-tonoise ratio or signal-to-interference ratio or signalto-interference-and-noise ratio for the trial beam pattern concerned.
19. 19. A method as claimed in any preceding claim, wherein each said quality measure is averaged over a predetermined averaging period.
20. 20. A method as claimed in any preceding claim, wherein the said processing of the receive signals and/or the said determination of the trial beam patterns for the next cycle involves the use of the first-order Taylor-series approximation ex9-1 + j6.
21. 21. A method as claimed in any preceding claim, wherein the said receive signals are derived from different respective antenna elements.
22. 22. A method as claimed in claim 21, wherein the antenna elements are equally spaced one from the next.
23. 23. A method as claimed in any preceding claim, wherein the said receive signals are time-divisionmultiple-access signals.
24. 24. A method as claimed in any one of claims 1 to 23, wherein the said receive signals are code-divisionmultiple-access signals.
25. 25. Receiving apparatus, for use in a mobile communications network, operative to carry out a series of cycles and including: processing means operable in each said cycle to process receive signals, derived from different locations in space, in accordance with at least two different trial beam patterns to produce for each trial beam pattern a measure of the quality of a wanted signal included in the said receive signal; and determining means operable in each said cycle to compare the respective quality measures produced for the trial beam patterns and to employ the comparison results to determine the trial beam patterns for use in the next cycle of the said series.
26. 26. Apparatus as claimed in claim 25, wherein in each said cycle the said processing means are operable to process the said receive signals in accordance with a first such trial beam pattern, having an angular direction that is less than a trial angular direction determined for the cycle concerned by a first predetermined amount, and with a second such trial beam pattern having an angular direction that is greater than the said trial angular direction by a second predetermined amount.
27. 27. Apparatus as claimed in claim 26, wherein the said processing means include: first beamformer means for producing a first beamformer signal S1 using the formula: SI = Wrt Y +1 where Wn represents a set of weights defining the said trial angular direction for the present cycle, and Yn+ represents the said receive signals to be processed in the present cycle; second beamformer means for producing a second beamformer signal jAoS2 using the formula S2 = WnT M Yn+1 where

    and Aç represents the said first and second predetermined amounts for the present cycle; and output signal forming means connected to the said first and second beamformer means for receiving therefrom the said first and second beamformer signals and operable to produce a first-beam-pattern output signal corresponding to the said first trial beam pattern in dependence upon the difference between the said first and second beamformer signals and to produce a second-beam-pattern output signal corresponding to the said second trial beam pattern in dependence upon the sum of the said first and second beamformer signals.
28. 28. Apparatus as claimed in claim 27, wherein at least one complex-conjugate multiplier is shared by the said first and second beamformer means.
29. 29. Apparatus as claimed in claim 27 or 28, wherein the said determining means include weight updating means for deriving a set of weights Wu,,. for the next cycle of the said series using the formula: Wn+1 = Wn + j##MWn
30. 30. Apparatus as claimed in claim 29, wherein at least one complex-conjugate multiplier is shared by the said weight updating means and one or both of the said first and second beamformer means.
31. 31. Apparatus as claimed in any one of claims 25 to 30, for use in a code-division-multiple-access mobile communications network, including RAKE receiving means having: a plurality of RAKE finger means, each including its own said processing means and its own said determining means; and RAKE combiner means for combining respective output signals of the different finger means to produce a combined signal from which the said wanted signal is detected.
32. 32. Apparatus as claimed in any one of claims 25 to 30, for use in a code-division-multiple-access mobile communications network, wherein the said processing means includes: beamforming means for producing an output signal, representing the said wanted signal, for each trial beam pattern; RAKE receiver means connected to the said processing means for receiving therefrom the said output signal for each trial beam pattern and operable to identify and combine output-signal components corresponding to different paths of the said wanted signal; and quality measuring means for producing the said quality measure for each said trial beam pattern based on the said combined components for the trial beam pattern concerned.
33. 33. A base station for use in a mobile communications network, including receiving apparatus as claimed in any one of claims 25 to 32.
34. 34. A receiving method substantially as hereinbefore described with reference to Figures 2 to 6 of the accompanying drawings.
35. 35. Receiving apparatus substantially as hereinbefore described with reference to Figures 2 to 6 of the accompanying drawings.
36. 36. A base station substantially as hereinbefore described with reference to Figures 2 to 6 of the accompanying drawings.