The present invention relates to an adaptive array apparatus, a radio base station, and a mobile phone for correcting the differences of the transfer characteristics between a transmission unit and a reception unit of a plurality of radio systems within an adaptive array apparatus designed for wireless communication.
With the recent increase in mobile stations such as PHS devices and mobile phones, there is an increasing social need for more effective use of radio wave frequency resources. An example of a communication method that addresses this need is the spatial multiplexing method.
The spatial multiplexing method is a communication method that involves the usage of an adaptive array apparatus that forms a different directivity pattern (called an adaptive array pattern) for each of a plurality of mobile stations. By doing this, the adaptive array apparatus is able to perform communication functions by multiplexing the transmission and reception signals of the plurality of mobile stations simultaneously on the same frequency.
An adaptive array apparatus includes a plurality of radio units made up of an antenna, a transmission unit, and a reception unit. And this apparatus forms directivity patterns (called array antenna patterns) for the entire antenna by adjusting the amplitudes and phases of the transmission signals and reception signals inputted to and outputted from each radio unit. More specifically, the array antenna patterns are formed by weighting the amplitudes and phases of the transmission signals and reception signals inputted to and outputted from each radio unit using a weight coefficient (also called a weight vector) The calculation of weight vectors is performed by a DSP (Digital Signal Processor) within the adaptive array apparatus.
When an adaptive array is used in a mobile phone system, the mobile phone unit has physical limitations such as size and possible number of antennas, making it impossible for the mobile phone to perform the function of controlling the directivity pattern. Therefore, it is the radio base station that forms the directivity patterns both at times of reception and times of transmission. In that case, the array antenna pattern formed by the radio base station at the time of transmission is same as the one that has been optimally formed at the time of reception.
However, even if the weight vector calculated at the time of reception is used at the time of transmission, it is not always the case that the same array antenna pattern will be formed for transmission and reception. This is due to the fact that, within each radio unit, the transfer characteristics of the transmission unit vary from those of the reception unit. These differences stem, for example, from the fact that the units are two physically separate circuits and there are variations between the characteristics of the circuit elements. In particular, this variation between the characteristics of the circuit elements come about in LNAs (Low Noise Amps) found in reception units and HPAs (High Power Amps) found in transmission units, these amps varying individually or undergoing temperature changes in certain usage environments. Due to these factors, variation comes about in the transfer characteristics such as phase rotation amounts and amplitude fluctuation amounts produced when a signal is passed through the transmission unit and the reception unit.
The differences between the transfer characteristics of reception units and transmission units have direct influences to errors in array antenna patterns at times of receiving and transmitting. For this reason, it is necessary that differences in transfer characteristics between transmission units and reception units be measured in order to perform calibration to compensate for those differences. An example of such a calibration method can be found in Japanese Laid-Open Patent Application No. H11-312917, titled “Array Antenna Apparatus”.
This array antenna includes, as additional equipment, a calibration-specific desired signal generation means, a calibration-specific interference signal generation means, an electrical power control means for controlling the electrical power of the calibration-specific interference signal generation means, a combining means for combining the calibration-specific desired signal with the calibration-specific interference signal whose electrical power is controlled, and a distribution means for distributing the combined signal to the various antennas, so as to compensate for the transfer characteristics within a reception system.
This additional equipment is required within the adaptive array apparatus of the prior art to assess the difference between transfer characteristics of transmission circuits and reception circuits in each separate radio unit, causing a problem wherein the scale of the circuits are increased. In other words, the problem with the prior art is that the scales of its circuits are increased in order to include a calibration-specific circuit that would not be required in regular communication.
In light of the above problems, an object of the present invention is to provide an adaptive array apparatus, a radio base station, and a mobile phone that reduce the size of the required circuitry and make corrections so that the same array antenna patterns are formed at times of receiving and times of transmitting.
In order to achieve the above object, the present invention provides an adaptive array apparatus, a radio base station, and a mobile phone each of which includes a plurality of radio units each made up of a transmission unit, a reception unit, and an antenna, comprising: a retaining means for retaining relative correction values that each indicate a difference of transfer characteristics between (a) a first radio unit out of the plurality of radio units and (b) each of the other radio units besides the first radio unit; and a correction means for correcting, based on the relative correction values retained by the retaining means, a signal either transmitted from or received by each of the other radio units besides the first radio unit.
With this arrangement, the correction values used for the forming of matching array antenna patterns at times of reception and times of transmission are relative correction values based on the first radio unit, making it unnecessary to include calibration (correction value measurement) circuitry within the apparatuses of the present invention. This results in a reduction in the scale of the circuitry. Moreover, the fact that there is no need for any correction of the first radio unit further leads to a reduction in the scale of the correction unit circuitry.
The present invention further provides an adaptive array apparatus wherein the plurality of radio units included therein are the first through an m'th radio unit where m is an integer greater than or equal to 2, and the retaining means retains the following values (1) and (2) as the relative correction values for an n'th radio unit where n is an integer that satisfies 2≦n≦m, (1) a phase correction value Δθ1n derived by a formula Δθ1n=((θTX1−θRX1)−(θTXn−θRXn)) and (2) an amplitude correction value Amp1n derived by a formula Amp1n=((ATX1/ARX1)/(ATXn/ARXn)) wherein, (a) θTX1 and θRX1 respectively represent phase fluctuation characteristics of the transmission unit and the reception unit within the first radio unit, (b) ATX1 and ARX1 respectively represent amplitude fluctuation characteristics of the transmission unit and the reception unit within the first radio unit, (c) θTXn and θRXn respectively represent phase fluctuation characteristics of the transmission unit and the reception unit within the n'th radio unit, and (d) ATXn and ARXn respectively represent amplitude fluctuation characteristics of the transmission unit and the reception unit within the n'th radio unit.
With the stated arrangement, the correction values used are derived from relative phase correction values and relative amplitude correction values.
Note here that the adaptive array apparatus may be arranged so that the correction means includes: a phase correction unit that corrects a phase of the signal to be transmitted from the n'th radio unit using the phase correction value Δθ1n; and an amplitude correction unit that corrects an amplitude of the signal to be transmitted from the n'th radio unit using the amplitude correction value Amp1n.
Additionally, the adaptive array apparatus may be arranged so that a gain of the antenna of a second radio unit is less than a gain of the antenna of the first radio unit, the retaining means further retains a compensation value that compensates for the gain of the antenna of the second radio unit, and the amplitude correction unit corrects the amplitude of the signal to be transmitted from the second radio unit using the compensation value.
This arrangement is particularly effective in mobile phones since they have limited space for the installation of antennas. In such a case, compensation can be made for the second radio unit, which would likely be, for instance, an internal pattern antenna or a chip antenna with less antenna gain than the first radio unit (such as a rod antenna).
Still further, it may be arranged so that each of the adaptive array apparatus, the radio base station, and the mobile phone further comprises: a calculating means for calculating a weight coefficient for each of the first through the m'th radio units for the purpose of adjusting, at a time of reception, phases and amplitudes used in forming array antenna patterns; and a weighting means for weighting the signals transmitted from the first through the m'th radio units, wherein the correction means (a) calculates a correction weight coefficient by combining the phase correction value and the amplitude correction value retained by the retaining means with the weight coefficient calculated for the n'th radio unit, and (b) instructs the weighting means to weight the signals transmitted from the second through the m'th radio units using the correction weight coefficient.
With this arrangement, since the correction means uses the correction weight coefficient to control weighting performed by the weighting means, it is possible to use the weighting means that is naturally provided within the adaptive array apparatus. This does away with the requirement for separate correction circuitry within the apparatus, which results in the reduction of the overall scale of circuitry.
The preferred embodiment of the radio base station, mobile phone, and measurement apparatus of the present invention will be explained in the following order.
The following is a general overview of an adaptive array apparatus of the preferred embodiment when it performs the function of a radio base station in a mobile communications network.
As indicated in this drawing, the adaptive array apparatus includes radio units 1 through 4 and DSP (Digital Signal Processors) 50. For the sake of clarity, 4 DSP 50 units are shown in the drawing, but, in actuality, only 1 is required. This adaptive array apparatus independently measures the correction values and, under normal communication conditions, it uses this measured correction value to perform its communication functions. In other words, this adaptive array apparatus also serves the function of a measurement apparatus.
Radio unit 1 is composed of an antenna 10, a transmission unit 111 (TX1 in the drawing), a reception unit 112 (RX1), and an antenna switch 113 (SW1). Radio units 2 through 4 are constructed in the same fashion.
θRX1 and ARX1 respectively represent the amount of phase fluctuation and the amount of amplitude fluctuation produced when a signal passes through the antenna 10, the antenna switch 113, and the reception unit 112. θTX1 and ATX1 respectively represent the amount of phase fluctuation and the amount of amplitude fluctuation produced when a signal passes through the transmission unit 111, the antenna switch 113, and the antenna 10. θRX2 through θRX4 and ARX2 through ARX4 similarly represent the amount of phase and amplitude fluctuation occurring in their respective radio units.
Additionally, Δθ12 and Amp12 respectively represent the phase fluctuation amount and amplitude fluctuation amount of radio unit 2 relative to radio unit 1, which is used as the standard. Δθ23, Δθ34, and Δθ41, as well as Amp23, Amp34, and Amp41 similarly represent the relative amounts of phase and amplitude fluctuation occurring in their respective radio units relative to radio unit 1. These are defined by the following formulas.
Δθ12=((θTX1−θRX1)−(θTX2−θRX2)) (1)
Δθ23=((θTX2−θRX2)−(θTX3−θRX3)) (2)
Δθ34=((θTX3−θRX3)−(θTX4−θRX4)) (3)
Δθ41=((θTX4−θRX4)−(θTX1−θRX1)) (4)
Amp12=((ATX1/ARX1)/(ATX2/ARX2)) (5)
Amp23=((ATX2/ARX2)/(ATX3/ARX3)) (6)
Amp34=((ATX3/ARX3)/(ATX4/ARX4)) (7)
Amp41=((ATX4/ARX4)/(ATX1/ARX1)) (8)
This adaptive array apparatus performs array-transmissions and array-receptions of known signals within radio units 1 through 4. Then, while changing the phase amounts and amplitude amounts, it seeks an adjustment value for the purpose of creating an array antenna pattern that is the same at the time of transmission and the time of reception. Accordingly, the adaptive array apparatus finds the relative phase fluctuation amounts and relative amplitude fluctuation amounts indicated by these formulas (1) through (8), and determines correction values for the purpose of compensating for the phase and amplitude fluctuation amounts. These correction values are defined by the following formulas (9) through (17).
θ_correction—1=0 (9)
θ_correction—2=Δθ12 (10)
θ_correction—3=Δθ12+Δθ23 (11)
θ_correction—4=Δθ12+Δθ23+Δθ34 (12)
A_correction—1=1 (13)
A_correction—2=Amp12 (14)
A_correction—3=Amp12*Amp23 (15)
A_correction—4=Amp12*Amp23*Amp34 (16)
θ_correction_x and A_correction_x are the correction values for the transmission signals from radio unit x (x ranging from 1 through 4).
The above correction values are correction values relative to radio unit 1, which serves as the standard. The reason that it is acceptable to use these relative correction values is as follows: If the difference of the phase fluctuation amounts and the ratio of the amplitude fluctuation amounts of the radio units are equal at the time of reception and transmission, then, when the weight vector calculated at the time of reception is used, the same array antenna pattern as the array antenna pattern that was present at the time of reception will result.
Additionally, in the above equations (9) through (16), radio unit 1 was used as the standard, however it is acceptable to use any of the radio units as the standard. For instance, if radio unit 3 were to serve as the standard, the following equations would define the phase correction values ((9′) through (12′)) and the amplitude correction values ((13′) through (16′)).
θ_correction—1=Δθ34+Δθ41 (9′)
θ_correction—2=Δθ34+Δθ41+Δθ12 (10′)
θ_correction—3=0 (11′)
θ_correction—4=Δθ34 (12′)
A_correction—1=Amp34*Amp41 (13′)
A_correction—2=Amp34*Amp41*Amp12 (14′)
A_correction—3=1 (15′)
A_correction—4=Amp34 (16′)
The following is an explanation of the general method for measuring the relative phase and amplitude fluctuation amounts.
The same drawing also shows the manner in which radio units 3 and 4 serve as a 2-antenna adaptive array apparatus, receiving the desired signal by forming an array antenna pattern that corresponds to radio unit 1 ({circle around (2)} in the drawing). That is to say, DSP 50 calculates a weight vector for the purpose of separating out the desired signal from a received signal on which both the desired signal and the interference signal are multiplexed.
As indicated by the solid line in the drawing, the array antenna pattern produced at the time of this array transmission will be the same as that produced at the time of array reception as long as the amount of phase fluctuation and amplitude fluctuation within the transmission units and reception units within radio units 3 and 4 are the same. In such a case, the directivity is oriented toward radio unit 2, and a null is directed toward radio unit 3. (When “a null is directed” to a point or a direction, it means that it is impossible or difficult for the radio waves to reach at such a point or a direction.)
In actuality, the amount of phase and amplitude fluctuation within the transmission units and reception units are never equal. Therefore, as indicated by the dashed line and the dashed-dotted line in the drawing, a divergence in the array antenna pattern is produced.
At this point, DSP 50 adds phase compensation amounts Δθ to the transmission signal from radio unit 4 while altering the compensation amounts by 360 degrees (for example, one degree at a time, from −180 to +180 degrees). While this occurs, radio unit 2 measures the signal reception level in accordance with these alterations ({circle around (4)} in the drawing). The phase compensation amount Δθ at the time when the reception signal level is the smallest satisfies the equation Δθ34=((θTX3−θRX3)−(θTX4−θRX4)). Accordingly, Δθ34 is set as the phase compensation amount Δθ at this time ({circle around (5)} in the drawing)
Furthermore, DSP 50 also gradually alters the transmission signal amplitude compensation amount Amp_coef of radio unit 4 (for example, from 0.5 to 2, by varying 0.1 at a time). While this occurs, radio unit 2 measures the signal reception level in accordance with these alterations ({circle around (6)} in the drawing). The amplitude compensation amount Amp_coef at the time when the reception signal level is the smallest satisfies the equation Amp34=((ATX3/ARX3)/(ATX4/ARX4)). Accordingly, Amp34 is set as the amplitude compensation amount Amp_coef at this time ({circle around (7)}).
In this manner, the adaptive array apparatus measures a relative phase fluctuation amount Δθ34 and a relative amplitude fluctuation amount Amp34. And in the same fashion, it measures a) Δθ41 and Amp41, b) Δθ12 and Amp12, and c) Δθ23 and Amp23.
Furthermore, DSP 50 determines whether the measured relative phase and amplitude fluctuation amounts are appropriate by using formulas (17) and (18) below.
|δθ12+Δθ23+Δθ34+Δθ41|<θthre (17)
In this case, θthre is, for example, the threshold value of 1 degree. The left section of this equation (17) involves the addition of the right sections of equations (1) through (4) listed earlier. Ideally, these amounts should add up to 0 (degrees), but in reality, measurement errors and mis-measurements due to waves of an external source are produced. Therefore, it is desirable that judgments be made based on the relationship to θthre.
A—thre—min<Amp12*Amp23*Amp34*Amp41<A—thre—max (18)
In this case, for example, A thre_min might be a threshold minimum value of 0.95 and A_thre max a threshold maximum of 1.05. The middle section of this equation (18) involves the multiplication of the right sections of equations (5) through (8) listed earlier. Ideally, these amounts should add up to 1, but for the same reasons mentioned above, it is desirable that judgments be made based on the relationship to A_thre_min and A_thre_max.
When the values satisfy the conditions of equations (17) and (18), the adaptive array apparatus uses them to calculate the correction values indicated in equations (9) through (16) (or equations (9′) through (16′)). Then, at the time of transmission, DSP 50 uses these values to correct the transmission signal.
The following is a supplemental explanation regarding the relative phase fluctuation values and the relative amplitude fluctuation values.
As indicated in
In like fashion, the amplitude fluctuation amount produced by radio unit 3 at the time of transmission relative to the time of reception is (ATX3/ARX3) and the amplitude fluctuation amount produced by radio unit 4 at the time of transmission relative to the time of reception is (ATX4/ARX4)
Fluctuation of the phase Δθ of the transmission signal of radio unit 4 is done in small increments, and when the reception level at radio unit 2 is at a minimum, this means that the phase fluctuation amounts at radio units 3 and 4 have been compensated for.
That is, (ΔTX3−ΔRX3)=(ΔTX4−ΔRX4)+Δθ34. Accordingly, Δθ34=((ΔTX3×ΔθRX3)−(ΔTX4−ΔRX4)).
In like fashion, when the reception level at radio unit 2 is at a minimum, this means that the amplitude fluctuation amounts at radio units 3 and 4 have been compensated for.
That is, (ATX3/ARX3)=(ATX4/ARX4)*Amp34. Accordingly, Amp34=((ATX3/ARX3)/(ATX4/ARX4)).
Base band unit 70 functions between a plurality of circuit lines that are connected via a telephone switching network and the modem unit 60. For each of a plurality of signals (base band signals that carry audio or data information) that require spatial multiplexing, the base band unit 70 performs a TDMA/TDD process for multiplexing and dividing the signals so that they fit into TDMA/TDD frames. A TDMA/TDD frame is made up of a 5 mS period divided into 8 equal time slots, 4 transmission time slots and 4 reception time slots.
Concretely speaking, for the purpose of time division multiplexing, the base band unit 70 multiplexes quadruply the signals from the circuit lines to the modem unit 60 for each TDMA/TDD frame. And further, for the purpose of spatial multiplexing, up to 4 signals can be sent to the modem unit 60 for every 1 transmission time slot. Additionally, the base band unit 70 inputs from the modem unit 60 to the circuit lines up to 4 signals for every 1 reception time slot, resulting in time division multiplexing that allows for an output to a plurality of circuit lines.
The modem unit 60 modulates the signals inputted from the base band unit 70, and it demodulates the signals inputted from DSP 50. The method used for this modulation and demodulation is π/4 shift QPSK (Quadrature Phase Shift Keying)
DSP 50 is a digital signal processor that runs programs for performing weight vector calculations and the like. In particular, during the calibration process, this unit calculates the correction value to be used to compensate for the change in transfer characteristics between the time of reception and the time of transmission at radio units 1 through 4.
At the time of array transmission, front end units 11, 21, 31, and 41 convert into RF signals the signals that were weighted by DSP 50 and transmit the converted signals from antennas 10 through 40. At the time of array reception, the front end units convert the signals from antennas 10 through 40 into base band signals and output the converted signals to DSP 50. Henceforth, the assembly of antenna 10 and front end unit 11 shall be referred to as radio unit 1. In the same manner, the remaining antenna and front end unit shall be referred to as radio units 2, 3, and 4.
As shown in
Control unit 80 serves to control the overall radio base station, including to switch between transmission and reception of each radio unit.
As indicated in the drawing, DSP 50 is provided with user signal processing units 51a through 51d, adders 551 through 554, switches 561 through 564 for switching between transmission and reception, correction value retaining unit 570, and correction units 571 through 574.
User signal processing units 51a through 51d are provided to correspond to the maximum of 4 user signals that are spatially multiplexed in each time slot. Normally (at times other than times of calibration processing), each user signal processing unit serves to control the array reception and array transmission that involves the usage of all 4 radio units. That is, at times of reception, the user signal processing units calculate weight vectors from the various reception signals coming from radio units 1 through 4. And, using these weight vectors, they extract a user signal by combining the reception signals that are input by the radio units 1 through 4 via the switches 561 through 564. At times of transmission, the user signal processing units use the weight coefficient calculated during the immediately preceding reception time slot to generate a weighted user signal and output that signal to each of the radio units 1 through 4.
At times of calibration processing, there are cases in which each user signal processing unit: a) controls array reception and array transmission so that it occurs through 2 antennas; b) controls the desired signal so that, rather than being transmitted and received as an array, it is an independently transmitted and received signal via 1 radio unit; and c) controls the interference signal so that, rather than being transmitted and received as an array, it is an independently transmitted and received signal via 1 radio unit. By performing the consecutive processing indicated in
The adder 551 combines the weighted components of each user transmission signal corresponding to radio unit 1. However, in cases such as that indicated in
Correction value retaining unit 570 retains the correction values (θ_correction_1 through θ_correction_4, and A_correction_1 through A_correction_4) that are calculated during the calibration process.
At times other than that of the calibration process, correction unit 571, using the correction values θ_correction_1 and A_correction_1 retained in the correction value retaining unit 570, corrects the transmission signals arising from adder 551 and outputs these signals to radio unit 1 via switch 561. And at times of calibration processing, the transmission signals arising from adder 551 are output directly to radio unit 1 via switch 561. However, if the relative phase and amplitude fluctuation amounts at radio unit 1 are being measured during a time of calibration processing, the phase compensation amount Δθ and the amplitude compensation amount Amp are applied to the transmission signals while being altered gradually.
Correction units 572 through 574 function in the same manner in relation to their respective radio units other than the fact the correction values retained by the correction value retaining unit 570 are different.
As indicated in the drawing, the user signal processing unit 51a is provided with a weight calculating unit 53, an adder 54, a memory 55, switches 56 and 57, multipliers 521 through 524, and multipliers 581 through 584.
When calibration processing is not taking place, the weight calculating unit 53 calculates a weight vector in order to minimize an error between a reference signal generated by the memory 55 and the sum of the weighted reception signals S1R through S4R from each of the radio units 1 through 4, during each symbol period of a fixed bit pattern period within a reception time slot. And at times of calibration, the weight vector calculation for the purpose of an array reception using 2 antennas is performed in the same manner. The following is an explanation of weight vector calculation for a 4-antenna array reception. Note that the process is the same for 2-antenna array receptions; simply the number of elements involved is reduced.
Concretely speaking, the weight calculating unit 53 uses the following formula (19) for minimizing the error e(t) by adjusting the W1(t−1) through W4(t−1) values and taking these adjusted values to serve as weight coefficients for the symbol at time t, W1(t) through W4(t).
e(t)=d(t)−(W1(t−1)*X1′(t)+W2(t−1)*X2′(t)+W3(t−1)*X3′(t)+W4(t−1)*X4′(t)) (19)
In the above formula: t represents the timing of a symbol unit; d(t) represents the symbol data within the known reference signal (or training signal); W1(t−1) through W4(t−1) represent either the weight coefficient of each antenna that was calculated for the immediately preceding symbol or the weight coefficient calculated at the immediately preceding reception time slot; and X1(t) through X4(t) represent the reception signals for antennas 10 through 40.
The weight vectors are adjusted in the above manner for every symbol. Therefore, even if the error e(t) is large at the beginning of a reference signal period within the reception time slot, it will be reduced to a minimum (or reduced to 0) by the end of that reference signal period.
Furthermore, at the time of the symbol period in which the weight coefficient within the reception time slot is calculated and at times of symbol periods to follow, the weight calculating unit 53 outputs the calculated weight coefficients to the multipliers 521 through 524. Additionally, during a given transmission time slot, the weight calculating unit 53 outputs to the multipliers 581 through 584 the weight coefficient calculated during the reception time slot that corresponds to the immediately preceding transmission time slot.
The memory 55 stores a) waveform data of series of symbols that represent the reference signals used at times other than calibration processing (times of normal communication with mobile stations), b) waveform data of series of symbols that represent the desired signals used for calibration processing, and c) waveform data of series of symbols that represent interference signals. The reference signal is retrieved by the weight calculating unit 53 in synchronization with the symbol timing during the reception period of known fixed bit patterns (fixed symbols) within a reception time slot. For example, in the case of PHS systems, the SS (start symbol), PR (preamble), and UW (unique word) are fixed symbols that appear in the front end of the reception time slot.
It is acceptable if the desired signal and the interference signal are known series of symbol data such as PN (Pseudo-random Noise) symbols, and it is desirable that they be arranged orthogonal to each other. If arranged orthogonal to each other, the weight vectors can be more rapidly converged and more accurately calculated. Note that when the same PN symbols or fixed-symbols are used, it is best to stagger the timing (for instance, by 0.5 symbol periods).
In cases such as that illustrated by Ant3 and Ant4 of
Although it is acceptable that each user signal processing unit have the same structure, for ease of explanation, each processing unit will be assumed to undergo a particular fixed process during the calibration process.
As indicated in the drawings, during the first half of the calibration process (that is, in cases such as that illustrated in
In this case, as indicated by the “First Half” column in
During the second half of the calibration process (that is, in cases such as that illustrated in
In this case, as indicated by the “Second Half” column in
After variable n is initialized (n=1, Step 81), DSP 50 selects which physically existing radio units 1 through 4 will serve as which theoretical radio units Ant1 through Ant4 (Step 82). In this case, Ant1 is selected to independently transmit and receive the desired signal, Ant2 to independently transmit and receive the interference signal, and Ant3 and Ant4 are selected for the purpose of array reception and array transmission.
DSP 50 then has Ant 1 transmit the desired signal and has Ant 2 transmit the interference signal (Step 83). Simultaneously, it has Ant3 and Ant4 function as an adaptive array apparatus to form an array antenna pattern for the desired signals sent from Ant1. That is to say, DSP 50 calculates a weight vector for the purpose of separating out the desired signal from the reception waves that include both desired signals and interference signals in a multiplexed form (Step 84). At this time, user signal processing unit 51a supplies the desired signal to Ant1 and user signal processing unit 51b supplies the interference signal to Ant2. Additionally, user signal processing unit 51c calculates the weight vectors of the reception signals from Ant3 and Ant4.
At this juncture, if the weight vectors are not sufficiently converged—that is, if the error e (t) of formula (19) is greater than a certain threshold value—it is acceptable to terminate the calibration process and recommence it from the beginning.
If the weight vectors are sufficiently converged, DSP 50 then a) uses Ant3 and Ant4 as a 2-antenna adaptive array apparatus to perform an array transmission of the desired signal by using the calculated weight vector, and b) switches Ant2 to independent reception (Step 85). At this time, user signal processing unit 51c weight the signal using the weight vectors. The phases and amplitudes of the transmission signals sent to Ant3 and Ant4 after being weighted are expressed by the following: θAnt3=θAnt3_est, θAnt4=θAnt4_est, A_Ant3=A_Ant3_est, and A_Ant4=A_Ant4_est.
At the time of this array transmission, DSP 50 maintains fixed values for phase θAnt3 and amplitudes A_Ant3 and A_Ant4, and adds the phase compensation amount Δθ to the phase amount of the transmission signal being sent to Ant4 while altering the amount Δθ by 1 degree at a time from −180 degrees to +180 degrees (θAnt4=θAnt4_est+Δθ). Then, for every Δθ, DSP 50 measures the reception signal level at Ant2. (Steps 86 through 89.) The phase compensation amount Δθ at this time is added to the transmission signal that is inputted by correction unit 574, shown in
Next, DSP 50 takes the phase compensation amount Δθ, when the reception signal level measured at Ant2 is at a minimum, to be equal to Δθ34, wherein Δθ34=(θTX3−θRX3)−(θTX4−θRX4) (Step 90).
Furthermore, DSP 50 maintains fixed values for phase θAnt3, θAnt4 (=θAnt4_est+Δθ34) and amplitude A_Ant3, and multiplies the amplitude compensation amount Amp_coef with the amplitude of the transmission signal of Ant4 while gradually altering that amount by, for example, 0.05 units at a time within a range of 0.5 to 2 (A_Ant4=A_Ant4_est*Amp_coef). Then, for every Amp_coef, DSP 50 measures the reception signal level at Ant2. (Steps 91 through 94.) The amplitude compensation amount Amp_coef at this time is multiplied by the transmission signal that is inputted by correction unit 574, shown in FIG. 4, from user signal processing unit 51c via adder 554. Then the amplitude compensation amount is outputted to Ant4 via switch 564.
Next, DSP 50 takes the phase compensation amount Amp_coef, when the reception signal level measured at Ant2 is at a minimum, to be equal to Amp 34 wherein, Amp 34=((ATX3/ARX3)/(ATX4/ARX4)) (Step 95).
By the above stated steps, the phase and amplitude fluctuation amounts of Ant4, θ34 and Amp34, relative to Ant3 are measured. Furthermore, by using the loop formed by Steps 96 and 97, DSP 50 alters the assembly of theoretical radio units Ant1 through Ant4, which were chosen from the physically existing radio units 1 through 4. While performing these alterations, DSP 50 measures Δθ41 and Amp41 during the second loop, Δθ12 and Amp12 during the third loop, and Δθ23 and Amp23 during the fourth loop.
Next, as shown in
When both equations (17) and (18) are satisfied, DSP 50 calculates phase correction values θ_correction_1 through θ_correction_4 and amplitude correction values A_correction_1 through A_correction_4 in accordance with the earlier-explained equations (9′) through (16′) (Steps 100 and 101). Then the calculated correction values are written to correction value retaining unit 570 and are used for correcting the transmission signal of the various radio units at times of normal array transmission when calibration is not taking place.
As explained above, the adaptive array apparatus of the present embodiment allows for array transmission to take place between 2 radio units selected from a plurality of radio units and other radio units. And in accordance with the resulting reception signals, the transfer characteristics of the selected radio units can be measured, making it possible to calculate the relative transfer characteristics of each radio unit without any additional apparatuses.
The adaptive array apparatus shown in
Below is an explanation of the structure of the present invention when its adaptive array apparatus is a mobile phone in a mobile communications network. That will be followed by an explanation of the above-mentioned measurement apparatus.
Either of the two antennas 210 and 220 can be any of an assortment of antennas such as a rod antenna, a flat-surfaced pattern antenna, a helical antenna at the end of a rod, and a chip antenna (made up of chip components installed upon a base board). However, in this particular case, it is assumed that antenna 210 is a rod antenna and antenna 220 is a chip antenna.
DSP 260, which is indicated by the dashed lines in the drawing, actually operates in accordance with a program. In this drawing, the operations are indicated by being divided into blocks that represent particular functions. The functions of DSP 260 correspond to multipliers 214, 224, 215, and 225, adder 230, demodulating circuit 231, re-modulating circuit 232, memory 233, switch 234, counter 235, weight calculating unit 236, memory 237, weight control unit 238, correction control unit 239, phase shifter 240, amplifier 241, and modulating circuit 242.
Multipliers 214 and 224 respectively weight the reception signals inputted from reception circuits 212 and 222 by multiplying them by weight vectors W1 and W2 produced by weight calculating unit 236.
Multipliers 215 and 225 respectively weight the transmission signals inputted by modulating circuit 242 by multiplying them by weight vectors W1 and W2 produced by weight control unit 238. Then they output these weighted transmission signals to transmission circuit 211 and phase shifter 240.
Adder 230 adds together the reception signals that were weighted by multipliers 214 and 224.
Demodulating circuit 231 demodulates the reception signals that result after addition is performed by adder 230. The demodulated reception signals are outputted as reception series of bit data.
Re-modulating circuit 232 re-modulates reception series of bit data that are inputted by demodulating circuit 231, converting the reception series of bit data into symbol data (waveform data of symbols).
Memory 233 retains the reference signal table. The reference signal table stores symbol data (waveform data of symbols) that represents reference signals used at times other than the calibration process (times of normal reception from the radio base station), and it stores symbol data that represents desired signals that are used at times of calibration. In this case, the reference signals and desired signals are the same as those included in the explanation of the radio base station.
During normal reception, counter 235 counts, in synchronization with the symbol timing, the number of symbols in a reception time slot from the front end to the tail end (in a PHS, this ranges from 0 to 120). The resulting count value is used for the purpose of distinguishing symbol periods with a fixed bit pattern from other periods. During normal reception, the symbol periods of the third symbol through the 16th symbol correspond to the fixed bit pattern periods of SS, PR, and UW.
During normal reception, when the count value from counter 235 indicates a symbol period with a fixed bit pattern, switch 234 selects the symbol data (waveform data) that represents the reference signal retrieved from memory 233. For all other periods, switch 234 selects the symbol data from re-modulating circuit 232. And during the calibration process, it selects the symbol data that represents the desired signal retrieved from the memory 233.
During times of normal reception or times of calibration, weight calculating unit 236 calculates a weight vector for every symbol in order to minimize the error between the sum of the weighted reception signals inputted from reception circuit 212 and reception circuit 222 and the symbol data inputted from switch 234. Regarding the calculation of weight vectors, in this case it is performed in the same manner as weight calculating unit 53.
Memory 237, which includes RAM and ROM, stores the weight vector calculated by weight calculating unit 236 and relative correction value for radio unit B that uses radio unit A as the standard. During normal reception, it is acceptable that this weight vector be the weight vector calculated according to the symbol at the tail end of a reception time slot. It is used in the transmission time slot that immediately follows the reception time slot. During the calibration process, memory 237 stores the weight vector calculated upon the reception of the desired signal, and this weight vector is used for the immediately following desired signal transmission. The weight vectors for radio units A and B are referred to respectively as W1 and W2.
Additionally, the correction values are expressed using the following formulas (20) and (21), and the values measured during the calibration process are written onto the ROM storage area within memory 237 before it is shipped from the manufacturers.
Δθ12=((θTX1−θRX1)−(θTX2−θRX2) (20)
Amp12=((ATX1/ARX1)/(ATX2/ARX2)) (21)
During normal transmission and within a transmission time slot, weight control unit 238 retrieves weight vectors W1 and W2 from memory 237. Then it outputs these weight vectors to multipliers 215 and 216. The same occurs when a desired signal is transmitted during a calibration process.
Also during normal transmission and within a transmission time slot, correction control unit 239 retrieves correction values Δθ12 and Amp12 from memory 237, then respectively outputting these values to phase shifter 240 and amplifier 241. Additionally, within the calibration process at the time of transmission of the desired signal, correction control unit 239 outputs Δθ to the phase shifter 240 while altering it from −180 to +180 degrees one degree at a time, for instance. And it outputs Amp to the amplifier 241 while gradually altering it (for example, between 0.5 and 2, 0.05 units at a time).
Phase shifter 240 corrects the phase of the transmission signal inputted from multiplier 225 according to the correction value Δθ12 inputted from correction control unit 239.
Amplifier 241 corrects the amplitude of the transmission signal inputted from phase shifter 240 according to the correction value Amp12 inputted from correction control unit 239. Then it outputs the corrected transmission signal to transmission circuit 221.
During normal transmission, modulating circuit 242 modulates the series of bit data to be transmitted, generating a transmission signal (symbol data).
External I/F 250 is a connector connected to the DSP 260 input/output port and the DSP 260 memory (including memories 233 and 237) port. It is installed onto the baseboard of the mobile phone. During the calibration process, this external I/F 250 is connected to the external measurement apparatus and is used for the input/output of the various programs, data, commands, and replies to those commands.
With a mobile phone with the above construction, during normal reception, reception takes place by the formation of an array antenna pattern based on weight vectors calculated in a reception time slot. At the same time, the weight vectors are stored in memory 237. And during the immediately following transmission time slot, transmission takes place by the formation of an array antenna pattern based on the stored weight vectors.
At the time of this transmission, correction control unit 239 uses correction values Δθ12 and Amp12 stored in memory 237 to correct the transmission signal of radio unit B. As a result, it is possible to make corrections that ensure that there will be no deviation in the array antenna pattern at the time of reception and at the time of transmission. In other words, despite a difference in the phase and amplitude fluctuation characteristics of radio unit A and those of radio unit B, directivity of the signal at the time of reception can be made to match the directivity of the signal at the time of transmission by simply correcting the transmission signal of radio unit B. There is no need to correct the transmission signal of radio unit A, which serves as the standard.
Furthermore, by including external I/F 250, if the calibration process is conducted under the control of an external measurement apparatus, measuring the above correction values can be done with ease.
Note that in the case of the above-mentioned mobile phone, since Δθ12 and Amp12 are the same physical amounts as the weight vectors it is acceptable that the structure of the invention be such that the correction weight vectors that represent Δθ12 and Amp12 be stored in memory 237, and multipliers are provided in place of phase shifter 240 and amplifier 241. Additionally, one should note that the circuits of correction units 571 through 574 are equivalent to those of phase shifter 240 and amplifier 241, as well as those the multipliers.
Finally, as antenna 210 is a rod antenna and antenna 220 is a chip antenna, when the gain of the two antennas vary, it is acceptable that the above-mentioned Amp12 value be a value compensated by an antenna gain compensation value A_cmp, as indicated below.
Amp12=A—cmp*((ATX1/ARX1)/(ATX2/ARX2))
As shown in the drawing, the measurement apparatus is provided with a transmission and reception apparatus 301, a transmission apparatus 302, a timing adjuster 331, a control PC 330, a clock generation circuit 332, and an I/F unit 333.
In order to fulfill the role of Ant2 indicated in FIGS. 2A and 2B, transmission and reception apparatus 301 is provided with an antenna 310, a transmission circuit 311, a signal selection unit 312, a reception circuit 313, a level measurement unit 314, and a switch 315. With this arrangement, it functions to receive a desired signal that is transmitted from the mobile phone 200 after an interference signal is transmitted.
Via switch 315, transmission-circuit 311 transmits from antenna 310 an interference signal that is inputted from signal selection unit 312.
Signal selection unit 312 stores series of symbol data of a plurality of interference signals. Then it selects one of the interference signals and outputs it to transmission circuit 311. This plurality of interference signals includes a) primary interference signals made up of PN codes, and b) secondary interference signals made up of known series of codes, including the fixed bit patterns (SS, PR, UW) that are the same as normal transmission time slots. The selection of interference signals is dependent on commands from control PC 330.
Via antenna 310 and switch 315, reception circuit 313 receives from the mobile phone 200 a transmission signal that directs a null toward transmission and reception apparatus 301.
Level measurement unit 314 measures the reception signal level of the reception signal of reception circuit 313. Then it notifies control PC 330 of the measured reception signal level.
In order to fulfill the role of Ant1 shown in
Via switch 325, transmission circuit 321 transmits from antenna 320 a desired signal that is inputted from signal selection unit 322.
Signal selection unit 322 stores series of symbol data of a plurality of desired signals. Then it selects one of the desired signals and outputs it to transmission circuit 321. This plurality of desired signals includes a) primary desired signals made up of PN codes that are orthogonal to the primary interference signal and b) secondary desired signals made up of known series of codes, including the fixed bit patterns (SS, PR, UW) that are the same as normal transmission time slots. The selection of interference signals is dependent on commands from control PC 330.
When signal selection units 312 and 322 respectively select a primary interference signal and a primary desired signal, timing adjuster 331 directly outputs to transmission and reception apparatus 301 the clock signal (symbol clock) inputted from signal selection unit 322. When signal selection units 312 and 322 respectively select a secondary interference signal and a secondary desired signal, timing adjuster 331 outputs to transmission and reception apparatus 301 the clock signal inputted from signal selection unit 322 after, for example, delaying the signal by a 0.5 symbol time. The reason for this delay lies in the fact that both the secondary interference signal and the secondary desired signal include the same fixed bit patterns (SS, PR, UW, etc.). That is, the delay makes it easier to separate out the desired signal occurring in the mobile phone 200. It is not necessary for the timing adjuster 331 to perform a delay when a primary interference signal and primary desired signal are selected, but, for ease of construction, it is acceptable for the timing adjuster 331 to be designed so that the delay takes place.
In like manner to the calibration process shown in
Clock generation circuit 332 outputs a clock signal, which indicates symbol timing, to transmission apparatus 302 and timing adjuster 331.
I/F unit 333 is connected to external I/F 250 that is within mobile phone 200. It serves as an interface for the input/output of commands and data to and from mobile phone 200.
Note that it is acceptable that rather than being a connector, external I/F 250 can be a plurality of pads provided on the baseboard. In such a case, I/F unit 333 can be a probe connected to the plurality of pads.
Additionally, at times of calibration, it is desirable that the measurement apparatus and mobile phone shown in
After variable n is initialized (n=1, Step 181), control PC 330 determines that transmission apparatus 302, transmission and reception apparatus 301, radio unit A (TX1, RX1 in drawing), and radio unit B (TX2, RX2 in drawing) will respectively serve as radio units Ant 1 through Ant 4 (Step 182). These correspond to Ant1 through Ant4 in
Control PC 330 commands Ant1 (transmission apparatus 302) and Ant2 (transmission and reception apparatus 301) to transmit a primary desired signal and a primary interference signal. This causes the primary desired signal waves and primary interference signal waves from transmission apparatus 302 and transmission and reception apparatus 301, respectively, to be transmitted on the same frequency (Step 183). Further, control PC 330 uses Ant3 and Ant4 as a 2-antenna adaptive array apparatus to form array antenna pattern for the desired signals sent from Ant1. That is to say, control PC 330 commands the mobile phone 200 to calculate a weight vector for the purpose of separating out the desired signal from the reception waves that include both desired signals and interference signals in a multiplexed form. This causes the weight calculating unit 236 of mobile phone 200 to calculate a weight vector for the purpose of receiving the primary desired signal (Step 184).
At this juncture, if the weight vectors are not sufficiently converged—that is, if the error e(t) of formula (19) is greater than a certain threshold value—it is acceptable that the mobile phone 200 notify the control PC 330 of this so that it can terminate the calibration process and recommence it from the beginning.
If the weight vectors are sufficiently converged, control PC 330 a) commands that Ant3 and Ant4 are used as a 2-antenna adaptive array apparatus to perform an array transmission of the desired signal by using the calculated weight vector, and b) commands Ant2 (transmission and reception apparatus 301) to receive a desired signal. This leads to the mobile phone 200 forming an array antenna pattern with a null directed toward transmission apparatus 302 and performing an array transmission of the desired signal (Step 185). At this time, the phases and amplitudes of the transmission signals sent to Ant3 and Ant4 after weighting are expressed by the following: θAnt3=θAnt3_est, θAnt4=θAnt4_est, A_Ant3=A_Ant3_est, and A_Ant4=A—Ant4_est.
During this array transmission, control PC 330 commands the mobile phone 200 to, while maintaining fixed values for phase θAnt3 and amplitudes A_Ant3 and A_Ant4, add the phase compensation amount Δθ to the phase amount of the transmission signal being sent from Ant4 while altering the amount Δθ by 1 degree at a time from −180 degrees to +180 degrees (θAnt4=θAnt4_est+Δθ). Then, control PC 330 obtains from transmission and reception apparatus 301 the reception signal measurement results for every Δθ and stores them in its memory. (Steps 186 through 189.) The phase compensation amount Δθ at this time is added to the transmission signal from multiplier 225 by means of correction control unit 239 and phase shifter 240 shown in
Further, control PC 330 stores up within its memory the reception signal level for each θ, control PC 330 takes the phase compensation amount Δθ corresponding to the minimum reception signal level as Δθ34 (=Δθ12 of
Furthermore, control PC 330 commands mobile phone 200 to, while maintaining fixed values for phase θAnt3, θAnt4 (=θAnt4_est+Δθ34) and amplitude A_Ant3, multiply the amplitude compensation amount Amp_coef with the amplitude of the transmission signal of Ant4 while gradually altering that amount by, for example, 0.05 units at a time within a range of 0.5 to 2 (A_Ant4=A_Ant4_est*Amp_coef). Then, control PC 330 a) commands transmission and reception apparatus 301 to measure the reception signal level at Ant2 for every Amp_coef, b) obtains these measurement results, and c) stores this information in its internal memory (Steps 191 through 194.) The amplitude compensation amount Amp_coef at this time is multiplied by the transmission signal from phase shifter 240 by means of correction control unit 239 and amplifier 241 shown in
Referring to the reception signal level stored in the memory, Control PC 330 takes the phase compensation amount Amp_coef corresponding to the minimum reception signal level as Amp34 (=Amp12 of
By the above stated steps, the phase and amplitude fluctuation amounts, Δθ12 and Amp12, of Ant4 (radio unit B) relative to Ant3 (radio unit A) within mobile phone 200 are measured.
Further, control PC 330 switches Ant3 with Ant4. That is, it performs the same processes (Steps 183 through 195) with Ant3 functioning as radio unit B and Ant4 functioning as radio unit A (Steps 196 and 197). However, in Steps 187 and 192, the phase shifter 240 and amplifier 241 within the mobile phone 200 do not alter the phase and amplitude. Rather, weight control unit 238 calculates a weight vector in which the above-mentioned Δθ and Amp_coef are added to the weight vector W2. Then multiplier 215 uses this calculated weight vector to perform a weighting function.
As a result of this second process the correction values Δθ12 and Amp12 for radio unit A, relative to radio unit B, are measured. These correction values are not used for the mobile phone 200, but they are used, as mentioned below, for determining the correctness of the correction values Δθ12 and Amp12.
That is to say, control PC 330 determines whether the measured relative phase fluctuation amounts (Δθ12 and Δθ21) and relative amplitude fluctuation amounts (Amp12 and Amp21) are appropriate (Steps 198 and 199). This determination is dependent on whether or not the equations (22) and (23) are satisfied. Other than the fact that only two correction values are used, these equations are the same as (17) and (18).
|Δθ12+Δθ21|<θthre (22)
A—thre—min<Amp12*Amp21<A—thre—max (23)
If control PC 330 does not satisfy either of these equations, the calibration process should be terminated and recommenced from the beginning. In such an instance, it is preferable that changes be made to the conditions, for example, to the desired signal and the interference signal.
When both equations (17) and (18) are satisfied, control PC 330 commands mobile phone 200 to write the correction values Δθ12 and Amp 12 into memory 237 (Step 200). This results in memory 237 of mobile phone 200 storing the correction values Δθ12 and Amp12.
As explained above, by means of the measurement apparatus, the relative correction values of radio unit B, relative to radio unit A within mobile phone 200, are measured, and the correction values are established within mobile phone 200.
The following are examples of variations of the structure of the above-explained embodiment of the present invention.
Additionally, the reason that the earlier-explained embodiment included the measurement of the relative phase and amplitude fluctuation amounts for every radio unit was for the purpose of using equations (17) and (18) to determine the correctness of the phase and amplitude fluctuation amounts.
Additionally, in the case of
In this case, it is acceptable to perform unit-by-unit comparisons/updates using the phase and amplitude fluctuation amounts for each radio unit stored by the correction value retaining unit 570 and newly measured phase and amplitude fluctuation amounts. If the results of this comparison show a great difference (that is, if the threshold value is surpassed), it is acceptable to have a calibration process performed on all radio units.
Additionally, in Steps 87 and 88 of
Also, it is acceptable for the measurement apparatus to measure correction values, which are relative to a standard radio unit, for every radio unit other than the one serving as the standard. And the mobile phone can be designed to correct the transmission signal of every radio unit other than the one serving as the standard. With such an arrangement, any of the radio units can serve as the standard for the same reasons as those given in (9). And in the calibration process of
Furthermore, when 4 or more radio units are provided in the structure of the mobile phone, it is possible to not include an external measurement apparatus in the design. Calibration can be done by the mobile phone itself, in the same way that it is done by the radio base station. In this case, the structure should be designed so that the calibration process program can be downloaded via external I/F 250 from the external apparatus to the memory within the mobile phone and erased after measurement is concluded. Also, the design can be such that the above-mentioned program is retained within a memory (stored in ROM). If the program is stored in ROM, calibration processes can be performed based on user operations after delivery of the product, thus reducing the aging that takes place in the radio units.
The present invention is made up of an adaptive array apparatus, a radio base station, and a mobile phone that perform corrections in order that identical array antenna patterns are formed during reception and transmission. These apparatuses have circuitry reduced in scale and are intended for use in mobile communication systems.
Number | Date | Country | Kind |
---|---|---|---|
2000-101497 | Apr 2000 | JP | national |
2000-215100 | Jul 2000 | JP | national |
Number | Date | Country | |
---|---|---|---|
Parent | 10240590 | Oct 2002 | US |
Child | 11344384 | Jan 2006 | US |