US20060256904A1 - Receiving method and receiver - Google Patents

Abstract

BB input units input baseband received signals. An initial weight data setting unit sets weighting coefficients to be utilized in the interval of a training signal as initial weighting coefficients. A gap compensating unit compensates control weighting coefficients with a gap error signal and outputs the updated weighting coefficients acquired as a result of the compensation. A weight switching unit selects the initial weighting coefficients in the interval of the training signal and selects the updated weighting coefficients in the interval of the data signal. Then the weight switching unit outputs the selected initial weighting coefficients and updated weighting coefficients as the weighting coefficients. A synthesizing unit weights the baseband received signals with the weighting coefficients and then sums them up.

Description

BACKGROUND OF THE INVENTION
• 1. Field of the Invention
• The present invention relates to a receiving technology. It particularly relates to a receiving method and a receiving apparatus which controls a weighting coefficient for synthesizing radio signals received by a plurality of antennas.
• 2. Description of the Related Art
• In wireless communication, it is in general desired to effectively use limited frequency resources. In order to use the frequency resources effectively, radio waves of same frequency are, for example, utilized as repeatedly as possible in short-range. In this case, however, communication quality degrades because of cochannel interference caused by a radio base station or mobile terminal closely located, which utilizes the same frequency. As one technology for preventing such communication quality degradation deriving from the cochannel interference, the adaptive array antenna technology can be named.
• In the adaptive array antenna technology, signals received by a plurality of antennas are respectively weighted with different weighting coefficients and synthesized. The weighting coefficients are adaptively updated so that an error signal between a signal to be transmitted and the signal after the synthesis might be small. Here, the signal to be transmitted is determined based on the signal after synthesis. In order to update the weighting coefficients adaptively, the RLS (Recursive Least Squares) algorithm, the LMS (Least Mean Squares) algorithm or the like is utilized. The RLS algorithm generally converges at high speed. The RLS algorithm, however, requires a high speed or a huge arithmetic circuit since computation performed is very complicated. The LMS algorithm can be realized with a simpler arithmetic circuit than that of the RLS algorithm. However, the convergence speed thereof is low.
• Related Art List
• 1. Japanese Patent Application Laid-open No. 2002-26788
• In utilizing the adaptive array antenna for a radio mobile terminal, it is suitable to use the LMS algorithm for updating weighting coefficients, since it is desirable that an arithmetic circuit is small. However, the convergence speed of the LMS algorithm is low in general. Thus, if it is desired to delay received signals to be synthesized until the LMS algorithm converges, processing delay becomes large and therefore it is possibly impossible to use the adaptive array antenna in a real time application such as TV conference system where permissible delay time is limited. On the other hand, a response characteristic generally degrades if the weighting coefficients at the timing where the LMS-algorithm has not converged yet in order to diminish the processing delay.
SUMMARY OF THE INVENTION
• The inventor of the present invention has made the present invention in view of the foregoing circumstances and an object thereof is to provide a receiver having simple arithmetic circuits, of which the processing delay is small. It is also an object of the present invention to provide a receiver of which the response characteristic hardly degrades even in the case the weighting coefficients have not converged yet. Moreover, it is also an object of the present invention to provide a receiver which can switch a plural types of weighting coefficients.
• A preferred embodiment according to the present invention relates to a receiver. This receiver includes: an input unit which inputs a plurality of signals on which a processing is to be performed; a switching unit which switches a plurality of weighting coefficients by which the plurality of inputted signals are multiplied between a plurality of first weighting coefficients to be temporarily utilized and a plurality of second weighting coefficients which have higher adaptabilities; a controller which instructs the switching unit to switch the weighting coefficients between the plurality of first weighting coefficients and the plurality of second weighting coefficients; and a synthesizer which synthesizes results of multiplications, where the multiplications are performed on the plurality of inputted signals and the plurality of weighting coefficients.
• The plurality of weighting coefficients include (A, B, C, D) of which the number of terms is equal to that of the plurality of signals, where the results of multiplications between them and (X1, Y1), (X2, Y2) become (AX1, BY1) and (CX2, DY2). The plurality of weighting coefficients also include (A, B) of which the number of terms is different from that of the plurality of signals, where the results of multiplications become (AX1, BY1) and (AX2, BY2).
• The receiver described above enables to acquire a response characteristic optimal in each timing by switching the weighting coefficients which have different characteristics.
• Another preferred embodiment of the present invention also relates to a receiver. The receiver includes: an input unit which inputs a plurality of signals on which a processing is to be performed; a switching unit which switches a plurality of weighting coefficients by which the plurality of inputted signals are multiplied between a plurality of first weighting coefficients and a plurality of second weighting coefficients; a controller which instructs the switching unit to switch the weighting coefficients between the plurality of first weighting coefficients and the plurality of second weighting coefficients in a prescribed interval, where the plurality of signals are inputted in a sequential manner during the interval; and a synthesizer which synthesizes results of multiplications, where the multiplications are performed on the plurality of inputted signals and the plurality of weighting coefficients.
• The “sequential manner” merely means that the known received signal is sequential. As long as the signals are inputted sequentially, the time length does not necessarily need to be long but may be short. Moreover, the sequential manner here may include a case where the signals are inputted in a discrete manner in accordance with a certain rule, if the apparatus recognizes the rule. That is, the “sequential manner” here includes every case where the receiver can recognize the manner of inputting the signals as “sequential” one.
• The plurality of first weighting coefficients may be set in a manner that, as results of multiplications by the plurality of inputted signals, a multiplication result corresponding to one signal among the plurality of inputted signals becomes effective. The one signal among the plurality of inputted signals may be a signal having a largest value among the plurality of inputted signals. The plurality of first weighting coefficients may be set by utilizing the plurality of second weighting coefficients which have already been set.
• The receiver may further include: a weighting coefficient updating unit which updates a plurality of third weighting coefficients adaptively based on the plurality of inputted signals; a gap estimator which estimates gaps between the plurality of first weighting coefficients and the plurality of third weighting coefficients by performing a correlation processing between at least one of the plurality of inputted signals and a known signal; and a gap compensator which generates the plurality of second weighting coefficients by compensating the plurality of third weighting coefficients based on the estimated gaps.
• The signals inputted during the prescribed interval in the sequential manner may include signals having different characteristics and the controller may instruct to switch the weighting coefficients between the first weighting coefficients and the second weighting coefficients when it is detected a shift point where the characteristics of the signals change. The controller may input sequentially the plurality of third weighting coefficients updated in the weight coefficient updating unit and may instruct the switching unit to switch the weighting coefficients between the first weighting coefficients and the second weighting coefficients when fluctuation of the plurality of third weighting coefficients converges within a prescribed range.
• The receiver described above enables to acquire a response characteristic optimal in each time by switching the weighting coefficients which have different characteristics during the interval.
• Still, another preferred embodiment according to the present invention relates to a receiving method. This method includes: inputting a plurality of signals on which a processing is to be performed; switching a plurality of weighting coefficients by which the plurality of inputted signals are multiplied between a plurality of first weighting coefficients to be temporarily utilized and a plurality of a second weighting coefficients which have higher adaptabilities; giving an instruction of switching the weighting coefficients between the plurality of first weighting coefficients and the plurality of second weighting coefficients; and synthesizing results of multiplications, where the multiplications are performed on the plurality of inputted signals and the plurality of weighting coefficients.
• Still another preferred embodiment according to the present invention relates to a receiving method. This method includes: inputting a plurality of signals on which a processing is to be performed; switching a plurality of weighting coefficients by which the plurality of inputted signals are multiplied between a plurality of first weighting coefficients and a plurality of second weighting coefficients; giving an instruction of switching the weighting coefficients between the plurality of first weighting coefficients and the plurality of second weighting coefficients in a prescribed interval, where the plurality of signals are inputted in a sequential manner during the interval; and synthesizing results of multiplications, where the multiplications are performed on the plurality of inputted signals and the plurality of weighting coefficients.
• The plurality of first weighting coefficients may be set in a manner that, as results of multiplications by the plurality of inputted signals, a multiplication result corresponding to one signal among the plurality of inputted signals becomes effective. The one signal among the plurality of inputted signals may be a signal having a largest value among the plurality of inputted signals. The plurality of first weighting coefficients may be set by utilizing the plurality of second weighting coefficients which have already been set.
• The receiving method may further include: updating a plurality of third weighting coefficients adaptively based on the plurality of inputted signals; estimating gaps between the plurality of first weighting coefficients and the plurality of third weighting coefficients by performing a correlation processing between at least one of the plurality of inputted signals and a known signal; and generating the plurality of second weighting coefficients by compensating the plurality of third weighting coefficients based on the estimated gaps.
• The signals inputted during the prescribed interval in the sequential manner may include signals having different characteristics. In giving the instruction of switching the weighting coefficients between the first weighting coefficients and the second weighting coefficients, the instruction may be given when it is detected a shift point where the characteristics of the signals change. The plurality of third weighting coefficients updated may be inputted sequentially in giving the instruction of switching the weighting coefficients between the first weighting coefficients and the second weighting coefficients, and the instruction may be given when fluctuation of the plurality of third weighting coefficients converges within a prescribed range.
• Yet another preferred embodiment of the present invention relates to a program. The program includes: inputting a plurality of signals on which a processing is to be performed; switching a plurality of weighting coefficients by which the plurality of inputted signals are multiplied between a plurality of first weighting coefficients to be temporarily utilized and a plurality of a second weighting coefficients which have higher adaptabilities; giving an instruction of switching the weighting coefficients between the plurality of first weighting coefficients and the plurality of second weighting coefficients; and synthesizing results of multiplications, where the multiplications are performed on the plurality of inputted signals and the plurality of weighting coefficients.
• Still another preferred embodiment according to the present invention relates to a program method. This program includes: inputting a plurality of signals on which a processing is to be performed; switching a plurality of weighting coefficients by which the plurality of inputted signals are multiplied between a plurality of first weighting coefficients and a plurality of second weighting coefficients; giving an instruction of switching the weighting coefficients between the plurality of first weighting coefficients and the plurality of second weighting coefficients in a prescribed interval, where the plurality of signals are inputted in a sequential manner during the interval; and synthesizing results of multiplications, where the multiplications are performed on the plurality of inputted signals and the plurality of weighting coefficients.
• The plurality of first weighting coefficients may be set in a manner that, as results of multiplications by the plurality of inputted signals, a multiplication result corresponding to one signal among the plurality of inputted signals becomes effective. The one signal among the plurality of inputted signals may be a signal having a largest value among the plurality of inputted signals. The plurality of first weighting coefficients may be set by utilizing the plurality of second weighting coefficients which have already been set.
• The receiving method may further include: updating a plurality of third weighting coefficients adaptively based on the plurality of inputted signals; estimating gaps between the plurality of first weighting coefficients and the plurality of third weighting coefficients by performing a correlation processing between at least one of the plurality of inputted signals and a known signal; and generating the plurality of second weighting coefficients by compensating the plurality of third weighting coefficients based on the estimated gaps.
• The signals inputted during the prescribed interval in the sequential manner may include signals having different characteristics. In giving the instruction of switching the weighting coefficients between the first weighting coefficients and the second weighting coefficients, the instruction may be given when it is detected a shift point where the characteristics of the signals change. The plurality of third weighting coefficients updated may be inputted sequentially in giving the instruction of switching the weighting coefficients between the first weighting coefficients and the second weighting coefficients, and the instruction may be given when fluctuation of the plurality of third weighting coefficients converges within a prescribed range.
• It is to be noted that any arbitrary replacement or substitution of the above-described structural components and the steps, expressions replaced or substituted in part or whole between a method and an apparatus as well as addition thereof, and expressions changed to a computer program, recording medium or the like are all effective as and encompassed by the present embodiments.
• Moreover, this summary of the invention does not necessarily describe all necessary features so that the invention may also be sub-combination of these described features.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1 shows a structure of a communication system according to a first embodiment of the present invention.
• FIG. 2 shows a burst format according to the first embodiment of the present invention.
• FIG. 3 shows a burst format according to the first embodiment of the present invention.
• FIG. 4 shows a structure of a receiver according to the first embodiment of the present invention.
• FIG. 5 shows a structure of a first pre-processing unit shown inFIG. 4.
• FIG. 6 shows a structure of the first pre-processing unit shown inFIG. 4.
• FIG. 7 shows a structure of the first pre-processing unit shown inFIG. 4.
• FIG. 8 shows a structure of a timing detection unit shown inFIGS. 5, 6 and7.
• FIG. 9 shows a structure of a rising edge detection unit shown inFIG. 4.
• FIG. 10 shows an operation procedure of the rising edge detection unit shown inFIG. 9.
• FIG. 11 shows a structure of an antenna determination unit shown inFIG. 4.
• FIG. 12 shows a structure of a first weight computation unit shown inFIG. 4.
• FIG. 13 shows a structure of a gap measuring unit shown inFIG. 4.
• FIG. 14 shows a structure of a gap compensating unit shown inFIG. 4.
• FIG. 15 shows a structure of a synthesizing unit shown inFIG. 4.
• FIG. 16 shows a structure of a receiver according to a second embodiment of the present invention.
• FIG. 17 shows a structure of an antenna determination unit shown inFIG. 16.
• FIG. 18 shows a structure of a gap measuring unit shown inFIG. 16.
• FIG. 19 shows a structure of a frequency error estimation unit shown inFIG. 18.
• FIG. 20 shows a structure of a gap measuring unit shown inFIG. 16.
DETAILED DESCRIPTION OF THE INVENTION
• The invention will now be described based on the preferred embodiments, which do not intend to limit the scope of the present invention, but exemplify the invention. All of the features and the combinations thereof described in the embodiment are not necessarily essential to the invention.
FIRST EMBODIMENT
• The first embodiment of the present invention relates to a receiver provided with an adaptive array antenna which receives radio signals with a plurality of antennas as burst signals and synthesizes the received signals with weighting them respectively by different weighting coefficients. The burst signal is composed of a known training signal which is disposed in the head part thereof and a data signal. The receiver, in order to reduce processing delay, synthesizes the received signals by weighting them with the weighting coefficients without scarcely delaying them. The weighting coefficients are updated by the LMS algorithm one after another. As the weighting coefficients in the training signal interval, however, precedently prepared weighting coefficients of an omni antenna pattern are utilized since it is often the case that the weighting coefficients have not converged yet in the initial period of the training signal interval. Weighting coefficients of adaptive array antenna pattern, which are updated by the LMS algorithm, are utilized as the weighting coefficients in the interval of the data signal.
• FIG. 1 shows a communication system including atransmitter100 and areceiver106 according to the first embodiment of the present invention. Thetransmitter100 includes amodulator102, aRF unit104, and anantenna132. Thereceiver106 includes afirst antenna134a,asecond antenna134b,a n-th antenna134n,aRF unit108, asignal processing unit110, and ademodulator112. Here thefirst antenna134a,thesecond antenna134band the n-th antenna134nare generically named antennas134.
• Themodulator102 modulates an information signal to be transmitted and generates the transmission signal (hereinafter one signal included in the transmission signal is also called as a “symbol”). Any arbitrary modulation scheme may be utilized, such as QPSK (Quadri Phase Shift Keying), 16QAM (16 Quadrature Amplitude Modulation), GMSK (Gaussian filtered Minimum Shift Keying). In the following embodiments, examples are described where the QPSK is utilized. Moreover, in a case of a multi carrier communication, thetransmitter100 is provided with the plurality ofmodulators102 or inverse Fourier transform units. In a case of a spectrum spreading communication, themodulator102 is provided with a spreading unit.
• TheRF unit104 transforms the transmission signal into radio frequency signal. A frequency transformation unit, a power amplifier, a frequency oscillator and so forth are included therein.
• Theantenna132 of thetransmitter100 transmits the radio frequency signals. The antenna may have arbitrary directivity and the number of the antennas may also be arbitrary.
• The antennas134 of thereceiver106 receive the radio frequency signals. In this embodiment, the number of the antennas134 is n. When it is described in this embodiment that the receiver has a n-th component thereof, it means that the number of the components provided to thereceiver106 is same as the number of the antennas134, where the first, second, . . . n-th component basically performs same operation in parallel.
• TheRF unit108 transforms the radio frequency signals into baseband received signals300. A frequency oscillator and so forth are provided to theRF unit108. In a case of the multi carrier communication, theRF unit108 is provided with a Fourier transform unit. In a case of the spectrum spreading communication, theRF unit108 is provided with a despreading unit.
• Thesignal processing unit110 synthesizes the baseband received signals330 with respectively weighting by the weighting coefficients and controls each weighting coefficient adaptively.
• Thedemodulator112 demodulates the synthesized signals and performs decision on the transmitted information signal. Thedemodulator112 may also be provided with a delay detection circuit or a carrier recovery circuit for coherent detection.
• FIG. 2 andFIG. 3 show other burst formats respectively utilized in different communication systems corresponding to the communication system shown inFIG. 1. Training signals and data signals included in the burst signals are also shown in those figures.FIG. 2 shows a burst format utilized in a traffic channel of the Personal Handyphone System. A preamble is placed in initial 4 symbols of the burst, which is utilized for timing synchronization. The signals of the preamble and a unique word can serve as a known signal for thesignal processing unit110, therefore thesignal processing unit110 can utilize the preamble and the unique word as the training signal. Data and CRC both following after the preamble and the unique word are unknown for thesignal processing unit110 and correspond to the data signal.
• FIG. 3 shows a burst format utilized in a traffic channel of the IEEE 802.11a, which is one type of wireless LAN (Local Area Network). The IEEE 802.11a employs OFDM (Orthogonal Frequency Division Multiplexing) modulation scheme. In the OFDM modulation scheme, the size of the Fourier transform and the number of the symbols of the guard interval are summated and the summation forms a unit. It is to be noted that this one unit is described as an OFDM symbol in this embodiment. A preamble is placed in initial 4 OFDM symbols of the burst, which is mainly utilized for timing synchronization and carrier recovery. The signals of the preamble can serve as a known signal for thesignal processing unit110, therefore thesignal processing unit110 can utilize the preamble as the training signal. Header and Data both following after the preamble are unknown for thesignal processing unit110 and correspond to the data signal.
• FIG. 4 shows a structure of thereceiver106 shown inFIG. 1. TheRF unit108 includes afirst pre-processing unit114a,asecond pre-processing unit114b,. . . and a n-th pre-processing unit114n,which are generically named pre-processing units114. Thesignal processing unit110 includes: a firstBB input unit116a,a secondBB input unit116b,. . . and a n-thBB input unit116nwhich are generically named BB input units116; asynthesizing unit118; a firstweight computation unit120a,a secondweight computation unit120b,. . . and a n-thweight computation unit120nwhich are generically named weight computation units120; a risingedge detection unit122; acontrol unit124; atraining signal memory126; anantenna determination unit10; an initial weightdata setting unit12; agap measuring unit14, agap compensating unit16; aweight switching unit18. Thedemodulator112 includes asynchronous detection unit20, adecision unit128 and a summingunit130.
• Moreover the signals utilized in thereceiver106 include: a first baseband received signal300a,a second baseband receivedsignal300b,. . . and n-th baseband receivedsignal300nwhich are generically named the baseband received signals300; atraining signal302; acontrol signal306; anerror signal308; a firstcontrol weighting coefficient310a,a secondcontrol weighting coefficient310b,. . . and a n-thcontrol weighting coefficient310nwhich are generically named control weighting coefficients310; anantenna selection signal314; agap error signal316; a first updatedweighting coefficient318a,a second updatedweighting coefficient318b,. . . and a n-th updatedweighting coefficient318nwhich are generically named updated weighting coefficients318; a firstinitial weighting coefficient320a,a secondinitial weighting coefficient320b,. . . and a n-thinitial weighting coefficient320nwhich are generically named initial weighting coefficients320; and afirst weighting coefficient322a,asecond weighting coefficient322b,. . . and a n-th weighting coefficient322nwhich are generically named weighting coefficients322.
• The pre-processing units114 translates the radio frequency signals into the baseband received signals300.
• The risingedge detection unit122 detects the starts of the burst signals which serve as a trigger of the operation of thesignal processing unit110 from the baseband received signals300. The timings of the detected starts of the burst signals are informed to thecontrol unit124. Thecontrol unit124 computes timings when the interval of thetraining signal302 ends, based on the timings of the starts of the burst signals. These timings are notified to each unit as control signals306 in accordance with necessity.
• Theantenna determination unit10 measures the electric power of each baseband received signal300 after the interval of thetraining signal302 is started in order to select one antenna134 to be made effective in the interval of thetraining signal302 and then determines the one baseband received signal300 of which the electric power becomes is largest. Moreover, theantenna determination unit10 outputs this information as theantenna selection signal314.
• The initial weightdata setting unit12 sets the weighting coefficients322 utilized in the interval of thetraining signal302 as the initial weighting coefficients320. The initial weightdata setting unit12 makes only oneinitial weighting coefficient302 effective by setting the value of the one initial weighting coefficient320 as 1 and by setting the values of the other weighting coefficients320 as 0. The one initial weighting coefficient320 to be made effective is decided according to theantenna selection signal314.
• Thetraining signal memory126 stores thetraining signal302 and outputs the training signal in accordance with necessity.
• The weight computation unit120 updates the control weighting coefficients310 based on the baseband received signals300 and after-mentionederror signal308 by the LMS algorithm.
• Thegap measuring unit14, based on the baseband received signals300 and thetraining signal302, estimates the gap between the results of a synthesis processing performed in the after-mentionedsynthesizing unit118, wherein one result is acquired by performing the synthesizing processing on the initial weighting coefficients320 and baseband received signals300 and the other is acquired by performing the synthesizing processing on the control weighting coefficients310 and the baseband received signals300. The synthesis result acquired by utilizing the initial weighting coefficients320 is the baseband received signal300 as it is, which is corresponding to one antenna134. Therefore following expression (1) can be acquired. Here, it is presumed that the one antenna134 is an i-th antenna134.
  x_i(t)=h_iS(t)exp(jΔωt)+n_it  (1)
• Here, hi is the response characteristic of the radio interval, S(t) is the transmission signal, Δω is the frequency offset between the frequency oscillators of thetransmitter100 and thereceiver106, and ni(t) is a noise. On the other hand, a control weighting coefficient310wiupdated from the head region of the burst signal is given by:
  Σh_iw_i=1  (2)
• Here, it is presumed assumed that the control weighting coefficients have already converged sufficiently.
• By performing the synthesis processing based on the ground of the above-described expression (2), following result of the synthesis processing can be acquired.
  y(t)=S(t)exp(jΔωt)+n(t)  (3)
• By comparing the synthesis results shown in (1) and (3), a gap error signal316C is given by:
  C=h_i  (4)
• Thegap compensating unit16 compensates the control weighting coefficients310 with thegap error signal316 and outputs the result of the compensation as the updated weighting coefficients318.
• Theweight switching unit18, based on the instruction of thecontrol signal306, selects the initial weighting coefficients320 in the interval of thetraining signal302 and selects the updated weighting coefficients318 in the interval of the data signal. Then, theweight switching unit18 outputs them as the weighting coefficients322.
• The synthesizingunit118 weights the baseband received signals300 with the weighting coefficients322 and then sums them up.
• Thesynchronous detection unit20 performs synchronous detection on the synthesized signals and also performs a carrier recovery necessary for the synchronous detection.
• Thedecision unit128 decides the transmitted information signal by comparing the signal acquired by the summation to a pre-determined threshold value. The decision may be either hard or soft.
• The summingunit130 generates theerror signal308 based on the difference value between the synchronous detected signal and the decided signal, which is to be utilized in the LMS algorithm in the weight computation units120. In an ideal situation, the error signal becomes zero since the LMS algorithm controls the weighting coefficients310 so that theerror signal308 might become small.
• FIG. 5 toFIG. 7 show various structures of thefirst pre-processing unit114a.Thefirst pre-processing unit114ain thereceiver106 can accept and treat various signals in different communication systems such as shown inFIG. 2 orFIG. 3, therefore thesignal processing unit110 following thereafter can operate ignoring the difference of the communication systems. Thefirst pre-processing unit114ainFIG. 5 is for the single carrier communication system shown inFIG. 2 such as Personal Handyphone System, cellular phone system or the like. Thefirst pre-processing unit114ainFIG. 5 includes afrequency translation unit136, a quasisynchronous detector138, an AGC (Automatic Gain Control)140, anAD conversion unit142, and atiming detection unit144. Thefirst pre-processing unit114ashown inFIG. 6 is for the spectrum spreading communication system such as the W-CDMA (Wideband-Code Division Multiple Access) or the wireless LAN implemented in relation to the IEEE 802.11b. In addition to thefirst pre-processing unit114ashown inFIG. 5, that shown inFIG. 6 further includes adespreading unit172. Thefirst pre-processing unit114ais for the multi carrier communication system shown in FIG.3 such as the IEEE 802.11a or the Hiper LAN/2. In addition to thefirst pre-processing unit114ashown inFIG. 6, that shown inFIG. 7 further includes aFourier transform unit174.
• Thefrequency translation unit136 translates the radio frequency signal into one intermediate frequency signal, a plurality of intermediate frequency signals or other signals. The quasisynchronous detector138 performs quadrature detection on the intermediate frequency signal utilizing a frequency oscillator and generates a baseband analog signal. Since the frequency oscillator included in the quasisynchronous detector138 operates independently from the frequency oscillator provided to thetransmitter100, the frequencies between the two oscillators differ from each other.
• TheAGC140 automatically controls gains so that the amplitude of the baseband analog signal might become an amplitude within the dynamic range of theAD conversion unit142.
• TheAD conversion unit142 converts the baseband analog signal into a digital signal. Sampling interval for converting the baseband analog signal to the digital signal is generally set to be shorter than symbol interval in order to constrict the degradation of the signal. Here, the sampling interval is set to the half of the symbol interval (Hereinafter, the signal digitalized with this sampling interval is referred to as a “high speed digital signal”).
• Thetiming detection unit144 selects a baseband received signal300 of an optimal sampling timing from the high speed digital signals. Alternatively, thetiming detection unit144 generates the baseband received signal300 having the optimal sampling timing by performing a synthesis processing or the like on the high speed digital signals.
• Thedespreading unit172 shown inFIG. 6 performs correlation processing on the baseband received signal300 based on a predetermined code series. TheFourier transform unit174 inFIG. 7 performs the Fourier transform on the baseband received signal300.
• FIG. 8 shows the structure of thetiming detection unit144. Thetiming detection unit144 includes: afirst delay unit146a,asecond delay unit146b,. . . and a (n−1)-th delay unit146n−1which are generically named delay units146; afirst multiplication unit150a,asecond multiplication unit150b,a (n−1)-th multiplication unit150n−1, . . . and a n-th multiplication unit150nwhich are generically named multiplication units150; afirst data memory152a,asecond data memory152b,a (n−1)-th data memory152n−1, . . . a n-th data memory152nwhich are generically named data memories152; a summingunit154; adecision unit156; a mainsignal delay unit158; and a selectingunit160.
• The delay units146 delay the inputted high speed digital signal for the correlation processing. The sampling interval of the high speed digital signal is set to half of the symbol interval. However the delay quantity of the delay units146 is set to the symbol interval, therefore the high speed digital signal150 is outputted from every other delay unit146 to the multiplication units150.
• The data memories152store 1 symbol of each preamble signal for the timing synchronism.
• The multiplication units150 perform multiplications on the high speed digital signals and the preamble signals, and the results thereof are summed up by the summingunit154.
• Thedecision unit156 selects an optimal sampling timing based on the result of the summation. The sampling interval of the high speed digital signal is half of the symbol signal and the interval of the high speed digital signal utilized for the summation is equal to the symbol interval, therefore there are two types of the summation results for every other high speed digital signal corresponding to each shifted sampling timing. Thedecision unit156 compares the two types of the summation results and decides a timing corresponding to larger summation result as the optimal sampling timing. This decision should not necessarily be made by comparing the two types of the summation results once, but may be made by comparing them for several times.
• The mainsignal delay unit158 delays the high speed digital signal until the optimal sampling timing is determined by thedecision unit156.
• The selectingunit160 selects a baseband received signal300 corresponding to the optimal sampling timing from the high speed digital signals. Here one high speed digital signal is selected sequentially from the two successive high digital speed signals.
• FIG. 9 shows the structure of the risingedge detection unit122 included in thesignal processing unit110. The risingedge detection unit122 includes apower computation unit162 and adecision unit164. Thepower computation unit162 computes the received power of each baseband received signal300 and then sums up the received power of each baseband received signal to acquire the whole power of the signals which are received by all the antennas134.
• Thedecision unit164 compares the whole received power of the signals with a predetermined condition and decides that the start of the burst signal is detected when the condition is satisfied.
• FIG. 10 shows the operation of the risingedge detection unit122. Thedecision unit164 sets an internal counter T to zero (S10). Thepower computation unit162 computes the received power from the baseband received signals300 (S12). Thedetermination unit164 compares the received power with a threshold value. When the received power is larger than the threshold value (Y in S14), thedecision unit164 adds 1 to the T (S16). When the T becomes larger than a predetermined value τ (Y in S18), it is decided that the start of the burst signal is detected. The processing described-above is repeated until the start of the burst signal is detected (N in S14, N in S18).
• FIG. 11 shows the structure of theantenna determination unit10. Theantenna determination unit10 includes: a firstlevel measuring unit22a,a secondlevel measuring unit22b,. . . and a n-thlevel measuring unit22nwhich are generically called level measuring units22; and a selectingunit24.
• The level measuring units22 detect the start timing of the burst signal based on thecontrol signal306 and measure the electric power of each baseband received signal300 during prescribed interval from the start timing.
• The selectingunit24 selects the baseband received signal300 which has the largest electric power by comparing the electric power of each baseband received signal300 and then outputs a result as theantenna selection signal314.
• FIG. 12 shows the structure of the firstweight computation unit120a.The firstweight computation unit120aincludes aswitching unit48, acomplex conjugate unit50, a mainsignal delay unit52, amultiplication unit54, a stepsize parameter memory56, amultiplication unit58, a summingunit60, and adelay unit62.
• The switchingunit48 selects thetraining signal302 in the interval of the training signals302 by detecting the start timing of the burst signal and the end timing of interval of thetraining signal302 based on thecontrol signal306 and then selects theerror signal308 in the interval of the data signal.
• The mainsignal delay unit52 delays the first baseband received signal300aso that the first baseband received signal300amight synchronize with the timing detected by the risingedge detection unit122.
• Themultiplication unit54 generates a first multiplication result by multiplying thephase error308 after complex conjugate transform in thecomplex conjugate unit50 by the first baseband received signal300awhich is delayed by the mainsignal delay unit52.
• Themultiplication unit58 generates a second multiplication result by multiplying the first multiplication result by a step size parameter stored in the stepsize parameter memory56. The second multiplication result is fed back by thedelay unit62 and the summingunit60 and added to a new second multiplication result. The result of the summation is then sequentially updated by the LMS algorithm. This summation result is outputted as thefirst weighting coefficient310a.
• FIG. 13 shows the structure of thegap measuring unit14. Thegap measuring unit14 includes acomplex conjugate unit44, a selectingunit64, abuffer unit66 and amultiplication unit68.
• The selectingunit64, based on theantenna selection signal314, selects the baseband received signal300 corresponding to the one initial weighting coefficient320 which has been made effective in the interval of thetraining signal302.
• Thebuffer unit66 detects the start timing of the burst signal based on thecontrol signal306 and outputs the baseband received signal300 at the start timing.
• Themultiplication unit68 multiplies thetraining signal302 after the complex conjugate processing in thecomplex conjugate unit44 by the one baseband received signal300 outputted from thebuffer unit66 and then outputs thegap error signal316. Here, it is presumed that both thetraining signal302 and baseband received signal300 are the head signal of the burst signal.
• FIG. 14 shows the structure of thegap compensating unit16. Thegap compensating unit16 includes afirst multiplication unit70a,asecond multiplication unit70b,. . . and a n-th multiplication unit70nwhich are generically named multiplication units70.
• The multiplication units70 detect the end timing of the interval of thetraining signal302 based on thecontrol signal306. Then the multiplication units70 multiply the control weighting coefficients310 by thegap error signal316 and outputs the updated weighting coefficients318.
• FIG. 15 shows the structure of the synthesizingunit118 which is included in thesignal processing unit110. The synthesizingunit118 includes: afirst delay unit166a,asecond delay unit166b,. . . and a n-th delay unit166nwhich are generically named delay units166; afirst multiplication unit168a,asecond multiplication unit168b,. . . and a n-th multiplication unit168nwhich are generically named multiplication units168; and a summingunit170.
• Since the delay time of the delay units166 is from the detection of the head of the burst signal by the risingedge detection unit122 until setting the weighting coefficients322 by the initial weightdata setting unit12 via theweight switching unit18, the processing delay of the delay units166 can be ignored in general. Therefore, synthesizing processing with less processing delay can be realized.
• The multiplication units168 multiply the baseband received signals300 which are delayed by the delay units166 by the weighting coefficients322. The summingunit170 sums up the whole results of the multiplications by the multiplications units168.
• Hereunder will be described the operation of thereceiver106 having the structure described above. The signals received by the plurality of antennas134 are translated to the baseband received signals300 by the quadrature detection and so forth. When the risingedge detection unit122 detects the starts of the burst signals from the baseband received signals300, the interval of thetraining signal302 is started. At the start timing of the interval of thetraining signal302, theantenna determination unit10 selects the one baseband received signal300. Then the initial weightdata setting unit12 sets the initial weighting coefficients320, where the only initial weighting coefficient320 corresponding to the selected baseband received signal300 is made effective.
• In the interval of thetraining signal302, theweight switching unit18 outputs the initial weighting coefficients320 as the weighting coefficients322 and the synthesizingunit118 sums up the baseband received signals300 weighting them with the weighting coefficients322. Meanwhile, the weight computation units120 update the control weighting coefficients310 by the LMS algorithm. In the interval of the data signal, thegap compensating unit16 compensates the control weighting coefficients310 with thegap error signal316 computed in thegap measuring unit14 and then outputs them as the updated weighting coefficients318. Moreover, theweight switching unit18 outputs the updated weighting coefficients318 as the weighting coefficients322 and the synthesizingunit118 weights the baseband received signals300 with the weighting coefficients322 and sums them up.
• According to the first embodiment, the processing delay can be reduced since the synthesizing processing is performed even in the interval of the training signal regardless of the convergence of the weighting coefficients. Moreover, communications with surrounding radio stations located in the vicinity can be realized since the omni antenna pattern is utilized for the weighting coefficients in the interval of the training signal. The weighting coefficients can be smoothly switched between the omni antenna pattern and the adaptive array antenna pattern.
SECOND EMBODIMENT
• In the second embodiment, same as the first embodiment, received signals are weighted with weighting coefficients and synthesized. The processing delay hardly occurs since the switching is performed between the omni antenna pattern which is precedently prepared and the adaptive array pattern updated by the LMS algorithm. In the first embodiment, the switching of the weighting coefficients between two types is performed in an undifferentiated manner at the timing where the training signal included in the burst signal ends. On the other hand, in the second embodiment, the switching of weighting coefficients between two types is performed adaptively at the timing where the LMS algorithm converges within a predetermined range.
• FIG. 16 shows the structure of thereceiver106 according to the second embodiment. The structure thereof is almost same as the structure of thereceiver106 shown inFIG. 4. However, thereceiver106 shown inFIG. 16 includes afirst convergence information324a,asecond convergence information324b,. . . and a n-th convergence information324nwhich are generically named convergence information324.
• Theweight switching unit18 shown inFIG. 4 performs the switching operation in a manner that the initial weighting coefficient320 is selected in the interval of thetraining signal302 and the updated weighting coefficient is selected in the interval of the data signal, wherein the end timing of the interval of the initial weighting coefficients320 severs as a trigger for theweight switching unit18. On the other hand, theweight switching unit18 utilizes the timing where the control weighting coefficients310 converge in the weight computation units120 (hereinafter this timing is referred to as a “convergence timing”). The convergence timing is generated by thecontrol unit124 when the fluctuation of the control weighting coefficients310 caused by updating them converges within in a range, wherein the range is determined precedently. Alternatively, the convergence timing may be generated by thecontrol unit124 when the updatederror signal308 becomes within a range, wherein the range is predetermined for theerror signal308.
• Thecontrol unit124 notifies the convergence timing to each unit in accordance with the necessity, and each unit performs its assigned processing according to the convergence timing.
• FIG. 17 shows the structure of theantenna determination unit10. Theantenna determination unit10 includes aswitching unit72, alevel measuring unit74, astorage76 and a selectingunit24.
• The switchingunit72 switches the plurality of baseband received signals300 at a prescribed timing and outputs one baseband received signal300. The switching may be performed on the plurality of burst signals.
• Thelevel measuring unit74 measures the electric power of the baseband received signal300 selected by the switchingunit72. Being different from theantenna determination unit10 shown inFIG. 11, the electric power of the plurality of baseband received signals300 is not measured at a time but measured for every baseband received signal300 one by one, therefore the size of an arithmetic circuit for thelevel measuring unit74 can be diminished.
• Thestorage76 stores the computed electric power of the baseband received signal300.
• FIG. 18 shows the structure of thegap measuring unit14. Thegap measuring unit14 shown inFIG. 18 is structured by adding a frequencyerror estimation unit78, aninterval measuring unit80, amultiplication unit82, a complexnumber transformation unit84, acomplex conjugate unit86 and amultiplication unit88 to thegap measuring unit14 shown inFIG. 13.
• In the second embodiment, being different from the first embodiment, the timing where the weight computation units120 start updating the control weight coefficients310 is the head of the long preamble of the burst format shown inFIG. 3. The control weighting coefficient310wiupdated from the head of the long preamble is given by the expression (5) below. Here, it is presumed that the control weighting coefficients310 have converged sufficiently.
  Σh_iw_iexp(JΔωsT)=1  (5)
• Here, sT is the time length of a short preamble interval. By performing the synthesizing processing based on the expression (5), the synthesis result is given by:
  y(t)=S(t)exp(jΔωt)exp(−jΔωsT)+n(t)  (6)
• By comparing these expressions, the gap error signal316C can be expressed as follows.
  C=h_iexp(−jΔωsT)  (7)
• The frequencyerror estimation unit78 estimates a frequency error Δω based on the baseband received signals300. Theinterval measuring unit80 measures the time sT of the short preamble interval based on thetraining signal302.
• Themultiplication unit82 multiplies the frequency error by the time of the short preamble interval and acquires the phase error in the interval of the short preamble. This phase error is transformed to a complex number by the complexnumber transformation unit84 and a complex conjugate processing is performed thereon by thecomplex conjugate unit86.
• Themultiplication unit88 multiplies, by the above-described phase error, the result of the multiplication processing on the one baseband received signal300 and the complexconjugated training signal302, and then generates thegap error signal316.
• FIG. 19 shows the structure of the frequencyerror estimation unit78. The frequencyerror estimation unit78 includes: a first mainsignal delay unit26a,a second mainsignal delay unit26b,. . . and a n-th mainsignal delay unit26nwhich are generically named main signal delay units26; afirst multiplication unit28a,asecond multiplication unit28b,. . . and a n-th multiplication unit28nwhich are generically named multiplication units28; afirst delay unit30a,asecond delay unit30b,. . . and a n-th delay unit30nwhich are generically named delay units30; a firstcomplex conjugate unit32a,a secondcomplex conjugate unit32b,. . . and a n-thcomplex conjugate unit32nwhich are generically named complex conjugate units32; afirst multiplication unit34a,asecond multiplication unit34b,. . . and a n-th multiplication unit34nwhich are generically named multiplication units34; an averagingunit36; aphase transformation unit38; and atraining signal memory42.
• The multiplication units28 acquires a received signal Zi(t) which does not include transmission signal component by multiplying the baseband received signals300 delayed in the main signal delay units26 by thetraining signal302 after the complex conjugate transform. The received signal Zi(t) is given by:$\hspace{1em}\begin{array}{cc}\begin{array}{c}{Z}_i\left(t\right)=x_i\left(t\right)S^{*}\left(t\right)\\ =h_i\exp \left(\mathrm{j\Delta \omega }t\right)\end{array}& \left(8\right)\end{array}$
• Here, it is assumed that a noise is sufficiently small and therefore the noise is ignored.
• The delay units30 and the complex conjugate units32 delay the Zi(t) and then transform the Zi(t) to the complex conjugate. The Zi(t) transformed to the complex conjugate is multiplied by the original Zi(t) in the multiplication units34. The result Ai of the multiplication is given by:$\hspace{1em}\begin{array}{cc}\begin{array}{c}{A}_i\left(t\right)=Z_i\left(t\right)Z_i^{*}\left(t-T\right)\\ =\exp \left(\mathrm{j\Delta \omega }t\right)\end{array}& \left(9\right)\end{array}$
• Here, the delay time of the delay units30 is set to the symbol interval T.
• The averagingunit36 averages the multiplication results corresponding to each antenna. The multiplication results of which the time is shifted may also be utilized.
• Thephase transformation unit38 transforms the averaged multiplication result A to a phase signal B by utilizing an arc tangent ROM.
  B=ΔωT  (10)
• FIG. 20 shows the structure of agap measuring unit14 which is different from thegap measuring unit14 shown inFIG. 18. Thegap measuring unit14 shown inFIG. 20 is structured by adding acounter unit90, amultiplication unit92, a complex number transformation unit94, a summingunit96, a summingunit98 and adivision unit40 to thegap measuring unit14 shown inFIG. 18. In thegap measuring unit14 shown inFIG. 18, the multiplication of the baseband received signals300 by thetraining signal302 is performed only on the head signal of the burst signal. On the other hand, in thegap measuring unit14 shown inFIG. 20, the multiplications are performed during prescribed time and the results thereof are averaged.
• The summingunit98 sums up the results of the multiplications by themultiplications unit96 during prescribed time interval (hereinafter referred to as “averaging time”) in order to average the results of the multiplications of the baseband received signals300 by thetraining signal302.
• Thecounter unit90 counts up the symbol intervals in order to acquire the phase error corresponding to the averaging time based on the frequency error outputted from the frequencyerror estimation unit78. Themultiplication unit92 acquires the phase error corresponding to each counter value by respectively multiplying each counter value by the frequency error. The phase errors are transformed to complex numbers in the complex number transformation unit94 and are summed up in the summingunit96 within the averaging time.
• Thedivision unit40 divides the results of the multiplications summed up by the summingunit98 with the phase errors summed up by the summingunit96. The succeeding processings are same as those of thegap measuring unit14 shown inFIG. 18.
• Hereunder will be described the operation of thereceiver106 having the structure described above. The signals received by the plurality of antennas134 are transformed to the baseband received signals300 by the quadrature detection and so forth. When the risingedge detection unit122 detects the start timings of the burst signals from the baseband received signals300, the interval of thetraining signal302 is started. At the start timing of the interval of thetraining signal302, theantenna determination unit10 selects the one baseband received signal300 and the initial weightdata setting unit12 sets the initial weighting coefficients320 among which only the one initial weighting coefficient320 corresponding to the selected baseband received signal300 is made effective. Thereafter, theweight switching unit18 outputs the initial weighting coefficients320 as the weighting coefficients322 and the synthesizingunit118 weights the baseband received signals300 with the weighting coefficients322 and sums them up.
• Meanwhile, the weight computation units120 update the control weighting coefficients310 by the LMS algorithm. When the control weighting coefficients310 converge within the prescribed range, thegap compensating unit16 compensates the control weighting coefficients310 with thegap error signal316 computed in thegap measuring unit14 according to the instruction from thecontrol unit124 and then outputs them as the updated weighting coefficients318. Moreover,weight switching unit18 outputs the updated weighting coefficients318 as the weighting coefficients322 and the synthesizingunit118 weights the baseband received signals300 with the weighting coefficients322 and sums them up.
• According to the second embodiment, the synthesis processing is performed regardless of the convergence of the weighting coefficients even in the interval of the training signal. Therefore, the processing delay can be reduced. Moreover, in the case that the adaptive algorithm converges during the training signal interval, the response characteristic can be improved by reflecting it to the weighting coefficients. This is because the switching of the weighting coefficients between two types is performed based on the convergence timing of the adaptive algorithm.
• Although the present invention has been described by way of exemplary embodiments, it should be understood that many changes and substitutions may be made by those skilled in the art without departing from the scope of the present invention which is defined by the appended claims.
• In the embodiments, the initial weightdata setting unit12 sets the effective value for the initial weighting coefficient320 for the one baseband received signal300 selected by theantenna determination unit10, which has the largest electric power, and theunit12 sets the value which is not effective for the other initial weighting coefficients320. The initial weighting coefficients320, however, do not necessarily need to be set based on the electric power. For example, one fixed initial weighting coefficient320 may be set to the effective value and the other initial weighting coefficients320 may be set to the value that is not effective. In that case, theantenna determination unit10 becomes unnecessary.
• In the embodiments, the initial weightdata setting unit12 sets the effective value for the initial weighting coefficient320 for the one baseband received signal300 selected by theantenna determination unit10, which has the largest electric power, and theunit12 sets the value which is not effective for the other initial weighting coefficients320. It is, however, not necessarily required to set the weighting of the omni antenna pattern for the initial weighting coefficients320. For example, the setting may be performed on the updated weighting coefficients318 or the control weighting coefficients310 which are utilized in the already received burst signal. When the fluctuation of the radio transmission environment is small, it is estimated that this setting will not cause a serious degradation of the response characteristic.
• In the embodiments, the weight computation units120 utilize the LMS algorithm as the adaptive algorithm. However, another algorithm such as the RLS algorithm may be utilized. Moreover, the weighting coefficients may not be updated. That is, it is sufficient if the adaptive algorithm is selected in accordance with the estimated radio transmission environment, the size of arithmetic circuits or the like.
• In the first embodiment, the risingedge detection unit122 computes the electric power of the baseband received signals300 and detects the rising edge of the burst signal based on the computation result. The rising edge of the burst signal may be, however, detected by implementing another structure. For example, the rising edge may be detected by a matched filter which is shown as the structure of thetiming detection unit144. That is, it is sufficient if the rising edge of the burst signal is detected accurately.
• In the first embodiment, the training signal interval is the time where the initial weighting coefficients320 are changed into the weighting coefficients322. However, the time does not need to be limited to the interval of the training signal. For example, the time may be shorter than the interval of the training signal. That is, the time can be set according to the length of the interval of the training signal and to the required estimation accuracy.
• In the second embodiment, the delay time of the delay units30 which are included in the frequencyerror estimating unit78 is set to 1 symbol. The delay time, however, is not limited to 1 symbol. For example, the delay time may be 2 symbols or may be symbols in the interval between the start and end of the training signal. That is, it is sufficient if an optimum delay time of the delay units30 is decided in accordance with the stability of the frequency oscillator and with the required accuracy of the frequency offset estimation.

Claims (9)

1.-24. (canceled)

25. A method for processing a received signal, the method comprising:

receiving, from a transmitter, a first signal from a first antenna terminal and a second signal from a second antenna terminal wherein the first signal and the second signal each contains at least one packet having a training interval and a data interval;

having the first signal and the second signal pass through a signal detector and then using a first set of weighting factors;

performing an adaptive algorithm on the first and second signals in a period during which the first set of weighting factors is used, to determine a second set of weighting factors; and

applying the second set of weighting factors to an array synthesizer, after the period during which the first set of weighting factors is used.

26. A method according toclaim 25, further receiving, from the transmitter, a third signal from a third antenna terminal wherein the third signal contains at least one packet having a training interval and a data interval.

27. A method according toclaim 25, further comprising:

measuring a channel response of the first signal or the second signal; and

using the measured channel response to generate a gap error signal for correcting a phase gap in an output signal of the array synthesizer.

28. A method according toclaim 27, further comprising updating the second set of weighting factors, based on the gap error signal.

29. A receiver comprising:

an input unit configured to receive, from a transmitter, a first signal from a first antenna terminal and a second antenna terminal from a second antenna terminal wherein the first signal and the second signal each contains at least one packet having a training interval and a data interval;

an initial weight data setting unit configured to have the first signal and the second signal pass through a signal detector and then to use a first set of weighting factors;

a weight computation unit configured to perform an adaptive algorithm on the first and second signals in a period during which the first set of weighting factors is used, to determine a second set of weighting factors; and

a weight switching unit configured to apply the second set of weighting factors to an array synthesizer, after the period during which the first set of weighting factors is used.

30. A receiver according toclaim 29, wherein the input unit is configured to receive, from the transmitter, a third signal from a third antenna terminal wherein the third signal contains at least one packet having a training interval and a data interval.

31. A receiver according toclaim 29, further comprising a gap measuring unit configured to measure a channel response of the first signal or the second signal and to use the measured channel response to generate a gap error signal for correcting a phase gap in an output signal of the array synthesizer.

32. A receiver according toclaim 31, further comprising a gap compensating unit configured to update the second set of weighting factors, based on the gap error signal.

Images