JP3838261B2 - Communication apparatus and communication method - Google Patents

Description

  The present invention relates to a communication device and a communication method of an ultra-wide band (UWB) communication system in which an information signal is configured using an impulse signal sequence having a very short period of about several hundred picoseconds and this signal sequence is transmitted and received. In particular, the present invention relates to a communication apparatus and a communication method that perform transmission / reception with pulses that avoid a spectrum problem in an ultra-wideband communication system.

  More specifically, the present invention relates to a communication device and a communication method of an ultra-wideband communication system that has enhanced interference wave capability in combination with DS-SS (direct spectrum spread).

  By connecting multiple computers to form a LAN (Local Area Network), information such as files and data can be shared, peripheral devices such as printers can be shared, e-mails, data and content can be transferred, etc. Exchange information.

  Recently, wireless LAN has attracted attention. According to this type of wireless LAN, since most of the wired cables can be omitted in a work space such as an office, a communication terminal such as a personal computer (PC) can be moved relatively easily. In addition, the demand for wireless LAN systems has increased significantly as the speed and price of wireless LAN systems have decreased. In particular, recently, in order to establish a small-scale wireless network between a plurality of electronic devices existing around a person and perform information communication, introduction of a personal area network (PAN) has been studied.

  Recently, a wireless LAN system to which an SS (Spread Spectrum) system is applied has been put into practical use. For applications such as PAN, a UWB (Ultra Wide Band) transmission method using the SS method has been proposed.

  The DS (Direct Spread: Direct Spread) method, which is a type of SS method, transmits on the transmitting side by spreading an occupied band by multiplying an information signal by a random code sequence called a PN (Pseudo Noise) code. On the receiving side, the received spread information signal is despread by multiplying it by the PN code to reproduce the information signal.

  In the UWB transmission system, an information signal is configured by using a DS-UWB system in which the spreading speed of a DS information signal is increased to the limit, and an impulse signal sequence having a very short period of about several hundred picoseconds. There are two types of impulse-UWB systems that transmit and receive columns.

  Although the DS-UWB system can control the spectrum according to the PN code rate, there is a problem that power consumption tends to increase because the logic circuit needs to be operated at a high speed on the order of GHz. On the other hand, the impulse-UWB system can be configured by a combination of a pulse generator and a low-speed logic circuit, so that there is a merit that current consumption can be reduced, but there is a problem that it is difficult to control the spectrum with the pulse generator.

  In addition, both systems can realize high-speed data transmission by performing transmission / reception by spreading in an ultra-high frequency band of 3 GHz to 10 GHz, for example. The occupied bandwidth is a band on the order of GHz such that a value obtained by dividing the occupied bandwidth by the center frequency (for example, 1 GHz to 10 GHz) is approximately 1, and is a so-called W-CDMA or cdma2000 system, SS or OFDM ( Compared to a bandwidth normally used in a wireless LAN using an Orthogonal Frequency Division Multiplexing) method, the bandwidth is extremely wide.

Conventionally, a Gaussian monocycle pulse (Gaussian Mono Cycle Pulse) has been used as an impulse signal for UWB transmission. Here, in order to examine the linearity requirement of the apparatus in pulse generation, a comparison is made between a Gaussian monocycle pulse and a rectangular waveform monocycle pulse. As an example, consider a monocycle pulse with a rectangular waveform with T _p = 200 [ps] and 1 [V]. A Gaussian monocycle pulse was considered by the following equation. However, the constants 3.16 and 3.3 in the equation are obtained as values having a spectrum equivalent to that of the rectangular waveform monocycle pulse.

  FIG. 1 shows a time waveform at this time. Further, FIG. 2 compares the frequency characteristics of the power spectral density of these monocycle pulses. However, the power spectrum density [W / Hz = J] is shown when the pulse of this voltage is transmitted at 1 [pulse / s] and driven at 50 [ohm].

  As can be seen from FIG. 2, if 100 mega pulses per second, the power density is further increased by 80 dB from this value. Since the power density of the peak of the pulse shown here is about -211 dBJ, at 100 mega pulses per second, it is exactly −41.3 [dBm / MHz] = − 131.3 [dBW / Hz], which is the regulation of FCC. = DBJ].

  Therefore, the following can be concluded.

(1) A monocycle pulse with a Gaussian waveform and a monocycle pulse with a rectangular waveform are almost the same in the transmission band.
(2) A Gaussian waveform monocycle pulse has a higher peak voltage than a rectangular waveform, requires linearity, and is difficult to process including power amplification.

  In conventional UWB communications, monocycle pulses have been used. In FIG. 3, the frequency characteristic of the power spectral density shown in FIG. 2 is displayed as a true number instead of the decibel. There is no particular need to be an exact number, but the energy is shown linearly and is often intuitively convenient.

  Here, there are the following two requirements for the spectrum.

(1) According to the FCC spectrum mask regulations, radiation below 3 GHz cannot be emitted.
(2) The 4.9 to 5.3 GHz band has a 5 GHz wireless LAN, which should be avoided.

  Moreover, the following matters can be considered when looking at the power spectrum of the linear display.

(1) If the above requirements are not observed, only about half of the power [3 dB] can be transmitted.
(2) The pulse waveform is expected to be greatly disturbed, and only about half of the energy passes through the matched filter on the receiving side.
(3) A loss of 6 [dB] or more occurs in total.

  FIG. 4 shows a configuration example (conventional example) of a receiver in the ultra-wideband communication system. The receiver configuration shown is similar to a DS-SS (direct spread spectrum) receiver.

  In the illustrated example, it is assumed that the VCO oscillates at the same frequency as the pulse period.

The receiver generates a pulse train whose data is all zero according to the timing of the VCO, generates a total of three waveforms each shifted by half of the pulse width T _p (T _p / 2), and multiplies the received signal To do.

  When the pulse position is detected, a time when the pulse timing coincides is reached by intentionally shifting the VCO frequency slightly (sliding correlation).

  When the pulse timings coincide with each other, the energy of the multiplication result becomes high, so that the pulse position can be detected.

  At the stage where the pulse position is detected, the frequency of the VCO that is intentionally shifted slightly is returned to the correct frequency, and at the same time, the tracking operation is performed to maintain this timing.

The energy obtained by multiplying the waveform shifted by ± T _p / 2 with respect to the center (Puncture) is obtained, and the subtracted one detects a positive / negative value corresponding to the positive / negative of the pulse position error. It is used as a control voltage for pulse position tracking through a filter.

However, in the case of the receiver configuration as shown in FIG. 4, it is necessary to divide the signal path into three branches and to have three systems of circuits after multiplication, which complicates the circuit.

  Further, it is necessary to change the frequency at the time of searching and at the time of tracking, and the synchronization establishment time becomes longer due to the time required for this switching.

  In addition, when detecting the pulse position, it is necessary to detect that the energy is increased multiple times in order to correctly detect the pulse position in a noisy environment. After intentionally shifting the frequency to be negligibly small and averaging the energy that is increased over a plurality of times, it is necessary to detect the pulse position, resulting in a longer synchronization establishment time.

  Further, the mechanism for shifting the frequency or performing tracking is constituted by an analog circuit, but the circuit is complicated and affected by variations and it is difficult to stabilize the operation.

  Further, since the energy value is used at the time of pulse position detection or tracking, the S / N deteriorates and the characteristics deteriorate.

Nikkei Electronics March 11, 2002 issue "Ultra Wideband, a wireless revolutionary child that raises birth" P.55-66

  An object of the present invention is to provide an excellent communication apparatus and communication method capable of performing transmission / reception with pulses that avoid a spectrum problem in an ultra-wideband communication system.

  The present invention provides, among other things, a high-speed propagation path measurement method in an ultra-wideband communication system combined with DS-SS (direct spectrum spread) in order to enhance interference wave resistance capability, that is, a high-speed detection method of pulse position, amplitude, and phase. Propose about.

The present invention has been made in consideration of the above problems, and the first aspect thereof is a communication method for performing ultra-wideband communication,
A spread processing unit that directly spreads the transmission data, and
An RF transmission processing unit is provided that generates a signal composed of pulses obtained by modulating a baseband pulse generated at a chip rate that is a fraction of an integer of the carrier frequency at the center of the transmission band with the carrier frequency. And
Prior to transmission of the transmission data, a training pattern consisting of a short code that repeats the same pattern for each symbol length is transmitted.
It is the communication system characterized by this.

A second aspect of the present invention is a communication method for performing ultra-wideband communication,
An RF reception processing unit that receives an input signal composed of a pulse obtained by modulating a baseband pulse generated at a chip rate that is a fraction of an integer of the carrier frequency at the center of the transmission band with the carrier frequency; ,
Providing a training unit by short code spreading that repeats the same pattern for each symbol length, and using the training unit, the length of the short code is used as a measurement section, and N periods of the carrier are used as time resolution for measurement to measure transmission path propagation. A propagation measuring unit for performing
It is the communication system characterized by this.

  Here, the chip timing is controlled by the time resolution of the measurement, and the value obtained by A / D converting the input signal at the chip timing is despread by the short code at all points given by the time resolution of the measurement section. Thus, a measured value can be obtained.

  Moreover, the speed of measurement can be increased by using a plurality of despreading units and giving appropriate despreading timing to each.

  Also, the period of the training section is set to a length necessary for measuring all points. As a result, the measurement process of obtaining the measurement value by despreading the value obtained by A / D conversion of the input signal at the chip timing by the short code is executed a plurality of times, and these measurement values are converted into complex numbers at each point. By averaging by cumulative addition or moving addition, the S / N of the measurement can be improved. Then, the place where the energy is the largest among the measured values obtained by performing cumulative addition or moving addition in a complex number for each point is detected as the pulse position, and the received symbol is demodulated at the detected pulse position, and the demodulation is performed. The result is correlated with the symbol pattern of the training unit, and the symbol position in the training unit is detected. The final transmission path measurement value can be obtained by removing the influence of the symbol pattern of the training unit included in the measurement value of each point.

  Therefore, according to the communication method according to the present invention, it is possible to estimate the transmission path at a higher speed than the method using the sliding correlation.

  In addition, high-speed transmission path estimation enables high-speed detection of symbols or pulse positions. Even in UWB communication where carrier sense is inherently difficult, the same operation as carrier sense such as CSMA is possible, and a multiple access scheme such as CSMA can be adopted.

  For example, the most recent measurement values can be obtained at any time by always storing the measurement values for a predetermined point in the past for the number of times necessary for averaging, and moving average the number of times. Such an operation makes it possible to detect a signal that cannot be located anywhere, and can be used as a propagation measurement method even in a multiple access system such as CSMA that communicates at random time using bursty.

  In addition, it is possible to grasp the state of multipath using the propagation measurement result according to the present invention and obtain parameters necessary for RAKE reception based on the state of the multipath from signals received through a plurality of paths. it can. For example, when RAKE combining is performed, it is not possible to combine signals having different chip phases. Therefore, multipath timing is obtained based on the measurement result of the propagation path, and only signals having the same chip phase are combined in each path. What is necessary is just to do. By adopting RAKE reception, the signal of each separated path is aligned in time and phase by using the effect of time decomposition by despreading of direct spectrum spreading, and the desired signal power dispersed in time is made effective. Can be synthesized.

  Moreover, since SINR can be estimated based on the final propagation measurement value by the propagation measurement which concerns on this invention, a highly accurate Link Adaptation is realizable using this. In direct spread spectrum communication, various bit rates can be realized by changing the spreading factor. If the noise level and interference level are small with respect to the signal level, the spreading rate can be reduced and the bit rate can be increased. In the opposite case, the spreading rate can be increased and the bit rate can be lowered.

  Further, according to the present invention, a stable receiving circuit can be configured by correcting the receiving pulse position by changing the carrier frequency division ratio for obtaining the chip timing. That is, if only carrier phase correction is performed, I / Q obtained by despreading may be digitally corrected, but the pulse position shifts, so that the received pulse can be changed by switching the carrier frequency division ratio. An analog correction is performed to correct the position. Since the carrier and chip timing are synchronized, tracking the phase of the carrier will simultaneously track the chip timing.

  In addition, according to the present invention, in addition to the I and Q components that are normally obtained by orthogonal detection of the received signal, the result of 45 degree phase rotation by analog addition and subtraction of the I and Q outputs is A / D converted. Thus, even when an A / D converter with a small number of bits is used, good phase tracking is possible, and the demodulation result can be improved.

  Particularly in the case of A / D conversion with a small number of bits, the resolution may not be sufficient when the spreading factor is lowered and transmission is performed at a high bit rate. In such a case, the I and Q components obtained by orthogonal detection of the normal reception signal are added to the signal obtained by A / D conversion at the chip timing, and the I and Q components are added and subtracted in an analog manner to 45 degrees. The signals I ′ and Q ′ that have undergone A / D conversion after the rotation of may be input. Looking at the results of four systems of A / D conversion, even if a 1-bit A / D converter is used, the phase direction is quantized in eight steps, so that even in high-speed transmission, the phase is The resolution of correction can be increased.

  Note that a training pattern in which a short code is repeated for a predetermined number of symbols may be added to the training pattern which is inverted to add a training length twice as long. In this case, a measurement process for obtaining a measurement value by despreading the value obtained by A / D conversion of the input signal at the chip timing by the short code is executed a plurality of times, and the measurement value is alternately added or subtracted for each point. You just have to do it.

  The FCC rule requires that the radiation spectrum power is measured with a 1 MHz bandwidth, and the power density is smoothed regardless of which point is measured. By doubling the training length, the spectrum is almost the same even if measured at any 1 MHz, and a maximum amount of power can be radiated as long as the FCC rule is satisfied. On the contrary, it is obvious that the shorter the training sequence period, the easier the measurement (or synchronization detection) to be performed. However, if this is used, the total power needs to be reduced to satisfy the FCC rule.

  ADVANTAGE OF THE INVENTION According to this invention, the outstanding communication apparatus and communication method in which a high-speed transmission line estimation is possible by the UWB transmission which combined direct spread spectrum can be provided.

  Further, according to the present invention, it is possible to detect symbols or pulse positions at high speed by high-speed transmission path estimation. As a result, even in UWB communication where carrier sense is inherently difficult, the same operation as carrier sense such as CSMA can be performed, and a multiple access scheme such as CSMA can be adopted.

  In addition, it is possible to grasp the state of multipath using the propagation measurement result according to the present invention and obtain parameters necessary for RAKE reception based on the state of the multipath from signals received through a plurality of paths. it can. For example, when RAKE combining is performed, it is not possible to combine signals having different chip phases. Therefore, multipath timing is obtained based on the measurement result of the propagation path, and only signals having the same chip phase are combined in each path. It can be. By adopting RAKE reception, the signal of each separated path is aligned in time and phase by using the effect of time decomposition by despreading of direct spectrum spreading, and the desired signal power dispersed in time is made effective. Can be synthesized.

  Moreover, since SINR can be estimated based on the final propagation measurement value by the propagation measurement which concerns on this invention, a highly accurate Link Adaptation is realizable using this.

  Further, according to the present invention, a stable receiving circuit can be configured by correcting the receiving pulse position by changing the carrier frequency division ratio for obtaining the chip timing.

  In addition, according to the present invention, in addition to the I and Q components that are normally obtained by orthogonal detection of the received signal, the result of 45 degree phase rotation by analog addition and subtraction of the I and Q outputs is A / D converted. Thus, even when an A / D converter with a small number of bits is used, good phase tracking is possible, and the demodulation result can be improved.

  Other objects, features, and advantages of the present invention will become apparent from more detailed description based on embodiments of the present invention described later and the accompanying drawings.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

A. In a PAN in which direct spread spectrum utility UWB is used, a base station or the like does not centrally manage frequency resources, but each ubiquitous radio station grasps the state of resource usage of surrounding radio stations, It is considered preferable to use the frequency by distributed control from the viewpoint of spatial frequency reuse. In particular, in the case of UWB, the requirement is particularly high because an ultra-wide band is used (described above) and spatial frequency reuse by frequency division cannot be performed.

  In this case, if spread spectrum is used, even if there is communication using the same frequency in the vicinity, the required C / I for enabling normal communication can be set to a level lower than 0 dB. In other words, even if another person's signal is detected at the same level as his / her own signal, he / she can still communicate. In particular, in the case of UWB, the bandwidth that is originally occupied is much wider than the required bit rate, which is easy to use.

  The bandwidth in UWB is determined by the pulse width regardless of the pulse rate. The wide bandwidth due to the narrow pulse width is a kind of spread spectrum, but cannot be applied to the story here. This is because the anti-interference wave capability holds if the pulse position happens to be different, but if the pulse position happens to coincide, the anti-interference wave capability cannot be expected, in other words, the luck of time. Therefore, it is preferable to perform effective spectrum spreading by direct spreading in addition to spreading by a small pulse width.

  In the following, a propagation path measurement method and the like in an ultra-wideband communication system combined with DS-SS (direct spectrum spread) in order to enhance interference wave resistance is proposed.

  FIG. 5 schematically illustrates the configuration of the ultra-wideband transmission device according to the present embodiment. FIG. 6 shows the operating characteristics of this RF transmission function module.

  First, the transmission data is directly spectrum spread in the spread processing unit and passed to the RF function module as binary data. This RF transmission function module has an oscillator with a center frequency of 4 GHz. By dividing the oscillation frequency by 4, a 1 GHz rectangular pulse is generated as a baseband pulse. The pulse width is 1/1 [GHz] = 1000 picoseconds.

  Then, this is modulated (multiplied) by a carrier wave of 4 GHz to generate a 4-cycle pulse. The 4-cycle pulse is amplified by a power amplifier (PA), subjected to RF filtering, and then transmitted from the antenna to the transmission line.

  Here, the pulse rate is 1 GHz obtained by dividing the carrier frequency by 4, and the spreading rate is Nss, and the symbol rate is fsym = 4 [GHz] / Nss. In order to estimate the propagation path at the receiver, a training pattern including a short code that repeats the same pattern for each symbol is transmitted prior to transmission of transmission data. Details of the training pattern will be described later.

  FIG. 7 schematically shows a functional configuration of an RF reception function module corresponding to the RF transmission function module shown in FIG.

  On the receiving side, the received signal is orthogonally detected with a carrier wave having the same frequency as that at the time of transmission, a baseband pulse train is detected, despreading and decoding processing are performed, and received data (binary data) is extracted.

  By combining FIGS. 5 and 7, a transceiver for UWB communication combined with direct spread spectrum is configured (see FIG. 8).

  FIG. 9 shows an example of a direct diffusion circuit. As shown in the figure, a spread information signal can be obtained by multiplying a transmission symbol (TxSymbol) by a spread chip sequence.

  First, the symbol rate is obtained by accurately setting the spread chip rate obtained by dividing the carrier frequency by 4 to 1 / Nss. In this embodiment, Nss = 10.

  Next, a spread sequence is generated based on the chip timing and the symbol timing.

  Then, a spread information signal (TxSignal) is obtained by multiplying the spread sequence by the transmission symbol. In the case of a digital circuit, this multiplication is completely equivalent to performing exclusive OR (EX-OR).

  The spreading sequence is composed of what is called a short code that repeats the same pattern for each symbol length.

  FIG. 10 shows how transmission symbols are spread by a spreading sequence consisting of short codes. As shown in the figure, the spreading sequence has a pattern in which chips having a width obtained by dividing a symbol length by a spreading factor Nss (= 10) are connected by the transmission symbol length, and the same pattern is repeated for each transmission symbol. Then, a transmission signal is obtained by multiplying the transmission symbol and the spreading sequence.

  In other cases, spreading may be performed in a sequence longer than a symbol length called a long code. These are distinguished and used depending on the application. A long code is preferable as the original direct spread spectrum, but a short code is effective when detecting synchronization.

  FIG. 11 shows how transmission symbols are spread with a spreading sequence consisting of long codes. As shown in the figure, the spreading sequence consists of a pattern in which chips having a width obtained by dividing a symbol length by a spreading factor Nss (= 10) are connected longer than the symbol length, and this pattern is repeatedly multiplied by a transmission symbol to transmit A signal is generated.

  On the other hand, FIG. 12 conceptually shows functional blocks for performing the basic operation of despreading. FIG. 13 shows an operation for despreading the received signal.

  First, the same spreading sequence is generated by the same mechanism as the transmission side. Then, the received signal (RxSignal) is multiplied by the same pattern ((De) SpreadSequence) as the spreading sequence used in transmission (After Multiple), and this is integrated in the symbol interval. At the end of the symbol period, this value (Accumulate Signal) is held as an integrated value, which becomes a received symbol. At this time, the integration part is sequentially cleared to prepare for the next integration.

However, with an actual receiver,
(1) Carrier phase uncertainty due to carrier phase rotation (2) Chip timing uncertainty (3) Spread sequence phase and symbol timing uncertainty.

  For this reason, it is not possible to demodulate only with the basic operation described above with reference to FIGS. Therefore, the receiving side despreading unit is actually configured by a circuit as shown in FIG.

  First, the received signal is quadrature detected with a carrier wave having the same frequency as that at the time of transmission, and each value of I (RxSignal I) and Q (RxSignal Q) of the detected received signal is A / D converted. And Next, these received signals are multiplied in a complex manner by the same pattern as the spreading sequence used for transmission. Then, these multiplication results are integrated in a complex manner in the symbol interval to obtain received symbols (RxSymbol I, RxSymbol Q) in a complex number. Next, the received symbol (RxSymbol) is obtained by performing carrier phase compensation on these received symbols.

  If the phase of the spreading sequence generated in synchronization with the chip timing and the chip timing is not known, this must be found first. Since it is difficult to find these when long codes are used, short codes are usually used. FIG. 15 shows a configuration of a digital signal processing portion of the receiver including a functional configuration for performing sliding correction and tracking when the carrier and the chip rate are synchronized. FIG. 16 shows the relationship between the input signal, spreading sequence, and correlation output when performing sliding correction. Hereinafter, the sliding correlation and the tracking loop will be described. However, each value of I (RxSignal I) and Q (RxSignal Q) of the received signal is A / D converted and performed as digital processing.

(1) A short code spreading sequence is periodically generated by a spreading sequence generator. The reference oscillator VCO is controlled so that the frequency is slightly shifted (slightly earlier) than the period of the received signal, and the chip timing is intentionally shifted with a sufficiently small shift.

(2) If such an operation for sliding correlation is continued, the timing at which both the chip timing and the short code spreading sequence coincide with each other will be reached. If they match, the signal after despreading (that is, the multiplication result of the short code spread sequence and the received signal) has a large correlation output power (I × I + Q × Q) (see Sync Detect in FIG. 16). Therefore, it is possible to know the coincidence, that is, the pulse position detection by detecting this power (Peak Detect).

(3) Return the frequency of the reference oscillator, which has been set to a slightly shifted frequency, to the correct frequency at the coincidence stage.

(4) Although the chip timing can be obtained by the processing so far, the phase of the carrier is not necessarily obtained. Therefore, by multiplying the output signals of I / Q (I × Q), ± 1 symbol information included in the received signal is removed to obtain carrier phase error information (Carrier Pulse Error).

(5) The carrier phase is synchronized by forming a control loop so that the carrier phase error is zero by feeding back the carrier phase error information to the reference oscillator.

(6) Since the carrier and chip timing are obtained from the same reference oscillator (baseband pulses are obtained at a time interval of 1 / integer of the carrier frequency), they are synchronized with each other. The carrier phase is synchronized, and at the same time, the chip timing is tracked.

  However, the following problems remain in the configuration in which the sliding correlation described above is performed.

● Sliding correlation requires slow sliding, so it takes a relatively long time to synchronize.
When trying to improve S / N by averaging at low S / N in timing detection, the information for pulse position detection uses the power value of the correlation output, and S / N improvement by averaging The degree is low.
When generating a frequency shift to obtain a sliding correlation and returning it to the time of synchronization detection, frequency switching by a PLL synthesizer is used. However, since this frequency settling time is long, it takes a relatively long time for synchronization. .

  A method for finding the position, amplitude, and phase of a pulse at a higher speed in a configuration in which the carrier and the chip rate are synchronized will be described below.

B. Transmission channel propagation measurement (Coherent Channel Measurement)
If the characteristics of the propagation path including the propagation delay amount can be measured, it can be applied to timing synchronization, carrier phase synchronization, RAKE reception configuration, and Link Adaptation. Hereinafter, this method will be described in detail.

  In this embodiment, a training unit is provided prior to the communication main body, and transmission path propagation measurement is performed at this portion. First, the following parameters are used as conditions. However, this parameter is a parameter suitable for satisfying the requirement of the FCC rule described above, and does not limit the present invention. FIG. 17 illustrates one cycle of training symbols.

● 1 giga chip per second, 1000 picosecond chip period ● 50 mega symbols per second, 20 nanosecond symbol period (spreading rate Nss = 20)
● The measurement period is the range of 20 nanoseconds of symbol period (ie, the length of the short code), and the resolution of measurement is 80 points in the period of 250 picoseconds (N periods of the carrier).・ Pattern repetition (480 nanosecond cycle) is used as a training pattern ● Frequency reference with sufficient accuracy is provided for both transmission and reception (480 nanosecond cycle)

  FIG. 18 shows a device configuration of the propagation path measurement unit. This propagation path measurement unit receives as input a value obtained by A / D converting each value of I (RxSignal I) and Q (RxSignal Q) of a received signal detected by quadrature detection at a chip timing (1 GHz). Consists of.

  As illustrated, the propagation path measurement unit divides the center frequency of 4 GHz to generate a chip timing, and a spreading code generation unit (Spread Code Generation) generates a spreading code based on the chip timing. ) Includes a control unit (Control & Calculation) and a despreading unit (DespreadBlock) that despreads the digitally processed received signal. One feature of this propagation path measurement unit is that four despreading units are provided for high-speed measurement. Of course, it is sufficient to provide a single despreading unit unless high speed is required.

  The spread code generation unit generates a short code consisting of 20 chips at four symbol timings according to a command from the control unit (see later and FIG. 20) (short codes are shifted at the same timing), A short code spreading sequence is passed to each despreading unit together with the symbol timing. Each despreading unit performs despreading using this signal.

  Basically, a chip rate of 1 [GHz] is generated by dividing the center frequency of 4 [GHz] by four. However, in order to adjust the timing of 250 picoseconds, which is the time resolution of measurement, the frequency can be divided by 5 by a command from the control unit.

The controller is responsible for measuring transmission path propagation.
(1) Four measurement points within the chip interval (2) 20 chips within the symbol interval (3) Grasping which point of the 24 symbols corresponding to the length of the training pattern is to be measured and spreading instructions for that The data is sent to the code generator and the frequency divider.

  FIG. 19 shows the configuration of the despreading unit in the propagation path measurement circuit shown in FIG.

  The despreading unit receives an A / D-converted value of I (RxSignal I) and Q (RxSignal Q) of the received signal detected by quadrature detection at the chip timing (1 GHz). Then, these received signals are multiplied in a complex number by a despreading sequence (the same pattern as the spreading sequence used for transmission) supplied from the spreading code generator. Then, these multiplication results are integrated in a complex manner in the symbol interval to obtain received symbols (RxSymbol I, RxSymbol Q) in a complex number. Then, a correlation output can be obtained by the sum of squares of received symbols after despreading (I × I + Q × Q).

  FIG. 20 illustrates a basic operation procedure in the propagation measurement unit. The procedure for measuring the propagation path will be described in detail with reference to FIG.

  Symbols 0 to 23 represent one cycle of training symbols (480 nanoseconds), and one symbol is composed of chip 0 to chip 19 (20 nanoseconds). Furthermore, 0 to 4 in-chip phase numbers are provided within one chip (1000 picoseconds). Further, when the carrier frequency is 4 GHz, one cycle of the carrier is 250 picoseconds, which is the time resolution of measurement.

(1) At the first stage of measurement, the control unit sets the combinations of “symbol” − “chip” − “in-chip phase” to 0-0-0 and 0-1-0 for the four despreading units, respectively. , 0-2-0, 0-3-0 is specified using a short code starting from 0-3. The four despreading units finish all of these correlations until reaching 1-2-0. The measured data is stored in the memory.

(2) The next measurement waits for 8 cycles of 4 [GHz] (that is, exactly 2 chips), and then the symbol-chip-in-chip phase combinations are 1-4-0 and 1-5-0, respectively. , 1-6-0, 1-7-0 is specified using a short code starting from 1-7-0.

(3) Combination of symbol-chip-in-chip phase in which similar operations are shown below 2-8-0, 2-9-0, 2-10-0, 2-11-0
3-12-0, 3-13-0, 3-14-0, 3-15-0
4-16-0, 4-17-0, 4-18-0, 4-19-0
Start at the point starting with. The measurement of the point where the in-chip phase is “0” ends at 5-18-0.

(4) Next, the control unit waits for 9 cycles of 4 GHz (that is, 2 + 1/4 chip), and then the symbol-chip-in-chip phase combinations are 6-0-1, 6-1-1, respectively. It is specified to take a correlation value using a short code starting from 6-2-1 and 6-3-1.

(5) Hereinafter, symbol-chip-in-chip phase combinations 7-4-1, 7-5-1, 7-6-1, 7-7-1 shown below by the same operation as in the procedure (3).
8-8-1, 8-9-1, 8-10-1, 8-11-1
9-12-1, 9-13-1, 9-14-1, 9-15-1
10-16-1, 10-17-1, 10-18-1, 10-19-1
Start at the point starting with.

(6) Subsequently, in order to measure the point (2) in the chip, the same operation as (4) and (5) is performed.

(7) Subsequently, in order to measure the point (3) in the chip, the same operation as (4) and (5) is performed. As described above, all the 80 points to be measured can be measured by the time the symbol-chip-in-chip phase combination reaches 23-18-3. By this time, the time corresponding to 24 symbols, which is just the training cycle, has elapsed.

(8) Average the measured values in order to measure the channel state at a higher S / N. FIG. 21 shows a circuit configuration for averaging the measurement results. In the example shown in FIG. 20, after waiting for 5 cycles of 4 GHz (that is, 1 + 1/4 chip), measurement starts again from the same measurement point. Also, the same training symbols should be transmitted. The above operations (1) to (7) are performed once more, and the S / N is improved by adding to the previous measurement value.

(9) The operation of (8) is repeated (for example, 10 times) until sufficient S / N is obtained.

  Here, as a new assumption, it is assumed that the training that is actually received starts, for example, at a timing of 8-4-2 when the symbol-chip-in-chip phase combination is started.

(10) FIG. 22 shows the result of adding the measured values at the measured 80 points, that is, the measurement result of the propagation path response (Measured Channel Response (Power)). Comparing the measured 80 point measured power, the highest power should be detected where the symbol-chip-in-chip phase combination is 13-4-2. Thereby, it can be detected that the received symbol starts from the timing of x-4-2. In other words, the detection of the chip timing and the phase of the spreading sequence is completed.

(11) However, in the procedure (10), it is not yet known which of the 24 training symbols has been received when the symbol-chip-in-chip phase combination is 13-4-2. Therefore, several (for example, 10 symbols) symbols are received and detected (demodulated) with a short code starting from the timing of x-4-2. By comparing the detected symbols with the training pattern, it is possible to determine which portion of the training 24 symbols is received. In the example shown in FIG. 22, it can be seen that the symbol at the time when the combination of the symbol-chip-in-chip phase is 13-4-2 is TS06.

(12) Since it can be determined which of the 24 training symbols has been measured at other measurement points, the 80-point measurement value is obtained by multiplying each 80-point measurement value by ± 1 of the training. Remove the effects of training patterns from Thereby, a propagation characteristic of 80 points with a resolution of 250 picoseconds can be finally obtained in a section of 20 nanoseconds.

  By the way, in the above description, a symbol pattern (480 nanosecond period) obtained by repeating a short code of 20 chips having a chip period of 1000 picoseconds only 24 times is used as a training pattern (see FIG. 17). In addition, a training length consisting of a period of 48 symbols in total can be obtained by adding an inverted version of the training after 24 symbols. FIG. 23 exemplifies a 48-symbol training pattern in which 24 symbols are trained and then inverted.

  In this way, when the preamble length is doubled, the same operation as described above is performed until the propagation path measurement procedures (1) to (7). In the procedures (8) to (9), when the measured values are averaged in order to measure the channel state at a higher S / N, addition and subtraction are alternately performed. FIG. 24 shows a circuit configuration for averaging the measurement results.

  According to the FCC rule, radiation spectrum power is measured at 1 MHz. In this embodiment, by adding an inverted version of a 24-symbol training pattern, the total period of 48 symbols is 960 nanoseconds (about 1 microsecond) in time, and 1 MHz in terms of spectrum. Even when measured with, the spectrum is almost the same, and it is possible to radiate as much power as possible within a range that satisfies the FCC rules. On the contrary, it is obvious that the shorter the period, the easier the measurement (or synchronization detection) is performed. However, if this is used, it is necessary to reduce the total power in order to satisfy the FCC rule.

  However, in this embodiment, the training length is set to 960 nanoseconds by adding an inverted symbol pattern. However, the training length may be set to about 960 nanoseconds by using other methods.

  Next, a propagation measurement method when the training location is not known will be described.

  The number of times required to average the measured values is 10 sets. In the above description of the basic operation of the propagation measurement, the addition is performed 10 times from the state where the accumulated value is zero.

  If you don't know where the training is, you can always store all 10 sets of past 80 points of measurement in a memory, etc., and move the 10 sets of moving averages to obtain the most recent measurement at any time. Obtainable. FIG. 25 shows the configuration of a circuit that averages measured values by moving average.

  Such an operation makes it possible to detect a signal that cannot be located anywhere, and can be used as a propagation measurement method even in a multiple access system such as CSMA (Carrier Sense Multiple Access) that communicates at random times at bursty times. It becomes like this.

C. Since the bit rate of the chip timing tracking method UWB is high, it is advantageous to shorten the transmission burst to 200 [us] or the like. In this case, it can be said that there is almost no variation in propagation characteristics. Therefore, the measurement value obtained by the propagation measurement (Coherent Channel Measurement) in the previous section is effective in the burst.

  However, even if the frequency reference possessed by the transmission / reception has sufficient accuracy, it is about 2 [ppm] TCXO, so the maximum frequency error between transmission and reception must be 4 [ppm]. If there is an error of 4 [ppm], a burst of 200 [us] results in about 200 [us] × 4 [ppm] = 800 [ps], and only timing and carrier phase tracking must be performed. However, in this case of correcting the deviation due to the frequency error, it is considered that the deviation direction is one direction and the deviation speed is constant.

  Immediately after performing the propagation measurement, the chip timing detected with a resolution of 250 picoseconds is also effective, and the phase of the complex number of the measured value is the phase of the carrier.

  Also, since the carrier and chip timing are synchronized, tracking the phase of the carrier will simultaneously track the chip timing. Hereinafter, a method for tracking the phase of the carrier will be described.

  If only carrier phase correction is performed, the phase of I / Q obtained by despreading may be digitally corrected (described later).

  However, in this case, since the pulse position is shifted, it is necessary to perform some analog correction. In the present embodiment, it is considered that correction is performed to move the timing back and forth in units of 250 picoseconds by controlling the frequency division of 4 GHz. As shown in FIG. 18, this is realized by dividing the 4 GHz carrier by 1/3 or 1/5 while it is normally divided by 1/4. FIG. 26 shows a functional configuration for shifting the chip timing back and forth in units of 250 picoseconds.

  If this is viewed in terms of the carrier phase, an error of ± 360 [deg] is obtained. This part performs digital phase correction. FIG. 27 shows a state where the 4 GHz 1/4 frequency divider is set to 1/3 or 1/5 only once and the phase is shifted by ± 360 degrees.

  If the frequency on the receiving side is fast, the phase shifts in the positive direction as time passes. If there is a deviation of +180 [deg] or more, the chip timing is returned by -250 picoseconds. Specifically, in the example shown in FIG. 28, it is observed that the state shown in (a) is shifted in the plus direction as shown in (b). In such a case, as shown in (c), it is possible to reduce the frequency of 4 GHz to 1/5 only once. As a result, the carrier phase is pulled back to −180 [deg], and the phase shift returns to 0 [deg] as time elapses from here.

  Conversely, when the frequency on the receiving side is slow, the phase shifts in the negative direction as time passes. If it is shifted by -180 [deg] or more, the chip timing is returned by +250 picoseconds. Specifically, in the example shown in FIG. 28, as shown in (d), the frequency division of 4 GHz is possible only by 1/3. As a result, the carrier phase is advanced to +180 [deg], and the phase shift returns to 0 [deg] as time elapses from here.

  In this way, in the method of correcting the carrier phase by moving the chip timing back and forth (shifted), the pulse position is allowed to be shifted to ± 125 picoseconds at worst. Since the pulse length is 1000 picoseconds, degradation due to this amount is considered to be slight. In the following, it is considered to perform a finer correction in order to further reduce deterioration.

  FIG. 29 shows a circuit configuration for performing pulse phase shift with fine resolution. The circuit shown in the figure can control a phase of 250 picoseconds or less using the following cosine theorem.

    COS (A + B) = COS (A) COS (B) -SIN (A) SIN (B)

  SIN (θ) and COS (θ) are obtained by referring to the detected pulse phase θ in the sine / cosine table, and these are respectively input to a phase conversion circuit composed of a multiplier. On the other hand, the center frequency 4 GHz output from the oscillator is input to the sine and cosine phase conversion circuits with a phase difference of 90 deg. And the difference of the output of each phase conversion circuit is taken.

  Further, by controlling the oscillator as a VCO, the phase can be controlled in an analog manner, that is, continuously. However, it is difficult to obtain a definite phase shift amount in a feedforward manner.

  Since the communication method according to the present embodiment is BPSK (Binary Phase Shift Keying), it is possible to detect a chip timing phase error using a circuit as shown in FIG.

Here, it is assumed that the phase error is φ. When the squares of the complex numbers I and Q are obtained, the uncertainty of the BPSK data can be removed. That is, I + jQ is squared, the value of I ² + Q ² + 2I × Q is averaged, and the angle φ is obtained by halving this declination. As a result, the phase point becomes 2φ as shown in FIG. At this stage, averaging is performed to improve the S / N, and then this phase is calculated and ½ is obtained by halving.

  When the phase error φ is obtained, digital phase compensation of chip timing can be performed by a phase rotation circuit as shown in FIG.

  Referring to the chip timing phase error φ detected by the phase error detection circuit shown in FIG. 30 in the sine / cosine table, SIN (φ) and COS (φ) are obtained, and these are respectively phase-converted by multipliers. The received symbol is taken out by inputting to the circuit, multiplying it with the I and Q components of the received symbol, and adding these.

D. Multipath fading can be cited as one of the problems when transmitting and receiving radio signals applied to RAKE reception . This is a phenomenon in which a communication signal is reflected on a building or other object to reach the receiving side through a different route, and a reception signal is disturbed by radio waves arriving from different directions interfering with each other.

  RAKE reception means that a plurality of radio waves are received, and a desired signal is separated by despreading processing from a reception signal on which a plurality of delayed waves are superimposed by a multipath propagation path, and the dispersed signal power is collected into one. That is, using the time-resolving effect of despreading that direct spectrum spreading has, the signals of the separated paths are synthesized with the same time and phase (for example, the maximum ratio is obtained by weighting according to the S / N ratio of the path). Synthesized). According to RAKE reception, desired signal power dispersed in time can be effectively combined.

  FIG. 33 shows a state of propagation characteristics measured by the above-described propagation channel measurement (Coherent Channel Measurement) in a multipath environment.

  In the example shown, it can be seen that there are two paths at 4-2 and 10-3, ie, 6250 picoseconds apart. In addition, since it is known what amplitude and phase these paths are, it is possible to RAKE-synthesize them.

  On the other hand, the above-described propagation path measurement unit (see FIG. 18) is configured to be equipped with four despreading units for speeding up the measurement. However, in the apparatus configuration shown in FIG. 15, since the A / D conversion processing of the received signal performed at 1 giga sample per second is one set of I / Q, when the RAKE synthesis is performed, those with different chip phases are used. It cannot be synthesized. That is, only those having the same chip phase are targeted for synthesis. Therefore, in the example shown in FIG. 33, the 4-2 timing and the 10-2 timing are synthesized.

  If it is desired to synthesize different chip phases, it is necessary to prepare a plurality of A / D converters and perform A / D conversion processing at different chip timings.

  FIG. 34 shows a device configuration of the propagation path measurement unit combined with RAKE reception.

  As illustrated, the propagation path measurement unit divides the center frequency of 4 GHz to generate a chip timing, and a spreading code generation unit (Spread Code Generation) generates a spreading code based on the chip timing. ) Is equipped with a RAKE reception processing unit (RAKE Combining) in addition to a control unit (Control & Calculation) and a despreading unit (DespreadBlock) that despreads the digitally processed received signal.

  As with the apparatus configuration example shown in FIG. 18, four despreading units are provided for high-speed measurement. The RAKE reception processing unit includes four amplifiers and a timing compensation unit (AMP & Timing Compensation) corresponding to each despreading unit, and has a configuration in which these outputs are RAKE combined (SUM).

  The control unit designates the multipath timing obtained by the propagation path measurement for the spreading code generation unit and each despreading unit.

  The spreading code generation unit generates a short code consisting of 20 chips at four symbol timings according to a command from the control unit (see the above and FIG. 20) (short codes are shifted at the same timing), A short code spreading sequence is passed to each despreading unit together with the symbol timing. Each despreading unit performs despreading using this signal.

  The despread signals are supplied to corresponding amplifiers and timing compensation units, respectively.

  The control unit designates the (complex) amplitude and delay value of the transmission path in each path to each amplifier and timing compensation unit.

  FIG. 35 shows the configuration of the amplifier and the timing compensation unit. The amplifier and the timing compensator perform (complex) multiplication on the input signal to adjust the amplitude and phase to further adjust the timing in order to perform maximum ratio synthesis. In this case, since it is BPSK, only the real part may be output.

  Outputs from the amplifiers and the timing compensation unit are added by a synthesis unit (SUM), and RAKE reception is completed.

E. Link Adaptation
In the direct spread spectrum communication according to the present invention, various bit rates can be realized by changing the spreading factor.

  If the noise level and interference level are small with respect to the signal level, the spreading rate can be reduced to increase the bit rate, and vice versa, the spreading rate can be increased to lower the bit rate. In the embodiment of the present invention described above, variations as shown in the following table are conceivable. In Table 1, diffusion rate 1 is synonymous with non-diffusion.

  Furthermore, depending on the encoding method, there are variations in required SINR (signal to interference and noise power ratio) and bit rate.

  By performing the propagation measurement as described above, it is possible to know the state of the propagation path, and whether the signal level in the communication and the other, that is, the noise level and the interference level (whether or not the above-described RAKE reception is performed). I can know.

  Here, taking as an example a case where information is transmitted from user A to user B, Link Adaptation will be described with reference to FIG.

(1) Prior to the net transmission of information from A to B, a training sequence is transmitted from A to B. On the B side, SINR is measured using a training sequence.

(2) This measurement result is transmitted from B to A. A receives the SINR, and determines the bit rate (spreading rate, encoding method) in the net information transmission from A to B.

(3) In net information transmission from A to B, send at this bit rate.

  Alternatively, B may not only measure the SINR based on the result of the propagation measurement, but may further determine the bit rate (spreading rate, encoding method, etc.) based on the SINR and return this to A. Good. The procedure in this case will be described with reference to FIG.

(1) Prior to the net transmission of information from A to B, a training sequence is transmitted from A to B. On the B side, SINR is measured using a training sequence.

(2) B further determines the bit rate (spreading rate, encoding method) in net information transmission from A to B based on the measured SINR. B then transmits this determination to A.

(3) In net information transmission from A to B, send at this bit rate.

  For example, in the RTS / CTS process in an autonomous distributed network method such as CSMA with RTS / CTS (Request to Send / Clear to Send), it is a transmission node prior to transmission of net information. A transmits RTS, and if B which is a receiving node receives this RTS and can receive data, CTS is returned as a response. If A can receive the CTS, the subsequent communication from A to B is executed.

  Messages such as RTS and CTS are also sent to the bursty. Therefore, according to the above-described embodiment, it is possible to incorporate Link Adaptation by adding a training sequence to a burst.

F. According to the device structure of the propagation measurement according to A / D converter this embodiment, the received signal quadrature detection is made in the carrier of the same frequency as the time of transmission, the detected reception signals I (RxSignal I) and Q (RxSignal Q ) Are A / D converted at 1 GHz chip timing as input and thereafter digital processing.

  For this reason, the A / D converter used requires a sample rate of 1 GHz, and it is difficult to realize multi-bit A / D at this speed. Therefore, consider adopting a small-bit A / D converter including 1-bit A / D.

  FIG. 38 shows a circuit configuration example of a 1-bit A / D converter. The analog output signal from the RF receiver is input to the amplifier, and is output to the D latch when the signal level exceeds a predetermined value. The D latch outputs this input signal as an A / D signal in synchronization with the chip timing of 1 GHz.

FIG. 39 shows a circuit configuration example of a 2-bit A / D converter. The analog output signal from the RF receiver is input to each amplifier that differentially amplifies between + V _th , ground, and −V _th , and is output to each D latch when the potential difference exceeds a predetermined value. . Each D latch outputs this input signal in synchronization with 1 GHz chip timing, and the decoder decodes these outputs and outputs them as a 2-bit A / D signal.

  In the UWB communication according to the present embodiment, since the direct spread spectrum method is combined, even if 1-bit A / D conversion is performed, a resolution of several bits can be obtained after despreading.

  Particularly in the case of A / D conversion with a small number of bits, the resolution may not be sufficient when the spreading factor is lowered and transmission is performed at a high bit rate. In such a case, the I and Q components normally obtained by orthogonal detection of the received signal are added to the signal obtained by A / D conversion at the chip timing, and the I and Q components are added to and subtracted from each other in an analog manner to rotate 45 degrees The signals I ′ and Q ′ that have been A / D converted after the execution may be input. FIG. 40 shows a circuit configuration example of an A / D converter that detects eight phases in this way. In this case, an A / D converter having a resolution of 8 phases can be used for phase tracking and demodulation.

  As shown in FIG. 40, when the result of A / D conversion of four systems is seen, even if a 1-bit A / D converter is used, the phase direction is quantized in 8 steps, so that high speed is achieved. Even during transmission, the phase correction resolution can be increased.

  The present invention has been described in detail above with reference to specific embodiments. However, it is obvious that those skilled in the art can make modifications and substitutions of the embodiment without departing from the gist of the present invention. That is, the present invention has been disclosed in the form of exemplification, and the contents described in the present specification should not be interpreted in a limited manner. In order to determine the gist of the present invention, the description of the scope of claims should be considered.

It is the figure which showed the time waveform of the Gaussian-shaped monocycle pulse. It is the figure which showed the frequency characteristic of the power spectral density of a Gaussian shape and a rectangular monocycle pulse. It is the figure which showed the frequency characteristic of the power spectral density of a Gaussian shape and a rectangular monocycle pulse. It is the figure which showed the structural example (conventional example) of the receiver in an ultra wide band communication system. It is the figure which showed typically the function structure of the RF transmission function module of the communication apparatus which concerns on one Embodiment of this invention. 6 is a chart showing operating characteristics of the RF transmission function module shown in FIG. 5. FIG. 6 is a diagram schematically illustrating a functional configuration of an RF reception function module corresponding to the RF transmission function module illustrated in FIG. 5. It is the figure which showed typically the structure of the transmitter / receiver for UWB communication combined with the direct spread spectrum. It is the figure which showed the circuit example of direct spreading | diffusion. It is the chart which showed a mode that the transmission symbol was spread | diffused with the spreading | diffusion sequence which consists of a short code. It is the figure which showed a mode that the transmission symbol was spread | diffused with the spreading | diffusion sequence which consists of long codes. It is the figure which showed notionally the functional block for performing the basic operation | movement of despreading. It is the chart which showed the operation | movement for despreading a received signal. It is the figure which showed the actual circuit structural example of the de-spreading part. It is the figure which showed the structure of the digital signal processing part of the receiver containing the function structure for performing a sliding correction | amendment and tracking. It is the chart which showed a mode that sliding correction | amendment was performed. It is the figure which illustrated one cycle of the training symbol. It is the figure which showed the apparatus structure of the propagation path measurement part. It is the figure which showed the structure of the de-spreading part in the propagation path measuring circuit shown in FIG. It is the figure which showed the procedure of the basic operation | movement in a propagation measurement part. It is the figure which showed the circuit structure for averaging the measurement result in a propagation measurement part. It is the figure which showed the addition result of the measured value in 80 points measured. It is the figure which illustrated the training pattern of 48 symbols formed by adding what inverted this after training of 24 symbols. It is the figure which showed the circuit structure for averaging a measurement result by adding or subtracting a measurement value alternately. It is the figure which showed the structure of the circuit which averages a measured value by a moving average. It is the figure which showed the function structure which shifts chip | tip timing back and forth in a unit of 250 picoseconds. It is the figure which showed a mode that the 1/4 frequency divider of 4 GHz was made into 1/3 or 1/5 only once, and the phase was pulled back only ± 360deg. It is the figure which showed a mode that the phase correction of a carrier was performed digitally. It is the figure which showed the circuit structure which performs the pulse phase shift of a fine resolution. FIG. 3 is a diagram illustrating a configuration of a circuit that detects a phase error of chip timing. It is the figure which showed a mode that the square of the complex numbers I and Q was calculated | required and the uncertainty of the data of BPSK was removed. It is the figure which showed the structure of the phase rotation circuit. It is the figure which showed the mode of the propagation characteristic measured in the multipath environment. It is the figure which showed the apparatus structure of the propagation path measurement part combined with RAKE reception. It is the figure which showed the structure of an amplifier and a timing compensation part. It is a figure for demonstrating the procedure of Link Adaptation performed based on the SINR estimated on the receiving side based on the result of propagation measurement on the receiving side. It is a figure for demonstrating the procedure of Link Adaptation performed based on the SINR estimated on the receiving side based on the result of propagation measurement on the receiving side. It is the figure which showed the circuit structural example of the 1-bit A / D converter. It is the figure which showed the circuit structural example of the 2-bit A / D converter. It is the figure which showed the circuit structural example of the A / D converter which detects 8 phases.

Claims (2)

A communication device for performing ultra-wideband communication,
A frequency generator that generates a carrier frequency that is the center of the transmission band;
A spread processing unit that directly spreads the transmission data at a chip rate obtained by fixedly dividing the carrier frequency ;
An RF transmission processing unit that generates a rectangular baseband pulse by modulating the spread transmission data , and further generates a signal including a pulse obtained by modulating the baseband pulse at the carrier frequency. ,
Prior to transmission of the transmission data, a training pattern consisting of a short code that repeats the same pattern for each symbol length is transmitted.
A communication device. A communication method for performing ultra-wideband communication,
A frequency generation step for generating a carrier frequency to be the center of the transmission band;
A spread processing step for directly spectrum-spreading transmission data at a chip rate obtained by fixedly dividing the carrier frequency ;
RF transmission processing steps for generating a rectangular baseband pulse by modulating the spread transmission data and generating a signal composed of pulses obtained by further modulating the baseband pulse at the carrier frequency ,
Prior to transmission of the transmission data, a training pattern consisting of a short code that repeats the same pattern for each symbol length is transmitted.
A communication method characterized by the above.

Images