JP2009089146A - Communication apparatus, communication system, communication method, communication program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a receiver for efficiently tracking a channel change by achieving, inexpensively, a configuration for speedily calculating modulation precision corresponding to a new channel status in a case where the status of a communication medium (channel) changes. <P>SOLUTION: In a case where a change in a channel status is recognized, a receiving terminal 1200 requests to a transmitting terminal 1000 transmission of a channel estimation packet and calculates from the received channel estimation packet an average and variance of a transfer function at that time in parallel in association with each other. The above calculation work is sequentially carried out each time the channel estimation packet is received, results are accumulated and an amplitude of a signal and noise is estimated from the average value and the variance at a time when the channel estimation packets of a predetermined number are received. On the basis of a result, the transmitting terminal 1000 is notified about modulation precision suitable for the channel status at that time. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a communication device, a communication system, a communication method, a communication program, and a recording medium on which the communication program is recorded that requires high-speed channel estimation.

  An example of a general network configuration for performing data transfer is shown in FIG. Here, the transmission side communication apparatus (hereinafter also referred to as “transmission terminal”) 1000 performs data transfer to the reception side communication apparatus (hereinafter also referred to as “reception terminal 1200”) 1200.

<Description of QoS>
Recently, there is an increasing demand for streaming large-capacity moving image data such as MPEG2-TS (Transport Stream). When transmitting such streaming data, real-time performance is required for communication. That is, an expiration date is determined for a QoS (Quality Of Service) frame constituting the streaming data, and it is necessary to transmit within the expiration date.

  FIG. 12 is a diagram schematically illustrating packet sequences of an example in which transmission of a QoS frame is successful and an example in which transmission is unsuccessful.

  In the successful case (upper part in FIG. 12), the frame 5 has failed to be transmitted when it is first transmitted, and has been successfully transmitted when retransmission is performed. Since the time of this retransmission is before the expiration date of the frame 5, the transmission of the frame 5 is successful.

  On the other hand, in the failure case (lower part in FIG. 12), the frame 5 has failed to be transmitted both at the first transmission and at the first retransmission. Then, before the second retransmission is performed thereafter, the expiration date of frame 5 has elapsed. In this case, the QoS frame 5 that could not be transmitted within the validity period cannot be used and becomes invalid (frame loss), and the video based on the moving image data is disturbed on the receiving terminal 1200 side. Therefore, when performing retransmission to compensate for transmission errors, it is important that retransmission be successful within the validity period of each frame.

  In the above description, the expiration date is described as a time constraint, but is not limited to time, and may be the upper limit of the number of retransmissions, for example.

<Description of HPAV standard>
Next, an outline of generation of transmission data of the HPAV (HomePlug Audio Video) standard for PLC (power line communication) will be described using the HPAV standard format shown in FIG.

  The HPAV standard is divided into two layers, MAC (Media Access Control) and PHY (physical layer), similarly to a wireless LAN (Local Area Network) and Ethernet.

  Data is input to the MAC layer as a stream of frames. The stream of input frames is divided into 512-byte units called “segments”.

  A unit in which data is transferred from the MAC layer to the physical layer is a long MPDU (MAC Protocol Data Unit), followed by a 16-byte AVFC (Audio Video Frame Control), and then n PBs (PHY Blocks). AVFC is a header for various information in the PB portion. The PB includes a header H, a PBB (PB Body), and an error check (C). The header H is 4-byte information for the PBB. The PBB is the content obtained by encrypting the Segment. C is a 4-byte error check code for H and PBB. In this case, the PB is 520 bytes, but may be 136 bytes. If the frame does not enter the segment in the last PB, the PAD is put in the remaining part to match the length of the segment.

  In the PHY layer, Long MPDU is modulated. First, a known signal is transmitted on the transmitting side and the receiving side called Preamble. Next, the AVFC is modulated into an OFDM (Orthogonal Frequency Division Multiplexing) symbol with a predetermined modulation accuracy. The subsequent n PBs are modulated into M OFDM symbols, and the number of OFDM symbols depends on the number of PBs and the transmission rate for modulating the PBs. The OFDM and transmission rate will be described next. Preamble is used to recognize Long MPDU and demodulate AVFC. In this specification, a signal handled by the physical layer is referred to as a packet.

<Description of multi-carrier data transfer>
Returning to FIG. 11, first, the transmission terminal 1000 generates and transmits a signal (TX) to be transmitted. The transmitted TX signal is received as an RX signal by the receiving terminal 1200 via the communication medium (channel) 1100. In this case, the signal is deformed in the channel 1100, and when the transfer function is represented by H, the relationship of the following expression (1) is established.

RX = TX * H + n (1)
Here, n is noise.

  In recent years, the multi-carrier method is frequently used as a method for performing data transfer. One example of this multicarrier scheme is OFDM. FIG. 14 shows an example of an OFDM signal. In the figure, the horizontal axis represents time, the vertical axis represents amplitude, and the third axis (oblique axis) represents frequency. The OFDM signal is transmitted in units of OFDM symbols. N carriers are assigned to one OFDM symbol, and each carrier corresponds to one frequency. Each carrier carries a part of transmitted data.

  When data is allocated to each carrier, it is necessary to determine how many bits of data are allocated to that carrier. Although this method will be described later, the carrier may be a complex number.

  FIG. 15 shows an example of 16QAM (Quadrature Amplitude Modulation) modulation accuracy in which 4 bits are allocated to one complex carrier. Here, I is a real part and Q is an imaginary part. By assigning 2 bits to each of I and Q, a total of 4 bits are assigned to the entire carrier. In this case, the relationship between the bit value and I or Q is defined as shown in Table 1 below.

従って、選択肢数は２⁴＝１６個である。図１５には、これら全ての選択肢が描かれている。

Therefore, the number of options is 2 ⁴ = 16. FIG. 15 depicts all these options.

  Here, for example, when “0010” is transferred, assuming that the upper 2 bits are normally assigned to the real part I and the lower 2 bits are assigned to the imaginary part Q, the carrier value from Table 1 is “00” is −3 of the real part. , “10” is the imaginary part +1, that is, −3 + j, where j = SQRT (−1). That is, it is -3 for the I axis and +1 for the Q axis.

  Thus, data transfer is performed by assigning M bits of data (M = 4 in the case of 16QAM) to each carrier.

  In addition, when the average of the absolute values is obtained over a plurality of OFDM symbols of the transmission signal TX, an OFDM signal is normally generated so that the average of the absolute values is constant for each carrier as shown in FIG. Is done. That is, in the case of 16QAM shown in FIG. 15, there are 16 options, but the average of absolute values is constant in any situation.

  Next, when the above equation (1) is rewritten for each carrier, the following equation (2) is obtained.

RX (i) = TX (i) * H (i) + n (i) (2)
Here, i is a carrier number.

  FIG. 17 shows an example of the relationship between the transmission signal TX and the reception signal RX after passing through the channel. It can be seen that the RX signal is deformed by the transfer function H compared to the transmission signal TX.

  As described above, the signal transmitted from the transmission terminal 1000 is not received as it is. Therefore, the receiving terminal 1200 needs to return the received signal to its original form in order to correctly demodulate the received signal. For this reason, the receiving terminal 1200 normally performs channel estimation processing for estimating H. The channel estimation processing method will be described later. If the estimation result is H ′, the receiving terminal 1200 performs channel correction processing using the H ′ as follows.

TX (i) ′ = 1 / H ′ (i) * RX (i)
= 1 / H '(i) * (TX (i) * H (i) + n (i))
= TX (i) + n '(i) (3)
By performing channel correction in this way, TX ′ close to the original TX can be obtained by the receiving terminal 1200. As the estimated H ′ is closer to H, a signal closer to the transmission signal TX can be restored.

  FIG. 18A shows an example in which the influence of noise is added to the signal RX shown in FIG. Here, the solid line is the average value of the RX signal (the average value of the channel response), and the dotted line is the noise fluctuation width. As shown in the figure, the fluctuation width of the noise is not constant on the frequency axis. In this example, the swing width is large on the low frequency side and the swing width is small on the high frequency side.

  The signal shown in FIG. 18B shows a signal TX ′ after channel correction with respect to the signal RX shown in FIG. Again, the solid line is the average value and the dotted line is the noise amplitude. At the frequency of the signal RX where the signal amplitude is small and the noise is large (for example, the portion indicated by the symbol A in the figure), the noise amount of TX ′ is large and the signal amplitude is large and the noise amount is small (for example, FIG. In the portion (indicated by reference symbol B), the noise amount of TX ′ is small. As described above, after the correction, the case of 16QAM of the signal TX described in FIG. 15 is described on the assumption that there is no noise. Therefore, the carrier values I and Q are just 3, 1, −1. , -3. On the other hand, FIG. 19 shows an example of 16QAM of the signal TX in consideration of noise, and FIG. 19A shows an example when the amount of noise of the signal of the receiving terminal 1200 is small. Here, the area of the noise amount of each value of the carrier is represented by ●. As shown in FIG. 19A, when the amount of noise is small, each region is sufficiently independent, so that communication is sufficiently possible.

  FIG. 19B shows an example in the case where the amount of noise of the signal of the receiving terminal 1200 is very close. Each area touches the adjacent area, but communication is possible by performing error correction processing.

  On the other hand, FIG. 19C shows a case where the noise amount of the signal of the receiving terminal 1200 is too large. In this case, since each area overlaps with the adjacent area in many parts, the signal is not restored even if error correction processing is performed. Therefore, in such a case, it is necessary to use a modulation accuracy lower than 16QAM. That is, in this case, QPSK (Quadrature Phase Shift Keying) (equal to 4QAM), that is, a modulation accuracy in which 2 bits are allocated to one carrier may be used.

  Note that the example shown in FIG. 19A can be communicated without any problem, but here, 64 QAM, which is one more, that is, 6 bits can be allocated to one carrier. In other words, using 16QAM with this carrier results in wasted bandwidth.

  As described above, the receiving terminal 1200 can obtain the optimum modulation accuracy by estimating the carrier strength and the noise amount, that is, the SNR (Signal To Noise Ratio). Therefore, it is possible to communicate at the fastest transmission rate by informing the transmission terminal 1000 of this.

<Description of Method for Estimating Channel Transfer Function H>
In order to estimate the transfer function H, a known signal is usually transmitted by both the transmitting terminal 1000 and the receiving terminal 1200. In the case of HPAV, the preamble is a known signal on the transmission side and the reception side, but only one is transmitted for one long MPDU. Therefore, in HPAV, a channel estimation packet for channel estimation is used. The PB of this packet is known by the transmitting terminal 1000 and the receiving terminal 1200 and is composed of a plurality of OFDM symbols. Since a plurality of OFDM symbols are used for one Long MPDU, channel estimation can be performed more efficiently for the preamble. This will be described next.

If X (i) is the OFDM symbol of the channel estimation packet transmitted by the transmission terminal 1000, this is the reception terminal 1200:
Y (i) = X (i) * H (i) + n (i)
It becomes. Since the receiving terminal 1200 knows that the received content is X, for example, Hxy (i) can be obtained by the following arithmetic expression (4).

Hxy (i) = Y (i) / X (i)
= (X (i) * H (i) + n (i)) / X (i)
= H (i) + n '(i) (4)
What is important here is to reduce the influence of n ′. For this reason, normally, as shown in FIG. 20, the transmitting terminal 1000 transmits a large number of X (i), and the receiving terminal 1200 obtains an average H ′ (i) of a large number of Hxy (i), for example. The average H ′ (i) of Hxy (i) is the amplitude. Similarly, the noise amount is a standard deviation of Hxy (i). In this case, the standard deviation (σ) is obtained by calculating the variance (σ ² ) calculated by the following equation (5).

上記式（４）が従来から知られている、ある状況にあるチャンネルを経た信号の振幅を推定する高速のＬＳ（Ｌｅａｓｔ Ｓｑｕａｒｅ）アルゴリズムである。このＬＳアルゴリズムを用いた場合、ＱｏＳの期限内にチャンネル推定を行うことが可能であるが、多数のＨｘｙ（ｉ）を記憶する必要があるなど回路規模が大きくなるので現実にはあまり実装されていなかった。

The above equation (4) is a high-speed LS (Least Square) algorithm for estimating the amplitude of a signal passing through a channel in a certain situation. When this LS algorithm is used, it is possible to perform channel estimation within the deadline of QoS. However, since the circuit scale becomes large, for example, it is necessary to store a large number of Hxy (i), it is not so implemented in reality. There wasn't.

  On the other hand, LMS (Least Mean Square) is an algorithm that can be easily realized that has been used frequently. However, since this algorithm requires several seconds to complete processing, channel estimation may be performed within the QoS deadline. could not.

  Incidentally, the transfer function H differs depending on the path between the transmitting terminal 1000 and the receiving terminal 1200. That is, H between the transmitting terminal 1000 and the receiving terminal 1200 is different from H between another transmitting terminal 1000 ′ and the receiving terminal 1200. Therefore, when channel correction is performed by the above method, it is necessary to identify which terminal has transmitted data. For this reason, in HPAV, the AVFC portion is safe so that communication is possible even if communication is performed through a path in any assumed communication situation (noise on the communication path, signal degradation due to impedance reduction, etc.). It is transmitted with QPSK which is low modulation accuracy. As a result, even if channel estimation is performed using the preamble, it is possible to sufficiently demodulate the AVFC. By demodulating the AVFC, it is possible to perform channel correction (assuming that the channel estimation is completed) and demodulation of the PB portion. Since PB can be understood, demodulation can be performed with more efficient Hxy (Hxy in Expression (4) described above) in PB.

  The method for estimating the SNR by the receiving terminal 1200 is used in, for example, DSL (Digital Subscriber Line) or PLC.

  Note that when the channel condition, for example, noise state or impedance changes, the transfer function H also changes. Here, when H1 is before change and H2 is after change, channel estimation / correction appropriate for H1 is performed before change and the optimum modulation accuracy is used for H1, but when the transfer function is changed to H2, H1 Under the optimum conditions, data transmission cannot be performed. That is, when the modulation accuracy obtained in the state before the channel state is changed is used for modulation of data after the channel state is changed, especially when the change in the channel state is large, the data at the receiving terminal 1200 is changed. The error rate can be high.

  In particular, when the PLC channel is compared with the DSL channel, only the telephone device and the DSL modem are connected to the DSL channel, whereas various home appliances are connected to the PLC channel. Furthermore, DSL modems are designed with DSL communications in mind, whereas household appliances do not consider PLC communications, so connecting household appliances to a PLC channel will change the channel impedance and affect it. In some cases, the signal amplitude and noise amount fluctuate rapidly. In such a case, it is necessary to perform the channel estimation process again.

  The PLC has been conventionally used for Internet communication. Specifically, since it was used for reading emails and browsing web pages, there was no problem even if communication was interrupted for several seconds. However, when video communication is performed, the validity period of the QoS frame is set to 100 to 200 ms in order to ensure real-time performance. Furthermore, the time limit in the case of VoIP (Voice Over Internet Protocol) used for voice such as telephone is 100 ms or less. Therefore, even when the channel status changes dramatically in the PLC, it is necessary to recognize the change in the channel status and estimate the signal amplitude and noise within the QoS deadline. Here, although the LMS algorithm can be used to estimate the amplitude of the PLC channel as in Non-Patent Document 1, it cannot be completed within the QoS deadline when LMS is used. Further, since the LMS algorithm is only an amplitude estimation and does not consider the noise factor, it is necessary to further estimate the noise in order to perform a good channel estimation. However, as shown in the above equation (5), the noise ( In order to obtain (σ), it is necessary to obtain in advance an average of Hxy, that is, an average of amplitude. An example of processing in this case is shown in FIG. Here, when the channel status changes, the receiving terminal 1200 first recognizes the change in the channel status, and the receiving terminal 1200 requests the transmitting terminal for a channel estimation packet. Next, the transmitting terminal starts transmission of a channel estimation packet, and the receiving terminal 1200 estimates the amplitude using LMS. When the estimation of the amplitude (average) is finished, the noise is estimated next. Therefore, it takes a very long time.

  As a method for estimating the amplitude and noise at high speed, for example, there is Patent Document 1. In Patent Document 1, amplitude estimation is performed using a preamble (in literature 1, two long training frames and names), and noise estimation is performed using both the preamble and the OFDM symbol. That is, in the preamble portion, the amplitude is estimated using the first long training frame (Equation 12 of Literature 1), and the noise is estimated using the estimated amplitude (Equation 14 of Literature 1). The same operation is performed on the second long training frame, and the final amplitude and noise estimate is the average of the estimates obtained on the two long training frames. Note that the amplitude estimation used here is a high-speed calculation method unlike the LMS. In this way, channel estimation can be performed at high speed by estimating the amplitude and noise for each known long training frame.

In Patent Document 1, the known information is set as “Preamble”. However, when the known information is used as a channel estimation packet, the mathematical expression of Patent Document 1 can be used as it is for HPAV.
S. Morosi, D. Marabissi, E. Del Re, R. Fantacci, N. Del Santo,: "A rate adaptive bit-loading algorithm for a DMT modulation system for in-building power-line communications", in Global Telecommunications Conference 2005, GLOBECOM '05. IEEE, Nov / Dec 2005, Vol 1, pp. 403-407 JP 2005-536103 A

  However, the mathematical formula for obtaining the amplitude in Patent Document 1 is a matrix operation. Patent Document 1 is for a wireless LAN, and the FFT length used in the wireless LAN is assumed to be 64. Even in the case of a 64 × 64 matrix operation, the amount of calculation becomes very large. On the other hand, the length of the FFT of HPAV is 3072. Therefore, when the method such as Patent Document 1 is applied to HPAV, the amount of calculation is further increased. There is a problem that the circuit size becomes enormous and enormous.

  In addition, the circuit scale of a simple LMS as in Non-Patent Document 1 is relatively simple. However, if an attempt is made to add a noise element, it takes more time to estimate the channel, and data transmission such as video transmission requires no QoS. When there is such a time limit, there has been a problem that channel estimation is not completed within the time limit, and there is a risk that the image may be disturbed due to a frame loss.

  The present invention was devised to solve such problems, and its purpose is to reduce the amount of computation and the circuit scale, and to perform high-speed and accurate channel estimation applicable to HPAV and the like, A communication system, a communication method, a communication program, and a recording medium are provided.

  In order to solve the above problems, a communication device according to the present invention includes a communication unit that transmits and receives a packet, a channel estimation packet request unit that requests a channel estimation packet, and a channel based on a plurality of received channel estimation packets. Channel estimation means for estimating, each time the channel estimation means receives the channel estimation packet transmitted from the transmission side device in response to the request from the channel estimation packet request means. Based on the average value and variance of the channel transfer function calculated based on the channel estimation packet received up to and the channel transfer function based on the newly received channel estimation packet, the average of the new channel transfer function Value and variance of the calculated channel transfer function It is characterized by performing channel estimation based on the average value dispersion and.

More specifically, the channel estimation means includes an average calculation unit, a dispersion calculation unit, and a modulation accuracy selection unit. The average calculation unit is a first storage unit that stores an average value H (n) calculated from packets received up to the previous time (n-th time), a predetermined operation involving the average value up to the previous time and the packet that has just been received. And the (n + 1) -th average value is calculated, and the (n + 1) -th average value H (n + 1) is stored in the second storage unit 401c for calculation with the next packet. used. The variance calculation unit entangles the first storage unit that stores the variance value σ ² (n) calculated from the packets received up to the previous time (n-th), the variance value up to the previous time, the currently received packet, and the average value and a second storage unit for storing an arithmetic unit for calculating a performing predetermined operation (n + 1) th variance sigma ² (n + 1), calculated (n + 1) th variance sigma ² to (n + 1) Te It is out. The modulation accuracy selection unit selects the modulation accuracy corresponding to the channel condition based on the values obtained from the average calculation unit and the dispersion calculation unit, and notifies the transmission terminal via the communication means.

  Here, the average value of the (k + 1) th channel transfer function H when the (k + 1) th channel estimation packet is received is expressed by the following equation (6).

（ｋ＋１）番目の前記分散σ²は下式（７）

The (k + 1) th variance σ ² is expressed by the following equation (7).

の式で算出される。なお、これらの算出式は、後述する通信システム、通信方法においても同様である。

It is calculated by the following formula. These calculation formulas are the same in the communication system and communication method described later.

  The communication apparatus of the present invention further includes channel estimation packet requesting means for requesting a channel estimation packet, and the channel estimation means performs channel estimation in response to a request from the channel estimation packet requesting means. I do.

  Furthermore, the communication apparatus of the present invention finishes channel estimation when (1) a predetermined number of channel estimation packets are received as timing for ending channel estimation. (2) When the SNR (signal-to-noise ratio) calculated by the average value / sqrt (dispersion) is calculated every time a channel estimation packet is received, and the absolute value of the change amount of the SNR becomes a predetermined value The channel estimation ends. (3) The channel estimation is terminated when a predetermined time has elapsed since the reception of the first channel estimation packet. This corresponds to the following three cases.

  Furthermore, the communication apparatus of the present invention is an apparatus that uses OFDM (Orthogonal Frequency Division Multiplexing) communication as a communication means, and notifies the OFDM modulation accuracy to the transmission side based on the channel estimation result.

  The communication system of the present invention is a communication system including a device that estimates a channel based on a plurality of channel estimation packets, and the transmission side device receives a plurality of channel estimation packets in response to a request from the reception side device. Each time the receiving device receives the channel estimation packet, the average value and variance of the channel transfer function calculated based on the previously received channel estimation packet and the newly received channel estimation are transmitted. Channel estimation means for calculating a mean value and variance of a new channel transfer function based on a channel transfer function based on a packet for use, and performing channel estimation based on the calculated mean value and variance of the channel transfer function It is characterized by providing.

  In the communication system of the present invention, the transmission side device transmits a plurality of channel estimation packets in response to a request from the reception side device.

  Also, according to the communication system of the present invention, the receiving side device (1) ends channel estimation when a predetermined number of the channel estimation packets are received. (2) When the SNR (signal-to-noise ratio) calculated by the average value / sqrt (dispersion) is calculated every time a channel estimation packet is received, and the absolute value of the change amount of the SNR becomes a predetermined value The channel estimation ends. (3) The channel estimation is terminated when a predetermined time has elapsed since the reception of the first channel estimation packet. This corresponds to the following three cases.

  Further, the communication system of the present invention uses OFDM (Orthogonal Frequency Division Multiplexing) communication as a communication means, and the receiving side apparatus notifies the transmitting side of the modulation accuracy of OFDM based on the result of channel estimation.

  Further, the communication method of the present invention includes a step of requesting a transmission side device to transmit a channel estimation packet when performing communication between a transmission side device and a reception side device, and in response to issuing the request. Each time the channel estimation packet transmitted from the transmission side device is received, the average value and variance of the channel transfer function calculated based on the channel estimation packets received so far, and the newly received channel estimation Calculating a new average value and variance of the channel transfer function based on the channel transfer function based on the packet for use, estimating a channel based on the calculated average value and variance of the channel transfer function, Calculated when a predetermined number of channel estimation packets are received or by the average value / sqrt (variance). SNR (signal-to-noise ratio) is calculated every time a channel estimation packet is received. When the absolute value of the amount of change in the SNR reaches a predetermined value, or after receiving the first channel estimation packet, And a step of ending channel estimation at any timing when time elapses.

  Note that the communication method can be provided as a communication program for causing a computer to execute each step, and the communication program can be provided by being recorded on a computer-readable recording medium.

  Since the present invention is configured as described above, the computation scale and circuit scale do not increase as in the prior art, and also in situations such as HPAV where there is a time limit on QoS for frame transmission such as video and audio transmission, High-speed channel estimation capable of completing the calculation for estimation within the time limit can be performed.

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

<Embodiment 1>
FIG. 1 is a block diagram illustrating a configuration of a main part of the communication device 100 according to the first embodiment. In the first embodiment, a case will be described in which the number of packets used for channel estimation is fixed to a predetermined number (K).

  In the figure, the communication apparatus 100 of the first embodiment is roughly composed of a communication unit 10, an estimation packet request unit 20, an estimation packet detection unit 30, a channel estimation unit 40, and a comparison unit 50.

  FIG. 2 is a block diagram showing a more detailed configuration of the channel estimation means 40.

  The channel estimation means 40 includes an average calculation unit 401, a dispersion calculation unit 402, and a modulation accuracy selection unit 403.

  The average calculation unit 401 is a first storage unit 401a that stores an average value H (n) calculated from packets received up to the previous time (n-th time), a predetermined value obtained by linking the average value up to the previous time and the currently received packet. And (n + 1) -th average value H (n + 1) is stored in the second storage unit 401c, and the next packet and the (n + 1) -th average value are calculated. Used for operations.

The variance calculating unit 402 stores a variance value σ ² (n) calculated from packets received up to the previous time (n-th time), a variance value up to the previous time, a packet received now, and an average value the performing predetermined calculation twine (n + 1) th arithmetic unit 402b for calculating the variance sigma ² (n + 1), calculated (n + 1) th second storage unit for storing the variance values sigma ² (n + 1) 402c.

  The modulation accuracy selection unit 403 selects the modulation accuracy according to the channel status based on the values obtained from the average calculation unit 401 and the dispersion calculation unit 402, and transmits the transmission terminal (the transmission shown in FIG. 11) via the communication means 10. To the terminal 1000).

  Since data communication is performed between the communication device on the transmission side and the communication device on the reception side, the communication device 100 of the first embodiment can be applied to both communication devices on the transmission side and the reception side. However, here, for convenience of explanation, a case where the present invention is applied to a receiving-side communication apparatus (corresponding to the receiving terminal 1200 shown in FIG. 11) will be described. Further, although OFDM (orthogonal frequency division multiplexing) is assumed as a communication method, the present invention is not limited to this.

  The communication means 10 is means for transmitting and receiving frames. Although the range included in the communication means depends on the configuration of the apparatus, the communication means 10 in this case is other than the estimation packet request means 20, the estimation packet detection means 30, and the channel estimation means 40 in the normal MAC / PHY layer. Means the part. This corresponds to, for example, an analog unit, FFT (Fast Fourier Transform) that converts a frequency axis signal into a time axis signal in the case of OFDM and a part that performs inverse FFT.

  The estimation packet requesting unit 20 monitors a change in the channel status based on the received packet, and when there is a change, requests the transmitting side to transmit the estimation packet via the communication unit 10. Further, following the transmission request for the estimation packet, the channel estimation unit 40 is requested to start the channel estimation process.

  The estimation packet detection means 30 recognizes this packet when receiving the estimation packet from the transmission side, and passes the information of the recognized packet to the channel estimation means 40.

  The channel estimation means 40 estimates the signal amplitude and noise based on the channel estimation packet. That is, the channel estimation unit 40 includes an average calculation unit 401, a dispersion calculation unit 402, and a modulation accuracy selection unit 40. The channel estimation packet notified from the estimation packet detection unit 30 is converted into an average calculation unit. The information is input to the operation unit 401 and the operation unit of the variance calculation unit 402. After the detailed estimation is completed, the channel estimation unit 40 passes the modulation accuracy information selected from the signal amplitude and noise estimation results by the modulation accuracy selection unit 403 to the communication unit 10.

  The communication means 10 informs the transmission side of this information and requests the transmission side to stop the transmission of the estimation packet. The transmitting side uses this information to modulate the next data frame. Note that the information received from the estimation packet detection means 30 may be Y (i) or Hxy (i) in the above equation (10). However, when Y (i) is received, it is necessary to calculate Hxy (i) = Y (i) / X (i) within the channel estimation means 40 to obtain Hxy (i).

  The comparison means 50 is a means for comparing the preambles of a plurality of packets (Long MPDU) in order to detect whether or not there has been a change in the channel status that requires an estimation packet. That is, the comparing unit 50 sequentially compares two preambles included in two adjacent frames that have been transmitted, and when the information of both preambles is different, the channel status has changed. judge. The result of the comparison is notified to the estimation packet request unit 20. In addition, although the comparison means 50 of this example detects the change of a channel condition using the comparison of adjacent preambles, it is not limited to the comparison of preambles. For example, it is possible to recognize that the channel communication status has changed by performing an error check on the received packet and detecting that the rate of increase / decrease in the error frequency exceeds a predetermined value. Also, when using preamble comparison, it is possible to compare not only adjacent preambles but also a predetermined number N of preambles. In this case, N may be a fixed value if the channel status change is slow, or may be changed as needed if the channel status change is rapid.

  Here, an outline of a procedure of channel estimation processing performed by the communication apparatus according to the first embodiment will be described with reference to a flowchart shown in FIG. Here, it is assumed that the communication apparatus is arranged as a receiving terminal 1200 that receives packets from the transmitting terminal 1000 via the channel 1100 as shown in FIG.

  In step S100, the receiving terminal 1200 extracts a preamble from a packet received as general data communication. Next, in step S101, this preamble is compared with the preamble of the previously received packet. Then, using the result of the comparison in step S101, it is determined in the next step S102 whether or not the channel status has changed. That is, if the previous and current preambles match, it is determined that there is no change in the channel status (No in step S102), and if they do not match, it is determined that the channel status has changed (determined Yes in step S102). To do.

  Here, if it is determined No in step S102, that is, if there is no change in the channel status, the process proceeds to step S103 to perform normal packet reception processing. On the other hand, if Yes is determined in step S102, that is, if there is a change in the channel status, the process proceeds to step S104.

  In step S104, a counter for counting the number of channel estimation packets to be received is reset. In next step S105, the transmission terminal 1000 is requested to transmit a channel estimation packet. Since the transmission terminal 1000 transmits a channel estimation packet in response to this request, the channel estimation packet is received in step S106 and 1 is added to the value of the counter described above. This channel estimation packet has a known content between the transmitting terminal 1000 and the receiving terminal 1200 as described above.

  Next, in step S107, an operation for obtaining an average of the transfer function H from the received packet is performed. This average is the estimated amplitude. Next, in step S108, the variance of the transfer function is calculated from the received packet. The square root of this variance is the standard deviation and becomes the estimated noise. Note that the details and meaning of this calculation have already been described in the above <Description of data transfer by multi-carrier method>, and a description thereof will be omitted here.

  Next, in step S109, it is determined whether or not a predetermined number of channel estimation packets determined in advance have been received by looking at the counter value. The number of received channel estimation packets is counted by a general counter (not shown). This predetermined number is determined in consideration of channel conditions, time limit due to application and QoS, required estimation accuracy, and the like. As an example, (reception time + calculation time) × the number of receptions may be determined as the number of receptions that falls within the processing time that satisfies the QoS required by the application. For example, when the channel condition is relatively stable, when rough estimation is sufficient, or when the time limit due to QoS is severe, the number of received channel estimation packets may be reduced with a small number. In the first embodiment, the counter is first reset and incremented each time a channel estimation packet is received until the predetermined number is reached. However, the predetermined number is first set and the channel estimation packet is It may be subtracted every time it is received, and the judgment condition may be that the counter value becomes 0 (zero) at S109.

  Here, when the determination in step S109 is No, the number of received channel estimation packets is smaller than the predetermined number, so the process returns to step S106 to receive the remaining channel estimation packets. Then, every time a channel estimation packet is received, the calculation of the average value and variance of the transfer function H is repeated each time it is received. This operation is repeated until a predetermined number of channel estimation packets are received. As a result, when it is determined as Yes in step S109, that is, when reception of a predetermined number of channel estimation packets is completed, the process proceeds to step S110, and the amplitude is calculated from the average value and variance of the transfer function H calculated so far. And noise estimation. Next, in step S111, the optimal modulation accuracy obtained from the estimated amplitude and noise, for example, 16QAM in OFDM is notified to the transmission terminal 1000.

  When the modulation accuracy is transmitted from the receiving terminal 1200 by the process of step S111, the transmitting terminal 1000 determines that the operation of channel estimation is completed and ends transmission of the channel estimation packet. Next, the data packet is transmitted with the modulation accuracy notified from the receiving terminal 1200.

  In the first embodiment, as described in the block diagram shown in FIG. 2 and the flowchart in FIG. 3, the average and variance are sequentially updated every time a channel estimation packet is received. As shown in FIG. In FIG. 4, the average of the transfer function Hxy is

で表す。（ｋ＋１）番目のチャンネル推定用パケットを受信したときの伝達関数Ｈｘｙの平均は下記の式（８）で計算される。

Represented by The average of the transfer function Hxy when the (k + 1) th channel estimation packet is received is calculated by the following equation (8).

また、（ｋ＋１）番目のチャンネル推定用パケットを受信したときの伝達関数Ｈｘｙの分散σ²は下記の式（９）で計算される。

Further, the variance σ ² of the transfer function Hxy when the (k + 1) th channel estimation packet is received is calculated by the following equation (9).

ここで、式（８）の演算を図式化したものが図４のａ部であり、式（９）の演算を図式化したものが図４のｂ部である。これら式（８）、式（９）、及び図４から明らかなように、本実施形態１では、伝達関数Ｈｘｙの平均と分散とはチャンネル推定用パケットを受信する度に双方を連動させて逐次計算している。つまり、１個のチャンネル推定用パケットを受信すると、そのパケットに対応して平均と分散がそれぞれ計算されるようにしている。そして、所定の数のチャンネル推定用パケットを受信し終わったときの伝達関数Ｈｘｙの平均から振幅推定を行い、分散から雑音推定を行っている。実際には、雑音推定を行うのに伝達関数Ｈｘｙの標準偏差を用いるので、分散の平方根を求めることになる。所定の数のチャンネル推定用パケットを受信し終わった時点で、伝達関数Ｈｘｙの平均と分散が求まっている。また、式（８）で分かるように、Ｈｘｙの平均を求める演算には（１／ｋ）の除算が必要である。除算は比較的演算に時間を費やすので、除算を避けるため、本実施形態１では、（１／ｋ）（ｋ＝１、２、３、・・・・所定数）の値を予め図示しないテーブルに保存しておき、（１／ｋ）の値についてはその都度演算せずに、このテーブルからｋに対応する値を読み出して乗算に使用する。しかし、高速除算が可能である、ＱｏＳで要求される時間制限が長いなど、種々の条件が許すのであれば、（１／ｋ）を、チャンネル推定用パケットを受信する度にその都度計算しても良い。

Here, the part a of FIG. 4 is a diagram of the calculation of equation (8), and the part b of FIG. 4 is a diagram of the calculation of equation (9). As is clear from these equations (8), (9), and FIG. 4, in the first embodiment, the average and the variance of the transfer function Hxy are sequentially linked each time a packet for channel estimation is received. I'm calculating. That is, when one channel estimation packet is received, an average and a variance are calculated for each packet. Then, amplitude estimation is performed from the average of the transfer function Hxy when a predetermined number of channel estimation packets have been received, and noise estimation is performed from the variance. Actually, since the standard deviation of the transfer function Hxy is used for noise estimation, the square root of the variance is obtained. At the time when a predetermined number of channel estimation packets have been received, the average and variance of the transfer function Hxy are obtained. Further, as can be seen from the equation (8), a division of (1 / k) is necessary for the calculation for obtaining the average of Hxy. Since division takes a relatively long time for calculation, in the first embodiment, a value of (1 / k) (k = 1, 2, 3,..., A predetermined number) is not shown in advance in the first embodiment. The value corresponding to k is read out from this table and used for multiplication without calculating each time the value of (1 / k). However, if various conditions permit, such as high-speed division is possible and the time limit required by QoS is long, (1 / k) is calculated each time a channel estimation packet is received. Also good.

  As described above, in the first embodiment, each time a channel estimation packet is received, both the average and variance of the transfer function H are calculated in association with each other, and the calculation result is received as the channel estimation packet. It is updated sequentially each time you do it. Therefore, since only the last calculation result needs to be used, a small amount of storage device is required for data storage.

  In addition, if the calculation is completed before the next packet is received, the calculation time is distributed, so that the processing can be performed more efficiently than when a predetermined number of packets are calculated after reception. This is shown in FIG. In FIG. 5, each time a channel estimation packet is received, a new average and variance are calculated using the previously calculated average and variance, and the value of the transfer function H of the received packet at that time. To do. When a predetermined number of packets are received while being sequentially calculated in this way, amplitude estimation and noise estimation are performed based on the last calculated average and variance, and modulation accuracy is determined based on the result. To be notified.

<Embodiment 2>
In the first embodiment, the number of packets used for channel estimation is set to a predetermined number (K). However, in the second embodiment, this number (K) varies depending on channel conditions. A description will be given of a case where the variation range of K has an upper limit.

  Here, the amplitude of the signal estimated from the average of the transfer function H is SC, the noise estimated from the standard deviation of the amplitude is NC, the estimated SNR (Signal To Noise Ratio) is SNRC and the symbol is defined, and SNRC = SC / NC is defined. To do. Using this symbol, the calculation represented by the following formula is performed.

  That is, n packets are received, the average and standard deviation of the transfer function H are obtained, and SNRC (n) in dB (decibel) units is calculated. Further, the next one packet is received, the average and standard deviation of the transfer function H up to (n + 1) packets are obtained, and SNRC (n + 1) is obtained. Then, the difference (n) in dB units = SNRC (n) −SNRC (n + 1) is calculated.

  The meaning of this equation is that when a plurality of channel estimation packets are received, the SNRC is calculated for each of the first n and (n + 1) additional channel estimation packets, and the difference between them is calculated. It is to take. If the absolute value of the difference (n) ≦ TH (predetermined value) as a result of these calculations, the channel estimation is terminated.

  That is, an appropriate number of channel estimation packets are received, a difference (n) between the two groups of SNRCs is obtained, and if the difference (n) is greater than a predetermined value TH, channel estimation is continued. If the difference (n) is equal to or smaller than the predetermined value TH, the channel estimation is terminated at that time. That is, the channel estimation is terminated when the change in the channel status is reduced.

  Note that n is determined to be an appropriate number in consideration of estimation accuracy and time limit defined by QoS. In the second embodiment, the number of additional packets received for difference calculation after receiving n channel estimation packets is one, but the number is not limited to one. An appropriate one is determined in consideration of the accuracy of target amplitude estimation and noise estimation and the processing time for determining them.

  As a value of TH, it is practically possible to change the modulation accuracy level (QPSK, 16QAM, etc.) by one step when the SNR is changed by 3 to 5 dB as a guideline. Therefore, for example, TH = 0.5 dB is sufficient. It is. In other words, when the SNRC difference (n) ≦ 0.5 dB, it is determined that the SNRC change is small and the channel condition is stable, and the reception of channel estimation packets and the calculation process are terminated. The noise is estimated.

Further, if reception and calculation for channel estimation are performed indefinitely until the SNRC difference (n) is equal to or less than a predetermined threshold, the time limited by the QoS required by the application may not be satisfied. . That is, in video transmission or the like, the video is disturbed unless the data packet reception interval is within a certain time. In this case, there is a predetermined upper limit time determined by the application, and the reception of the channel estimation packet and the calculation of the channel estimation must be completed within the predetermined upper limit time. When the number of received channel estimation packets reaches the upper limit number defined from this upper limit time, the reception of channel estimation packets and the calculation process are terminated even if the SNRC difference (n) is not equal to or less than a predetermined threshold, The amplitude and noise of the signal are estimated from the average and variance of the transfer function H at that time. In this example, SNRC and TH are compared in dB, but it is not necessarily limited to dB. For example, it is good also as a simple ratio of SC and NC as SNRC. In this case, TH also has a value corresponding to the simple ratio.
When the channel estimation is completed, the amplitude and noise are estimated from the average value and variance of the transfer function H calculated and accumulated until the last round, and the optimum modulation accuracy (QPSK, 16QAM, etc.) obtained from the estimated amplitude and noise is estimated. Is sent to the transmission terminal 1000.

  The above operation will be described again in detail with reference to the flowchart shown in FIG. 6 (FIGS. 6A and 6B).

  The processing from step S200 to step S203 in FIG. 6 is the same as the processing from step S100 to step S103 in FIG.

  If it is determined in step S202 that the channel status has changed, the process proceeds to step S204. In step S204, a request for a channel estimation packet is transmitted to transmitting terminal 1000. In step S205, n channel estimation packets transmitted from the transmission terminal 1000 are received, which is a number determined as appropriate from the estimation accuracy and QoS. In the next step S206, an operation for obtaining an average (n) of the transfer functions H from the received n channel estimation packets is performed. This average is the estimated amplitude.

  Next, in step S207, calculation is performed to obtain the variance (n) of the transfer function from the received n channel estimation packets. The square root of this variance is the standard deviation and becomes the estimated noise. The details and meaning of this calculation have already been described in <Description of data transfer by multi-carrier method> as described above in the first embodiment, and thus description thereof is omitted here.

  Next, in step S208, SNRC (n) is calculated as an average (n) / √ variance (n) (standard deviation). Thereafter, in step S209, one channel estimation packet is received. Next, in step S210, an average (n + 1) based on (n + 1) channel estimation packets is calculated, and in a next step S211, a variance (n + 1) based on (n + 1) channel estimation packets is calculated. . In step S212, SNRC (n + 1) is calculated as an average (n + 1) / √ variance (n + 1). Next, in step S213, the difference (n) = SNRC (n) −SNRC (n + 1) is calculated. Next, in step S214, it is determined whether the difference (n) is greater than a predetermined threshold value TH.

  As a result, if it is determined in step S214 that it is smaller than or equal to the predetermined threshold value TH (if it is determined Yes in step S214), the channel status has stabilized to a level at which the operation of channel estimation may be completed. Judgment is made, and reception of a new channel estimation packet is terminated, and the routine goes to Step S215. In step S215, the signal amplitude and noise are estimated from the average and standard deviation of the transfer function H calculated so far. Next, in step S216, the optimum modulation accuracy is selected from the amplitude and noise estimated in step S215, and the selected modulation accuracy is notified to the transmission terminal 1000. The modulation accuracy is, for example, 16QAM or QPSK.

  On the other hand, if the difference (n) is larger than the predetermined threshold value TH in step S214 (if it is determined No in step S214), it is determined that the level at which the channel estimation work may be finished is not reached, The process proceeds to S217. In step S217, it is checked whether or not the number of receptions n is an upper limit number. The upper limit number of times is set under the condition that (reception time + calculation time) × reception number falls within the processing time that satisfies the QoS required by the application. If n has reached the upper limit number in step S217 (if it is determined Yes), reception of any more channel estimation packets is terminated, and the process proceeds to step S215 to estimate amplitude and noise, and further in step S216. To notify the modulation accuracy.

  On the other hand, if n has not reached the upper limit number in step S217 (if it is determined No), the process proceeds to step S218. In step S218, n is incremented by 1, and the process returns to step S209 to repeat channel estimation. The addition of n by 1 in step S218 corresponds to the reception of one channel estimation packet in step S209. That is, if the number of additional channel estimation packets received in step S209 is p, the number added in step S218 is p.

  FIG. 7 is a diagram schematically showing a channel estimation packet and an estimation time relationship using the packet in the second embodiment. When the difference (n) is smaller than or equal to the threshold value TH, the channel estimation is terminated, amplitude estimation and noise estimation are performed based on the average and variance at that time, and the modulation accuracy is transmitted to the transmitting terminal based on the result. Notice. The calculation for obtaining the average and variance of the transfer function H is sequentially performed in conjunction with each other every time a channel estimation packet is received, as in the first embodiment.

<Embodiment 3>
In the second embodiment, the reception of packets used for channel estimation and the calculation for obtaining the average and variance of the transfer function H are finished when the change in SNRC, which is the estimated SNR, becomes a certain value or less. However, in the third embodiment, a case will be described in which the channel estimation is terminated when the cumulative calculation time for channel estimation reaches a predetermined value. However, also in the third embodiment, as in the second embodiment, when the number (K) of packets used for channel estimation varies according to the channel condition (however, variation in K is limited by time). ).

  FIG. 8 is a diagram schematically showing a channel estimation packet and an estimation time relationship using the same in the third embodiment.

  Since the request and reception of the channel estimation packet and the calculation for obtaining the average and variance of the transfer function H when the packet is received are the same as in the first and second embodiments, the description thereof is omitted here.

  As shown in FIG. 8, the channel estimation is finished and the modulation accuracy is notified to the transmitting terminal when a predetermined time has elapsed since the reception of the first channel estimation packet. The starting point of time measurement may be when a change in channel status is recognized or when a request for a packet for estimation is transmitted to the transmitting terminal. Alternatively, another arbitrary time may be set as the starting point. This predetermined time is set on the condition that “channel estimation packet reception time + calculation time” falls within a processing time that satisfies the QoS required by the application. When this predetermined time has elapsed, the channel estimation process is terminated at that time regardless of the result, amplitude estimation and noise estimation are performed from the average and variance of the transfer function H at that time, and the modulation accuracy is determined based on the result. To the sending terminal. The elapsed time after recognizing the change in channel status is measured by a general timer (not shown).

  The processing operation at this time will be described with reference to the flowchart shown in FIG.

  The processing from step S300 to step S303 shown in FIG. 9 is the same as the processing from step S100 to step S103 in the flowchart shown in FIG.

  If it is determined in step S302 that the channel status has changed (Yes is determined), the process proceeds to step S304. In step S304, the transmission terminal 1000 is requested to transmit a channel estimation packet. In step S305, the first estimation packet sent from the transmission terminal 1000 is received. Next, in step S306, a timer for measuring time is started. In step S307, the average is calculated from the received channel estimation packets, and in step S308, the variance is calculated. Thereafter, in step S309, it is determined by checking a timer whether a predetermined time has elapsed. As a result, if the determination result in step S309 is No, that is, if the predetermined time has not elapsed, the process proceeds to step S312 to receive the next channel estimation packet, and then returns to step S307 to transfer function. Repeat the calculation to find the mean and variance of H. On the other hand, if the determination result in step S309 is Yes, that is, if a predetermined time has elapsed, the process proceeds to step S310, where amplitude estimation and noise estimation are performed from the average and variance obtained at that time. Do. Next, the process proceeds to step S311, and the optimal modulation accuracy obtained from the estimated amplitude and noise, for example, 16QAM, is notified to the transmission terminal 1000. As in the first and second embodiments, the calculation for obtaining the average and variance of the transfer function H is sequentially performed in conjunction with each other every time a channel estimation packet is received.

<Embodiment 4>
In the fourth embodiment, the function of the channel estimation means 40 of the first to third embodiments is realized as a program, recorded on a recording medium such as a ROM, and the program recorded on the recording medium is read by a computer to obtain a channel. It is an Example in case estimation is implement | achieved by software. An example of the fourth embodiment is shown in FIG.

  Inside the communication apparatus 100, in addition to the means shown in FIG. 1, a CPU 60 and ROM 70 and RAM 80, which are recording media, are further provided. Note that the channel estimation means 40 shown in FIG. 1 is not implemented because it is realized by software in the fourth embodiment. The CPU 60 also serves to control each unit by exchanging commands and data via the bus 90. Here, the ROM is exemplified as the recording medium, but it goes without saying that various recording media can be applied as will be described later. The ROM 70 stores a program that realizes the function of the channel estimation unit 40 described in the first to third embodiments. The RAM 80 stores various data necessary for controlling the communication device 100. Further, it is a condition of when to end the channel estimation, (1) reception of a predetermined number of channel estimation packets, (2) the absolute value of the estimated SNR, that is, the SNRC difference is lower than a predetermined threshold, (3) Any one of the conditions such as elapse of a predetermined time since the reception of the first channel estimation packet is written in the ROM 70 in advance. However, all the above three end conditions may be written, and one of them may be selected. Further, the value of (1 / k) (k = 1, 2, 3,..., A predetermined number) described in the first embodiment is also written in the ROM 70. Further, the modulation accuracy corresponding to the channel estimation is also written in the ROM 70.

  The CPU 60 reads a program for performing channel estimation from the ROM 70. When receiving the channel estimation packet, the CPU 60 calculates the average value and the variance by performing the calculations shown in the equations (14) and (15). When the last channel estimation packet is received, channel estimation is performed from the average value and variance of the transfer function H calculated there. Based on the result, the modulation accuracy is selected from those written in the ROM 70, and the communication means 10 is instructed to notify the transmission terminal 1000 of the modulation accuracy.

  Thus, in the fourth embodiment, the CPU 60 can realize various functions and various processes of the communication device of the present invention simply by reading the program recorded in the ROM 70 and executing the program. In addition, by recording the program on a removable recording medium, the various functions and various processes described above can be realized on an arbitrary computer.

  In addition to the ROM 70, the recording medium may be a program medium that is provided with a program reading device (not shown) as an external storage device and can be read by inserting the recording medium therein.

  In any case, the stored program is preferably configured to be accessed and executed by the microprocessor. Furthermore, it is preferable that the program is read out, and the read program is downloaded to a program storage area of the microcomputer and the program is executed. It is assumed that this download program is stored in advance in the main unit.

  Further, the program media include tape systems such as magnetic tapes and cassette tapes, disk systems such as magnetic disks such as flexible disks and hard disks, and disks such as CD / MO / MD / DVD, IC cards (including memory cards), etc. Or a recording medium that carries a fixed program including a semiconductor memory such as a mask ROM, EPROM, EEPROM, flash ROM or the like.

  In addition, if the system configuration is capable of connecting to a communication network including the Internet, the recording medium is preferably a recording medium that fluidly carries the program so as to download the program from the communication network. Further, when the program is downloaded from the communication network as described above, it is preferable that the download program is stored in the main device in advance or installed from another recording medium.

  In addition, all said embodiment is an illustration, It is not limited to the description content here. That is, the scope of the present invention includes meanings equivalent to the claims and all modifications within the scope.

It is a block diagram which shows the structure of the principal part of the communication apparatus which concerns on Embodiment 1 of this invention. It is a block diagram which shows the structure of a channel estimation means. 3 is a flowchart for explaining the operation of the first embodiment. It is explanatory drawing which showed the outline | summary which updates the average and dispersion | distribution sequentially in this invention. It is a figure which shows the relationship between the packet which concerns on Embodiment 1, and the estimation using it. 10 is a flowchart for explaining the operation of the second embodiment. 10 is a flowchart for explaining the operation of the second embodiment. It is a figure which shows the relationship between the packet which concerns on Embodiment 2, and estimation using it. It is a figure which shows the relationship between the packet which concerns on Embodiment 3, and estimation using it. 10 is a flowchart for explaining the operation of the third embodiment. It is a block diagram when reading a program from a recording medium and performing channel estimation. It is explanatory drawing which shows an example of the general network structure for performing data transfer. It is explanatory drawing which showed typically the packet sequence of the example which succeeded in the transmission of a QoS frame, and the example which failed. It is explanatory drawing which shows the format of HPAV specification. It is explanatory drawing which shows the example of an OFDM signal. It is explanatory drawing which shows the example of the modulation accuracy of 16QAM which allocated 4 bits to the carrier of one complex number. It is explanatory drawing which shows the signal waveform at the time of calculating | requiring the average of an absolute value over several OFDM symbols of the transmission signal TX. It is explanatory drawing which shows the example of the relationship between the transmission signal TX and the reception signal RX after passing through the channel. It is explanatory drawing which shows the example which added the influence of noise to signal RX shown in FIG. It is explanatory drawing which shows the example of 16QAM of signal TX which considered the noise. It is explanatory drawing of the outline | summary which estimates an amplitude and noise from the average and standard deviation of a received signal. It is a figure which shows the packet of the conventional channel estimation, and the relationship of estimation using it.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Communication means 20 Estimation packet request means 30 Estimation packet detection means 40 Channel estimation means 50 Comparison means 60 CPU
70 ROM
80 RAM
90 Bus 100 Communication Device 401 Average Calculation Unit 401a First Storage Unit 401b Operation Unit 401c Second Storage Unit 402 Dispersion Calculation Unit 402a First Storage Unit 402b Operation Unit 402c Second Storage Unit 403 Modulation Accuracy Selection Unit 1000 Transmission Terminal 1100 Communication Medium (Channel)
1200 receiving terminal

Claims (20)

Communication means for receiving at least a packet; and channel estimation means for estimating a channel based on a plurality of received channel estimation packets,
Each time the channel estimation means receives a channel estimation packet transmitted from the transmission side device, the channel estimation function calculates an average value and variance of the channel transfer function calculated based on the channel estimation packet received so far, and a new one. Calculating a new average value and variance of the channel transfer function based on the channel transfer function based on the received channel estimation packet, and performing channel estimation based on the calculated average value and variance of the channel transfer function. A communication device characterized by the above. The communication device according to claim 1,
The average value of the (k + 1) -th channel transfer function H when the (k + 1) -th channel estimation packet is received is as follows:

The (k + 1) th variance σ ² is given by

A communication device calculated by The communication device according to claim 1,
Further comprising channel estimation packet request means for requesting a channel estimation packet;
The channel estimation unit performs channel estimation in response to a request from the channel estimation packet request unit. The communication device according to claim 3.
The channel estimation packet request means requests a channel estimation packet when the preamble information of the received packet and the packet previously received from the packet are different from each other by a predetermined number. . The communication device according to claim 3.
The channel estimation packet request means requests a channel estimation packet when the rate of increase or decrease in error frequency of a received packet exceeds a predetermined value. The communication device according to claim 1 or 2,
The channel estimation means ends the channel estimation when a predetermined number of the channel estimation packets are received in advance. The communication device according to claim 1 or 2,
The channel estimation means calculates an SNR (signal-to-noise ratio) calculated by the average value / sqrt (variance) every time a channel estimation packet is received, and the absolute value of the change amount of the SNR becomes a predetermined value. The communication apparatus is characterized by terminating channel estimation when The communication device according to claim 1 or 2,
The channel estimation means ends the channel estimation when a predetermined time has elapsed since the reception of the first channel estimation packet. The communication device according to any one of claims 1 to 8,
A communication apparatus using OFDM (Orthogonal Frequency Division Multiplexing) communication as a communication means and notifying the transmitter of the modulation accuracy of OFDM based on the result of the channel estimation. A communication system including an apparatus for estimating a channel based on a plurality of channel estimation packets,
The transmitting device transmits a plurality of channel estimation packets to the receiving device,
Each time the receiving side apparatus receives the channel estimation packet, it calculates the average value and variance of the channel transfer function calculated based on the channel estimation packet received so far, and the newly received channel estimation packet. A channel estimation unit that calculates an average value and variance of a new channel transfer function based on the channel transfer function based on the channel transfer function, and performs channel estimation based on the calculated average value and variance of the channel transfer function. A communication system. The communication system according to claim 10,
The receiving side apparatus calculates the average value of the (k + 1) th channel transfer function H when the (k + 1) th channel estimation packet is received by the following equation:

The (k + 1) th variance σ ² is expressed as

The communication system characterized by calculating with the type | formula. The communication system according to claim 10,
The transmission side device transmits a plurality of channel estimation packets in response to a request from the reception side device. The communication system according to claim 10 or claim 11,
The communication system according to claim 1, wherein the receiving side device ends the channel estimation when receiving a predetermined number of the channel estimation packets. The communication system according to claim 10 or claim 11,
The receiving side device calculates an SNR (signal-to-noise ratio) calculated by the average value / sqrt (variance) every time a channel estimation packet is received, and the absolute value of the change amount of the SNR becomes a predetermined value. A communication system, wherein channel estimation is terminated when The communication system according to claim 10 or claim 11,
The communication apparatus according to claim 1, wherein the receiving side device ends channel estimation when a predetermined time has elapsed since the reception of the first channel estimation packet. The communication system according to any one of claims 10 to 15,
A communication system using OFDM (Orthogonal Frequency Division Multiplexing) communication as a communication means, wherein the receiving side device notifies OFDM modulation accuracy to the transmitting side based on the channel estimation result. A communication method for performing channel estimation of a communication channel in the receiving device when performing communication between a transmitting device and a receiving device,
Each time a channel estimation packet transmitted from the transmission side device is received, the average value and variance of the channel transfer function calculated based on the channel estimation packets received so far, and the newly received channel estimation Calculating a mean and variance of the new channel transfer function based on the channel transfer function based on the packet for use;
Performing channel estimation based on the average value and variance of the calculated channel transfer function;
When a predetermined number of channel estimation packets are received, or an SNR (signal-to-noise ratio) calculated by the average value / sqrt (variance) is calculated for each reception of the channel estimation packets, and the SNR Ending channel estimation at any timing when the absolute value of the amount of change reaches a predetermined value or when a predetermined time has elapsed since the reception of the first channel estimation packet;
A communication method comprising: The communication method according to claim 17,
The step of calculating an average value and variance of the channel transfer function includes:
The average value of the (k + 1) th channel transfer function H when the (k + 1) th channel estimation packet is received is expressed by the following equation:

The (k + 1) th variance σ ² is expressed as

The communication method characterized by calculating with the type | formula.   A communication program for causing a computer to execute each step of the communication method according to claim 17 or 18.   A computer-readable recording medium on which the communication program according to claim 19 is recorded.

Images