JPWO2013105538A1 - IQ mismatch correction method and RF transceiver - Google Patents

Abstract

To achieve IQ mismatch correction, there is no standard dependency, no problem of interoperability, and no additional components such as a test signal generation circuit that are likely to increase in area. Based on the reception result of the receiver obtained in the first step using the signal of pattern 1 and the reception result of the receiver obtained in the second step of using the signal of pattern 2 In addition to canceling the gain error and the phase error, the correction matrix of the receiver is obtained and the data of each subcarrier of the receiver is corrected. [Selection] Figure 6

Description

  The present invention relates to a method for correcting IQ mismatch in OFDM direct conversion RF transmission and reception, and an RF transmission and reception apparatus that implements the method.

  The direct conversion method is a method of directly converting a baseband signal into an RF signal (up-conversion) or directly converting an RF signal into a baseband signal (down-conversion). FIG. 1A shows an example of a direct conversion type transmitter and receiver.

In the transmitter 10 shown in FIG. 1A, when a plurality of subcarriers d _k , d _−k ,... Are subjected to inverse discrete Fourier transform by the IDFT circuit 11 and converted from a frequency domain signal to a time domain signal. At the same time, the signals are separated into an I channel baseband signal TxI and a Q channel baseband signal TxQ which are orthogonal to each other. The I channel baseband signal TxI is converted from a digital signal to an analog signal by the DA converter 12I, the high frequency component is removed by the low pass filter 13I, the gain is adjusted by the amplifier 14I, and the I channel local signal TxLo_I is converted by the mixer 15I. The result of multiplication is input to the adder 16 as an RF signal. The Q channel baseband signal TxQ is converted from a digital signal to an analog signal by the DA converter 12Q, a high frequency component is removed by the low pass filter 13Q, the gain is adjusted by the amplifier 14Q, and the Q channel local signal TxLo_Q is converted by the mixer 15Q. The result of multiplication is input to the adder 16 as an RF signal. The output RF signal of the adder 16 is amplified by the power amplifier 17 and radiated from the antenna (not shown) as the transmission signal Tx_out.

The IQ mismatch in the transmitter 10 refers to, for example, a gain error and a phase error between the I channel route and the Q channel route. In the following description, it is assumed that the Q channel local signal TxLo_Q input to the Q channel side mixer 15Q has a gain error g and a phase error φ as shown in FIG. In other words, the local signal TxLo_Q is ideally a -sinω _c t, and a _{-g · sin (ω c t +} φ). Note that the gain error g and the phase error φ have dependency on the carrier frequency, and the error differs for each subcarrier. Therefore, the assumption of g and φ is made for the subcarrier.

In the receiver 20 shown in FIG. 1 (b), the analog RF signal Rx_in inputted from an antenna (not shown) and amplified by the low noise amplifier 21 is multiplied by the I channel local signal TxLo_I by the mixer 22I, and low-pass. The high frequency component is removed by the filter 23I, the gain is adjusted by the amplifier 24I, the analog signal is converted to the digital signal by the AD converter 25I, and the I channel baseband signal RxI is obtained. Further, the RF signal Rx_in is multiplied by the Q channel local signal TxLo_Q by the mixer 22Q, the high-frequency component is removed by the low-pass filter 23Q, the gain is adjusted by the amplifier 24Q, and converted from an analog signal to a digital signal by the AD converter 25Q. The Q channel baseband signal RxQ. Then, both baseband signals RxI and RxQ are subjected to discrete Fourier transform by the DFT circuit 26 and converted from a time domain signal to a frequency domain signal, thereby generating a plurality of subcarriers R _k , R _−k,. The

The IQ mismatch in the receiver 20 refers to, for example, a gain error and a phase error between the I channel route and the Q channel route. In the following description, it is assumed that the Q channel local signal RxLo_Q input to the Q channel side mixer 15Q has a gain error h and a phase error ψ, as shown in FIG. In other words, the local signal RxLo_Q is ideally is a -sinω _c t, and a _{-h · sin (ω c t +} ψ).

  When a signal is transmitted by the OFDM direct conversion method, a baseband signal in which data is carried on subcarriers is expressed as shown in FIG. 3 (the number of subcarriers is N). When considering the effect of IQ mismatch, a pair of data of a certain subcarrier (kth subcarrier) and data of the opposite subcarrier (-kth subcarrier) centered on the origin of the subcarrier is obtained. In the following, direct conversion will be described focusing on this pair.

In FIG. 4, the influence of IQ mismatch in the transmitter 10 will be described. In FIG. 4, it is assumed that data of d _k is transmitted on the _kth subcarrier and d− _k is transmitted on the _−kth subcarrier. As shown in FIG. 4, the following signals are input as the input baseband signals TxI and TxQ of the transmitter 10. * Represents a conjugate complex number.

ただし、

である。The baseband signal TxI, if TxQ is input, transmission output Tx_out, especially when attention to _{T k} and T ^* _-k, the influence of IQ mismatch, can be expressed as follows.

However,

It is.

しかし、ＩＱミスマッチがあることで、Ｔ_ｋにはｄ^* _−kの成分、Ｔ^* _−kにはｄ_ｋの成分が影響を与えることになる。When there is no IQ mismatch, since g = 1 and φ = 0, J ₁ = 1 and J ₂ = 0, and Equation (2) can be expressed as T _k and T ^* ₋ as shown in Equation (4) below. the _k, should only the component of d _k and d ^* _-k remain respectively.

However, that there is IQ mismatch, the _{T k} d ^* _-k ingredients, will affect the component of _{d k} in T ^* _-k.

Next, the influence of IQ mismatch in the receiver 20 will be described with reference to FIG. In FIG. 5, it is considered that the following signal is input as the input signal Rx_in of the receiver 20. c _k and c _−k are data in the input signal of the receiver 20 corresponding to the data (d _k , d _−k ) carried on the k th and −k th subcarriers.

ただし、

である。When this RF reception signal is processed by the DFT circuit 26 and converted into a baseband signal, the signal can be expressed as follows, particularly focusing on R _k and R ^* _−k .

However,

It is.

しかし、ＩＱミスマッチがあることで、Ｒ_kにはｃ^* _−kの成分、Ｒ^* _−kにはｃ_kの成分がそれぞれ影響を与えることになる。When there is no IQ mismatch, h = 1 and ψ = 0, so that K ₁ = 1 and K ₂ = 0, and Equation (6) can be expressed as R _k and R ^* _{− as} shown in Equation (8) below. Only _{k k} and c ^* _−k components should remain in _k , respectively.

However, that there is IQ mismatch, components c ^* _-k to R _k, so that the component of the c _k affects each of the R ^* _-k.

  In order to suppress the influence of the IQ mismatch described above, there are conventional techniques described in Patent Document 1 and Non-Patent Document 1. These two documents introduce a method of detecting and correcting IQ mismatch by transmitting a known pilot signal from the transmitter side to the receiver side. This is because “the standard stipulates that the pilot tone (test signal) is included in the preamble” or “a signal that is specific to the pilot tone for frequency synchronization and timing synchronization by each company that makes the transmitter / receiver (regardless of the standard). Include a pattern "is used, but in any case, it involves standards dependence and interoperability issues, so it is versatile.

  In addition, the detected mismatch includes not only the IQ mismatch on the receiver side but also the IQ mismatch on the transmission side. Therefore, when the reception and transmission destinations change one after another on the network, the mismatch detection is performed for each transmission / reception. There is a problem that requires.

  Next, Non-Patent Document 2 describes detection and correction of IQ mismatch in the frequency domain, which is not standard dependent. However, this document focuses on “correction only on the receiver side”. Since the contents relate to the present invention, the details will be described below.

とすると、補正出力は

と表される。ただし

である。FIG. 6 shows a correction method. In Non-Patent Document 2, the received signals RxI and RxQ are subjected to discrete Fourier transform by the DFT circuit 26 to obtain subcarriers, and then a correction unit 27 is obtained by a correction matrix 27 obtained by performing a calculation represented by the following equation. 28 is corrected.

Then, the correction output is

It is expressed. However,

It is.

であるとすると、式（１０）は、

となり、補正が完了する。

If so, equation (10) becomes

Thus, the correction is completed.

ただし、

である。
Ｅ｛ ｝は、平均をとった場合の期待値を表す。つまり、Ｅ｛Ｒ_ｋＲ_−k｝は、受信したシンボルでＲ_ｋＲ_−kの計算を行って、その時間的平均を取ることを意味する。ｃ_kとｃ_−kには相関がないと仮定すると、Ｅ｛ｃ_ｋｃ_−k｝＝０、Ｅ｛ｃ^* _ｋｃ^* _−k｝＝０が成立することもある。The detection method of the correction matrix 27 is as follows.

However,

It is.
E {} represents an expected value when an average is taken. That is, E {R _k R _−k } means that R _k R _−k is calculated on the received symbol and the time average is obtained. Assuming that there is no correlation between c _k and c _−k , E {c _k c _−k } = 0 and E {c ^* _k c ^* _−k } = 0 may hold.

によって、Ｋ_１Ｋ_２が求まる。このＫ_１Ｋ_２は

であるので、

により、ｈとψが求まり、式（７）からＫ_１とＫ_２が個々に求まるので、補正用行列２７を実現することができる。Therefore,

Thus, K ₁ K ₂ is obtained. This K ₁ K ₂ is

So

Thus, h and ψ are obtained, and K ₁ and K ₂ are obtained individually from Equation (7), so that the correction matrix 27 can be realized.

JP 2008-17145 A

Andreas Schuchert, Ralph Hasholzner, and Patrick Antoine, “A novel IQ imbalance compensation scheme for the reception of OFDM signals,” IEEE Trans. Consumer Electron. Vol. 47, no. 3, pp. 313? 318, Aug. 2001. M. Windisch and G. Fettweis.Standard-Independent I / Q Imbalance Compensation in OFDM Direct-Conversion Receivers 9th International OFDM Workshop (InOWo'04), Dresden, Germany, September 2004

  However, the methods as described in Patent Document 1 and Non-Patent Document 1 are not versatile because they depend on standards. For this reason, a method that does not depend on the standard as in Non-Patent Document 2 is desirable. However, Non-Patent Document 2 also has the following problems. First, detection and correction cannot be performed before communication, and IQ mismatch cannot be corrected at the initial stage of communication. This can be solved by providing a circuit for generating a test signal in the transmitter / receiver, but an increase in the area is inevitable. Next, the transmitter side IQ mismatch cannot be detected and corrected. This may be corrected by including a similar device in the other transmitter / receiver that forms a communication pair, including a mismatch on the transmission side, but this naturally involves an interoperability problem.

  The present invention has been made in view of the above points, and its purpose is no dependency on standards, no problem of interoperability, no test signal is required for correction of the receiver, and no test signal generation circuit is required. It is an object of the present invention to provide an IQ mismatch correction method having a feature of minimizing an increase in area due to a circuit, and an RF transceiver device implementing the correction method.

により、前記受信機の補正用行列としてのＫ₁，Ｋ₂，Ｋ^* ₁，Ｋ^* ₂を求めて、前記受信機のｋ番目のサブキャリアのデータに対して補正を行うことを特徴とするＩＱミスマッチ補正方法。
ただし、＊は、複素共役である。Ｒ_K1 は、前記送信機のｋ番目のサブキャリアに配置したデータをｄ_ｋとし、送信Ｉチャネルベースバンド信号をＴｘＩ_１とし、送信Ｑチャネルベースバンド信号をＴｘＱ_１とし、前記パターン１の信号を、

とするとき、前記送信機から出力する前記ＲＦ信号をループバック経由で前記受信機に入力させて前記ｋ番目のサブキャリアに得られる受信データである。Ｒ_−ｋ1 は、前記パターン１の信号を用いるとき、ＩＱミスマッチのために前記受信機で−ｋ番目のサブキャリアに得られる受信データである。Ｒ_ｋ2 は、送信Ｉチャネルベースバンド信号をＴｘＩ_２とし、送信Ｑチャネルベースバンド信号をＴｘＱ_２とし、前記パターン２の信号を、

とし、かつ前記送信機のローカル信号をＩチャネルとＱチャネルで互いに入れ替え、前記送信機から出力する前記ＲＦ信号をループバック経由で前記受信機に入力させたとき、前記受信機でｋ番目のサブキャリアに得られる受信データである。Ｒ_−ｋ2 は、前記パターン２の信号を用いかつ前記ローカル信号をＩチャネルとＱチャネルで互いに入れ替えた条件のとき、ＩＱミスマッチのために受信機で−ｋ番目のサブキャリアに得られる受信データである。ｈは、前記受信機における利得誤差である。ψは、前記受信機における位相誤差である。Ｒｅ[ ]は、実数部を示す。Ｉｍ[ ]は、虚数部を示す。
請求項３にかかる発明は、請求項１に記載のＩＱミスマッチ補正方法において、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記パターン１の信号を生成し、かつ前記パターン１の信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン１の信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第３のステップと、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記第１の送信ベースバンド信号を送信Ｉチャネルベースバンド信号とし、前記第２の送信ベースバンド信号の極性反転した信号を送信Ｑチャネルベースバンド信号としたパターン３の信号を生成し、前記パターン３の信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン３の信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第４のステップとを備え、前記第３のステップで得られた前記受信機の受信結果と、前記第４のステップで得られた前記受信機の受信結果基づき、前記送信機の補正用行列を求め、得られた該補正用行列により前記送信機のｋ番目のサブキャリアのデータに対して利得誤差および位相誤差を補正することを特徴とする。
請求項４にかかる発明は、請求項２に記載のＩＱミスマッチ補正方法において、前記受信機の補正用行列としてのＫ₁，Ｋ₂，Ｋ^* ₁，Ｋ^* ₂による補正を前記受信機のｋ番目のサブキャリアのデータについて行い、

により、前記送信機の補正用行列としてのＪ₁，Ｊ₂，Ｊ^* ₁，Ｊ^* ₂を求めて、前記送信機のサブキャリアのデータに対して補正を行うことを特徴とする。ただし、ＲＣ_ｋ1 は、前記送信機で前記パターン１の信号を用い、前記受信機の補正用行列としてのＫ₁，Ｋ₂，Ｋ^* ₁，Ｋ^* ₂による補正を前記受信機の各サブキャリアのデータについて行ったときの、前記受信機でｋ番目のサブキャリアに得られる受信データである。ＲＣ_ｋ3 は、送信Ｉチャネルベースバンド信号をＴｘＩ_３とし、送信Ｑチャネルベースバンド信号をＴｘＱ_３とし、パターン３の信号を

として、前記送信機から出力するＲＦ信号をループバック経由で前記受信機に入力させ、前記受信機の補正用行列としてのＫ₁，Ｋ₂，Ｋ^* ₁，Ｋ^* ₂による補正を前記受信機の各サブキャリアのデータについて行ったときの、前記受信機でｋ番目のサブキャリアに得られる受信データである。ｇは、前記送信機における利得誤差である。φは、前記送信機における位相誤差である。
請求項５にかかる発明は、請求項１に記載のＩＱミスマッチ補正方法において、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記パターン１の信号を生成し、かつ前記パターン１の信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン１の信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第３Ａのステップと、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記第１の送信ベースバンド信号を送信Ｉチャネルベースバンド信号とし、前記第２の送信ベースバンド信号を送信Ｑチャネルベースバンド信号としたパターン３Ａの信号を生成し、前記パターン３Ａの信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン３Ａの信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号を極性反転した信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第４Ａのステップとを備え、前記第３Ａのステップで得られた前記受信機の受信結果と、前記第４Ａのステップで得られた前記受信機の受信結果基づき、前記送信機の補正用行列を求め、得られた該補正用行列により前記送信機のｋ番目のサブキャリアのデータに対して利得誤差および位相誤差を補正することを特徴とする。
請求項６にかかる発明は、請求項２に記載のＩＱミスマッチ補正方法において、前記受信機の補正用行列としてのＫ₁，Ｋ₂，Ｋ^* ₁，Ｋ^* ₂による補正を前記受信機のｋ番目のサブキャリアのデータについて行い、

により、前記送信機の補正用行列としてのＪ₁，Ｊ₂，Ｊ^* ₁，Ｊ^* ₂を求めて、前記送信機のサブキャリアのデータに対して補正を行うことを特徴とする。ただし、ＲＣ_ｋ1 は、前記送信機で前記パターン１の信号を用い、前記受信機の補正用行列としてのＫ₁，Ｋ₂，Ｋ^* ₁，Ｋ^* ₂による補正を前記受信機の各サブキャリアのデータについて行ったときの、前記受信機でｋ番目のサブキャリアに得られる受信データである。ＲＣ_ｋ3 は、送信Ｉチャネルベースバンド信号をＴｘＩ_３とし、送信Ｑチャネルベースバンド信号をＴｘＱ_３とし、パターン３Ａの信号を

とし、かつ前記送信器のローカル信号のうちのＱチャンネルのローカル信号の極性を反転して、前記送信機から出力するＲＦ信号をループバック経由で前記受信機に入力させ、前記受信機の補正用行列としてのＫ₁，Ｋ₂，Ｋ^* ₁，Ｋ^* ₂による補正を前記受信機の各サブキャリアのデータについて行ったときの、前記受信機でｋ番目のサブキャリアに得られる受信データである。ｇは、前記送信機における利得誤差である。φは、前記送信機における位相誤差である。
請求項７にかかる発明は、互いに直交関係にある複数のサブキャリアを入力し逆フーリエ変換して互いに直交関係にある送信Ｉチャネルベースバンド信号と送信Ｑチャネルベースバンド信号とを生成する逆フーリエ変換手段と、該送信Ｉチャネルベースバンド信号をＩチャンネルローカル信号で直交変調するとともに前記送信Ｑチャネルベースバンド信号をＱチャンネルローカル信号で直交変調する直交変調手段と、該直交変調手段で該直交変調した後の両信号を加算してＲＦ信号にする加算手段とを備えた送信機と、ＲＦ信号を受信して直交復調することで受信Ｉチャネルベースバンド信号と受信Ｑチャネルベースバンド信号に分離する直交復調手段と、該受信Ｉチャネルベースバンド信号と該受信Ｑチャネルベースバンド信号をフーリエ変換して互いに直交関係にある複数のサブキャリアを生成するフーリエ変換手段とを備えた受信機と、を有するＯＦＤＭダイレクトコンバージョン方式のＲＦ送受信装置において、前記送信機のｋ番目のサブキャリアにデータを入力したときに得られる前記送信機の互いに直交関係にある第１の送信ベースバンド信号と第２の送信ベースバンド信号のうちの前記第１の送信ベースバンド信号を送信Ｉチャネルベースバンド信号とし、前記第２の送信ベースバンド信号を送信Ｑチャネルベースバンド信号としたパターン１の信号を生成し、かつ前記パターン１の信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン１の信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第１の制御手段と、前記第１の送信ベースバンド信号を送信Ｑチャネルベースバンド信号とし、前記第２の送信ベースバンド信号を送信Ｉチャネルベースバンド信号としたパターン２の信号を生成し、かつ前記パターン２の信号の内の前記送信Ｉチャネルベースバンド信号を前記Ｑチャネルローカル信号で直交変調し、前記パターン２の信号の内の前記送信Ｑチャネルベースバンド信号を前記Ｉチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第２の制御手段とを備え、前記第１の制御手段で得られた前記受信機の受信結果と、前記第２の制御手段で得られた前記受信機の受信結果とに基づき、前記送信機での利得誤差および位相誤差をキャンセルするとともに、前記受信機の補正用行列を求め、得られた該補正用行列によって前記受信機で得られるｋ番目のサブキャリアのデータに対して利得誤差および位相誤差を補正することを特徴とする。
請求項８にかかる発明は、請求項１に記載のＲＦ送受信装置において、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記パターン１の信号を生成し、かつ前記パターン１の信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン１の信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第３の制御手段と、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記第１の送信ベースバンド信号を送信Ｉチャネルベースバンド信号とし、前記第２の送信ベースバンド信号の極性反転した信号を送信Ｑチャネルベースバンド信号としたパターン３の信号を生成し、前記パターン３の信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン３の信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第４の制御手段とを備え、前記第３の制御手段で得られた前記受信機の受信結果と、前記第４の制御手段で得られた前記受信機の受信結果基づき、前記送信機の補正用行列を求め、得られた該補正用行列により前記送信機のｋ番目のサブキャリアのデータに対して利得誤差および位相誤差を補正することを特徴とする。
請求項９にかかる発明は、請求項１に記載のＲＦ送受信装置において、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記パターン１の信号を生成し、かつ前記パターン１の信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン１の信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第３Ａの制御手段と、前記受信機のｋ番目のサブキャリアのデータを前記受信機の前記補正用行列によって補正した上で、前記第１の送信ベースバンド信号を送信Ｉチャネルベースバンド信号とし、前記第２の送信ベースバンド信号を送信Ｑチャネルベースバンド信号としたパターン３Ａの信号を生成し、前記パターン３Ａの信号の内の前記送信Ｉチャネルベースバンド信号を前記送信機のＩチャネルローカル信号で直交変調し、前記パターン３Ａの信号の内の前記送信Ｑチャネルベースバンド信号を前記送信機のＱチャネルローカル信号を極性反転した信号で直交変調し、両直交変調された信号を加算してＲＦ信号としてループバック経由で前記受信機に入力させる第４Ａの制御手段とを備え、前記第３Ａの制御手段で得られた前記受信機の受信結果と、前記第４Ａの制御手段で得られた前記受信機の受信結果基づき、前記送信機の補正用行列を求め、得られた該補正用行列により前記送信機のｋ番目のサブキャリアのデータに対して利得誤差および位相誤差を補正することを特徴とする。To achieve the above object, according to a first aspect of the present invention, there is provided a transmission I channel baseband signal and a transmission Q channel which are orthogonal to each other by inverse Fourier transforming a plurality of subcarriers which are orthogonal to each other. A baseband signal is generated, the transmission I channel baseband signal and the transmission Q channel baseband signal are orthogonally modulated and then added and transmitted as an RF signal, and the receiver receives the RF signal and performs orthogonal demodulation. To separate the received I channel baseband signal and the received Q channel baseband signal, and Fourier transform the received I channel baseband signal and the received Q channel baseband signal to generate a plurality of subcarriers that are orthogonal to each other. IQ mismatch in OFDM direct conversion RF transceiver Among the first transmission baseband signal and the second transmission baseband signal which are correct methods, and are obtained by inputting data to the kth subcarrier of the transmitter, which are orthogonal to each other of the transmitter. The first transmission baseband signal is a transmission I channel baseband signal, the second transmission baseband signal is a transmission Q channel baseband signal, and a pattern 1 signal is generated. The transmission I channel baseband signal is orthogonally modulated with the I channel local signal of the transmitter, and the transmission Q channel baseband signal of the pattern 1 signal is orthogonally modulated with the Q channel local signal of the transmitter A first step of adding the signals subjected to bi-orthogonal modulation to input to the receiver via a loopback as an RF signal; A pattern 2 signal is generated in which the first transmission baseband signal is a transmission Q channel baseband signal, the second transmission baseband signal is a transmission I channel baseband signal, and among the signals of the pattern 2 The transmission I channel baseband signal is quadrature modulated with the Q channel local signal, and the transmission Q channel baseband signal of the pattern 2 signal is quadrature modulated with the I channel local signal, and both quadrature modulated signals And a second step of inputting to the receiver via a loopback as an RF signal, the reception result of the receiver obtained in the first step, and the second step obtained Based on the reception result of the receiver, the gain error and the phase error in the transmitter are canceled, and the correction of the receiver A matrix is obtained, and a gain error and a phase error are corrected with respect to the k-th subcarrier data obtained by the receiver using the obtained correction matrix.
The invention according to claim 2 is the IQ mismatch correction method according to claim 1,

To obtain K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes for the receiver, and correct the k-th subcarrier data of the receiver. IQ mismatch correction method.
However, * is a complex conjugate. R _K1 sets the data arranged in the kth subcarrier of the transmitter to d _k , sets the transmission I channel baseband signal to TxI ₁ , sets the transmission Q channel baseband signal to TxQ _1, and sets the signal of the pattern 1 to ,

Then, the RF data output from the transmitter is input to the receiver via a loopback, and is received data obtained on the kth subcarrier. R- _k1 is received data obtained on the _{-kth subcarrier} by the receiver due to IQ mismatch when the signal of the pattern 1 is used. R _k2 is a transmission I channel baseband signal is TxI ₂ , a transmission Q channel baseband signal is TxQ _2, and the signal of the pattern 2 is

And the local signal of the transmitter is exchanged between the I channel and the Q channel, and when the RF signal output from the transmitter is input to the receiver via a loopback, the receiver receives the k-th sub-signal. Received data obtained on a carrier. R- _k2 is received data obtained on the -k-th subcarrier at the receiver due to IQ mismatch when the pattern 2 signal is used and the local signal is switched between the I channel and the Q channel. is there. h is a gain error in the receiver. ψ is a phase error in the receiver. Re [] indicates the real part. Im [] indicates an imaginary part.
The invention according to claim 3 is the IQ mismatch correction method according to claim 1, wherein the data of the k-th subcarrier of the receiver is corrected by the correction matrix of the receiver, and then the pattern 1 Generating a signal, and orthogonally modulating the transmission I channel baseband signal in the pattern 1 signal with an I channel local signal of the transmitter, and transmitting the transmission Q channel baseband signal in the pattern 1 signal Is quadrature modulated with the Q-channel local signal of the transmitter, and both quadrature modulated signals are added together and input to the receiver via a loopback as an RF signal, and the kth of the receiver After correcting the subcarrier data by the correction matrix of the receiver, the first transmission baseband signal is transmitted to the transmission I channel baseband. And a signal of pattern 3 in which a signal obtained by inverting the polarity of the second transmission baseband signal is used as a transmission Q channel baseband signal, and the transmission I channel baseband signal among the signals of the pattern 3 is Quadrature modulation is performed with the I channel local signal of the transmitter, the transmission Q channel baseband signal of the pattern 3 signal is quadrature modulated with the Q channel local signal of the transmitter, and the signals subjected to both orthogonal modulation are added. And a fourth step of causing the receiver to input an RF signal via a loopback, and the reception result of the receiver obtained in the third step and the reception obtained in the fourth step. Based on the reception result of the transmitter, a correction matrix for the transmitter is obtained, and the gain for the k-th subcarrier data of the transmitter is obtained by the obtained correction matrix. And correcting the difference and the phase error.
The invention according to claim 4 is the IQ mismatch correction method according to claim 2, wherein correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as a correction matrix of the receiver is performed by the k of the receiver. The second subcarrier data,

Thus, J ₁ , J ₂ , J ^* ₁ and J ^* ₂ as the correction matrix of the transmitter are obtained, and correction is performed on the subcarrier data of the transmitter. However, RC _k1 uses the signal of the pattern 1 at the transmitter, and performs correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes of the receiver for each subcarrier of the receiver. Received data obtained for the kth subcarrier by the receiver. RC _k3 is a transmission I channel baseband signal as TxI _3, the transmission Q channel baseband signal and TxQ _3, the signal pattern 3

As described above, an RF signal output from the transmitter is input to the receiver via a loopback, and correction using K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as a correction matrix of the receiver is performed on the receiver. The received data obtained for the k-th subcarrier by the receiver when the data of each subcarrier is obtained. g is a gain error in the transmitter. φ is a phase error in the transmitter.
The invention according to claim 5 is the IQ mismatch correction method according to claim 1, wherein the data of the kth subcarrier of the receiver is corrected by the correction matrix of the receiver, and then the pattern 1 of the pattern 1 is corrected. Generating a signal, and orthogonally modulating the transmission I channel baseband signal in the pattern 1 signal with an I channel local signal of the transmitter, and transmitting the transmission Q channel baseband signal in the pattern 1 signal Is quadrature modulated with the Q channel local signal of the transmitter, and the two orthogonally modulated signals are added together and input to the receiver via a loopback as an RF signal, and the kth of the receiver After correcting the subcarrier data by the correction matrix of the receiver, the first transmission baseband signal is transmitted to the transmission I channel baseband. A signal of pattern 3A is generated using the second transmission baseband signal as a transmission Q channel baseband signal, and the transmission I channel baseband signal of the pattern 3A signal is used as an I channel of the transmitter Quadrature modulation is performed with a local signal, and the transmission Q channel baseband signal of the signal of the pattern 3A is quadrature modulated with a signal obtained by inverting the polarity of the Q channel local signal of the transmitter, and the signals subjected to bi-orthogonal modulation are added. 4A, which is input to the receiver via a loopback as an RF signal, and the reception result of the receiver obtained in the step 3A and the reception obtained in the step 4A. Based on the reception result of the transmitter, a correction matrix for the transmitter is obtained, and the k-th subcarrier data of the transmitter is determined based on the obtained correction matrix. And correcting the gain error and the phase error with respect.
The invention according to claim 6 is the IQ mismatch correction method according to claim 2, wherein correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as the correction matrix of the receiver is performed by the k of the receiver. The second subcarrier data,

Thus, J ₁ , J ₂ , J ^* ₁ and J ^* ₂ as the correction matrix of the transmitter are obtained, and correction is performed on the subcarrier data of the transmitter. However, RC _k1 uses the signal of the pattern 1 at the transmitter, and performs correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes of the receiver for each subcarrier of the receiver. Received data obtained for the kth subcarrier by the receiver. RC _k3 is a transmission I channel baseband signal as TxI _3, the transmission Q channel baseband signal and TxQ _3, the signal pattern 3A

And the polarity of the local signal of the Q channel among the local signals of the transmitter is inverted, and the RF signal output from the transmitter is input to the receiver via a loopback, for correction of the receiver This is the received data obtained for the kth subcarrier in the receiver when correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as a matrix is performed on the data of each subcarrier of the receiver. . g is a gain error in the transmitter. φ is a phase error in the transmitter.
The invention according to claim 7 is an inverse Fourier transform that inputs a plurality of subcarriers that are orthogonal to each other and performs inverse Fourier transform to generate a transmission I channel baseband signal and a transmission Q channel baseband signal that are orthogonal to each other. Means, orthogonal modulation means for orthogonally modulating the transmission I channel baseband signal with an I channel local signal and orthogonally modulating the transmission Q channel baseband signal with a Q channel local signal, and performing orthogonal modulation with the orthogonal modulation means A quadrature that separates the received I-channel baseband signal and the received Q-channel baseband signal by receiving a quadrature demodulator by receiving the RF signal and performing quadrature demodulation by adding a means for adding both later signals to an RF signal Demodulating means, Fourier transform of the received I channel baseband signal and the received Q channel baseband signal In other words, an OFDM direct-conversion RF transmitter / receiver having a Fourier transform means for generating a plurality of subcarriers that are orthogonal to each other, and transmitting data to the kth subcarrier of the transmitter The first transmission baseband signal of the first transmission baseband signal and the second transmission baseband signal that are orthogonal to each other of the transmitter obtained when input is used as a transmission I channel baseband signal, A signal of pattern 1 is generated using the second transmission baseband signal as a transmission Q channel baseband signal, and the transmission I channel baseband signal of the pattern 1 signal is used as an I channel local signal of the transmitter The transmission Q channel baseband signal in the pattern 1 signal is orthogonally modulated by First control means for performing quadrature modulation with the Q channel local signal, adding the signals subjected to both quadrature modulation and inputting the result as an RF signal to the receiver via a loopback, and transmitting the first transmission baseband signal A pattern 2 signal is generated using a Q channel baseband signal, the second transmission baseband signal as a transmission I channel baseband signal, and the transmission I channel baseband signal of the pattern 2 signal as the transmission I channel baseband signal. Perform quadrature modulation with a Q channel local signal, perform quadrature modulation on the transmission Q channel baseband signal of the signal of the pattern 2 with the I channel local signal, add the signals subjected to bi-orthogonal modulation, and loop back as an RF signal And a second control means for inputting to the receiver via, and the reception result of the receiver obtained by the first control means And based on the reception result of the receiver obtained by the second control means, the gain error and the phase error at the transmitter are canceled and the correction matrix of the receiver is obtained and obtained. According to the correction matrix, the gain error and the phase error are corrected with respect to the k-th subcarrier data obtained by the receiver.
The invention according to claim 8 is the RF transmitter / receiver according to claim 1, wherein the signal of the pattern 1 is obtained after correcting the k-th subcarrier data of the receiver by the correction matrix of the receiver. And orthogonally modulates the transmission I channel baseband signal of the pattern 1 signal with the I channel local signal of the transmitter, and converts the transmission Q channel baseband signal of the pattern 1 signal to Third control means for performing quadrature modulation with the Q channel local signal of the transmitter, adding the signals subjected to both quadrature modulation and inputting the result as an RF signal to the receiver via a loopback; After correcting the subcarrier data by the correction matrix of the receiver, the first transmission baseband signal is used as a transmission I channel baseband signal. A pattern 3 signal is generated with a signal obtained by inverting the polarity of the second transmission baseband signal as a transmission Q channel baseband signal, and the transmission I channel baseband signal among the signals of the pattern 3 is generated by the transmitter. Quadrature modulation is performed with an I channel local signal, and the transmission Q channel baseband signal of the pattern 3 signal is quadrature modulated with the Q channel local signal of the transmitter, and the signals subjected to both quadrature modulation are added to obtain an RF signal. And a fourth control means for inputting to the receiver via a loopback as a reception result of the receiver obtained by the third control means and the reception obtained by the fourth control means. Based on the reception result of the transmitter, a correction matrix for the transmitter is obtained, and the gain matrix and the gain error and the k-th subcarrier data of the transmitter are obtained by the obtained correction matrix. And correcting the phase error.
The invention according to claim 9 is the RF transmitter / receiver according to claim 1, wherein the signal of the pattern 1 is obtained after correcting the data of the kth subcarrier of the receiver by the correction matrix of the receiver. And orthogonally modulates the transmission I channel baseband signal of the pattern 1 signal with the I channel local signal of the transmitter, and converts the transmission Q channel baseband signal of the pattern 1 signal to 3A control means for performing quadrature modulation with the Q channel local signal of the transmitter, adding the signals subjected to both quadrature modulation and inputting the result as an RF signal to the receiver via a loopback; After correcting the subcarrier data by the correction matrix of the receiver, the first transmission baseband signal is changed to a transmission I channel baseband signal. , Generating a signal of pattern 3A using the second transmission baseband signal as a transmission Q channel baseband signal, and using the transmission I channel baseband signal of the pattern 3A as an I channel local signal of the transmitter Is orthogonally modulated with the signal obtained by reversing the polarity of the Q channel local signal of the transmitter, and the signals obtained by bi-orthogonal modulation are added to each other. 4A control means for inputting to the receiver via a loopback as a signal, the reception result of the receiver obtained by the 3A control means, and the result obtained by the 4A control means Based on the reception result of the receiver, the correction matrix of the transmitter is obtained, and the k-th subcarrier data of the transmitter is obtained by the obtained correction matrix. And correcting the gain and phase errors.

  According to the present invention, a correction matrix for subcarrier correction of a receiver is obtained using the signals of patterns 1 and 2, and a correction matrix for subcarrier correction of a transmitter is a correction matrix for subcarriers of a receiver. And using the signals of patterns 1 and 3. Therefore, the correction can be performed at the time of starting up the hardware or at the time of non-communication, there is no dependency on the standard, and there is no problem of interoperability. In addition, since a test signal is not required for correction of the receiver, a test signal generation circuit is unnecessary, and an effect of minimizing the area increase due to the additional circuit can be obtained.

It is a circuit diagram which shows the structure of the transmitter and receiver of the conventional direct conversion system. It is explanatory drawing of the gain error and phase error of the transmitter and receiver of the conventional direct conversion system. It is explanatory drawing of a subcarrier. It is explanatory drawing of the influence of IQ mismatch in a transmitter. It is explanatory drawing of the influence of IQ mismatch in a receiver. It is explanatory drawing of the subcarrier correction | amendment of a receiver. It is explanatory drawing of the subcarrier correction | amendment of a transmitter. It is explanatory drawing of operation | movement when the signal of the pattern 1 is input into the transmitter. It is explanatory drawing of operation | movement when the signal of the pattern 2 is input into the transmitter. It is explanatory drawing of operation | movement when the signal of the pattern 3 is input into the transmitter. It is a circuit diagram which shows the structure of RF transmitting / receiving apparatus of the Example of this invention.

  In the present invention, the IQ mismatch of both the transmitter and the receiver is corrected. Also, as a test signal, an arbitrary signal on the transmitter side is looped back to the receiver side in the transceiver, thereby minimizing the added hardware. Furthermore, detection and correction are not performed in pairs with another transmitter / receiver so that there is no problem of standard dependency and interoperability.

  For this purpose, a loopback path is formed in the transmitter / receiver to pass the signal on the transmitter side to the receiver side. Also, baseband signal processing means for performing the same calculation (detection and correction) as in Non-Patent Document 2 is provided on both the transmitter side and the receiver side. In addition, a local oscillator is prepared that can change the phase relationship between the transmission I channel local signal TxLo_I and the transmission Q channel local signal TxLo_Q, and can invert the polarity of the transmission Q channel local signal TxLo_Q.

を用いて、以下のような計算で補正を行う。ｄＣ_ｋ，ｄＣ^* _−ｋは、サブキャリアのデータｄ_ｋ，ｄ^* _−ｋを補正した後のサブキャリアのデータである。

<First embodiment>
FIG. 7 shows a correction method on the transmitter 10 side. Similar to the correction on the receiver 20 side described with reference to FIG. 6, a correction matrix as an inverse matrix of IQ mismatch factors in the frequency region preceding the IDFT circuit 11.

Is used to correct by the following calculation. dC _k and dC ^* _−k are subcarrier data after correcting the subcarrier data d _k and d ^* _−k .

となる。Therefore, the corrected transmission signal is

It becomes.

The correction matrix is calculated in the following steps.
(1) A transmission signal based on the baseband signal of pattern 1 is looped back from the transmitter 10 to the receiver 20. Specifically, in this embodiment, there is a path for inputting from the output of the adder 16 to the low noise amplifier 21 of the receiver 20.
(2) A transmission signal based on the baseband signal of pattern 2 is looped back from the transmitter 10 to the receiver 20 (however, the phase relationship between the local signals TxLo_I and TxLo_Q of the transmitter 10 is reversed).
(3) The correction matrix for the receiver 20 is calculated.
(4) A transmission signal based on the baseband signal of pattern 1 is looped back from the transmitter 10 to the receiver 20 (the receiver 20 side is assumed to have been corrected with a correction matrix).
(5) A transmission signal based on the baseband signal of pattern 3 is looped back from the transmitter 10 to the receiver 20 (the receiver 20 side is assumed to have been corrected with a correction matrix).
(6) A correction matrix on the transmitter 10 side is calculated.
Hereinafter, description will be given in the above order.

(1) A transmission signal based on the baseband signal of pattern 1 is looped back from the transmitter 10 to the receiver 20. Specifically, in this embodiment, there is a path for inputting from the output of the adder 16 to the low noise amplifier 21 of the receiver 20. The operation is shown in FIG.
Connect the transceiver so that the output signal Tx_out of the transmitter 10 is input signal Rx_in the receiver 20, the signal of the pattern 1 based on the data d _k of the k-th subcarrier, as shown in Figure 8, the transmitter 10 are input as baseband signals TxI ₁ and TxQ ₁ . The signal of pattern 1 is expressed by the following equation (26).

とし、送信機１０と受信機２０の間のループバック経路での利得と位相回転を

とすると、ループバックを経由した受信機２０の入力信号は、

となるので、受信機２０のＤＦＴ回路２６による離散フーリエ変換の後の信号は、

となる。The output signal of the transmitter 10

And gain and phase rotation in the loopback path between the transmitter 10 and the receiver 20

Then, the input signal of the receiver 20 via the loopback is

Therefore, the signal after the discrete Fourier transform by the DFT circuit 26 of the receiver 20 is

It becomes.

ただし、その際に、Ｉチャネルのミキサ１５Ｉに−sinω_cｔを、Qチャネルのミキサ１５Qにcosω_cｔを入力して、位相関係を逆転させる。(2) A transmission signal based on the baseband signal of pattern 2 is looped back from the transmitter 10 to the receiver 20. The baseband signals TxI ₂ and TxQ ₂ of pattern 2 are shown in equation (31). The operation is shown in FIG.

However, in this case, the -sinω _c t to the mixer 15I of I channel, enter the cos .omega _c t to the mixer 15Q of Q channel, to reverse the phase relationship.

となり、ループバックを経由した受信機２０の入力信号は、

となり、受信機２０のＤＦＴ回路２６による離散フーリエ変換の後の信号は、

となる。At this time, the output signal of the transmitter 10 is

The input signal of the receiver 20 via the loopback is

The signal after the discrete Fourier transform by the DFT circuit 26 of the receiver 20 is

It becomes.

となるので、

となり、

となる。なお、

であり、パワーを表す。(3) Calculation of receiver correction matrix The receiver correction matrix is calculated from the result obtained by inputting the baseband signals of pattern 1 and pattern 2. From equation (30) and equation (34),

So,

And

It becomes. In addition,

And represents power.

により、Ｋ_１Ｋ_２が求まる。このＫ_１Ｋ_２は

であるので、

により、ｈとψが求まり、式（７）からＫ_１とＫ_２が個々に求まるので、Ｋ_１、Ｋ_２、Ｋ^* _１、Ｋ^* _２からなる補正用行列２７を取得することができる。Therefore,

Thus, K ₁ K ₂ is obtained. This K ₁ K ₂ is

So

Thus, h and ψ are obtained, and K ₁ and K ₂ are obtained individually from the equation (7), so that the correction matrix 27 including K ₁ , K ₂ , K ^* ₁ , and K ^* ₂ can be obtained.

This completes the acquisition of the correction matrix 27 on the receiver 20 side. By applying this correction matrix 27 to FIG. 6, the correction unit 28 corrects the data arranged on each subcarrier on the receiver 20 side, and compensates for IQ mismatch. The gain error g and the phase error φ on the transmitter 10 side are canceled by the steps (1) to (3). That is, when looking at the equation (39), J ₁ (g and φ constituting it) disappears by a series of processes, and an equation of only K ₁ K ₂ has come out. ing. Since both pattern 1 and pattern 2 pass through the path from transmitter 10 to receiver 20, as can be seen from equations (30) and (34), the values R _k1 and R _k2 obtained from the output of receiver 20 are , J ₁ and J ₂ are included. Although only two equations of R _k 1 and R ^* _−k1 can be obtained in the equation (30), _variables of K ₁ , K ₂ , L _k J ₁ d _k , and L ^* _−k J ^* ₂ d _k Entered and cannot be obtained. Therefore, providing a pattern 2 (exchange of important point here is the local signal of the transmitter 10), without increasing the variable, it is possible to erase the ^{_{L * -k J * 2 d k}} . This is the first cancellation. FIG. 8 and FIG. 9 explain this in an easy-to-understand manner. Looking at the transmitter output Tx_out, the sign of ^{J *} ₂ d _k it has reversed in Figure 8 and Figure 9. When the expression (35) is reached, the variable L _k J ₁ d _k is deleted by the expressions (36) to (39), and this is the second cancellation.

(4) The transmission signal based on the baseband signal of the pattern 1 described above is looped back from the transmitter 10 to the receiver 20 (the receiver 20 side applies the previously obtained correction matrix).
On the receiver 20 side, the result after correction of the subcarrier of the receiver 20 is as follows.

(5) A transmission signal based on the baseband signal of pattern 3 is looped back from the transmitter 10 to the receiver 20 (the receiver 20 side applies the previously obtained correction matrix).
As baseband signals of pattern 3, the following baseband signals TxI ₃ and TxQ ₃ are input to the transmitter 10. This signal, in the baseband signal pattern 1 is obtained by polarity inverting the Q channel baseband signals TxQ _3. The operation is shown in FIG.

であり、ループバックを経由した受信機２０の入力信号は、

であるので、受信機のＤＦＴ回路２６による離散フーリエ変換後に更に補正したサブキャリア信号は、

となる。At this time, the output signal of the transmitter 10 is

The input signal of the receiver 20 via the loopback is

Therefore, the subcarrier signal further corrected after the discrete Fourier transform by the DFT circuit 26 of the receiver is

It becomes.

となり、

である。(6) The correction matrix 18 on the transmitter 10 side is calculated.
A correction matrix on the transmitter side is obtained using the output of the correction result after the discrete Fourier transform by the DFT circuit of the receiver 20 that has received the transmission signal based on the baseband signals of the pattern 1 and the pattern 3. From Equation (43) and Equation (47),

And

It is.

により、Ｊ_１Ｊ^* _２が求まる。このＪ_１Ｊ^* _２は

であるので、

により、ｇとφが求まり、式（３）からＪ_１とＪ_２が個々に求まるので、Ｊ_１、Ｊ_２、Ｊ^* _１、Ｊ^* _２からなる補正用行列１８を取得することができる。Therefore, from Equation (49) and Equation (50),

Thus, J ₁ J ^* ₂ is obtained. This J ₁ J ^* ₂ is

So

Thus, g and φ are obtained, and J ₁ and J ₂ are obtained individually from the equation (3), so that the correction matrix 18 composed of J ₁ , J ₂ , J ^* ₁ , J ^* ₂ can be obtained.

  The above operation is performed for each subcarrier, and acquisition of the correction matrix 18 on the transmitter 10 side is completed. By applying this correction matrix 18 to FIG. 7, the correction unit 19 performs pre-correction on the data arranged on each subcarrier on the transmitter 10 side, and compensates for the IQ mismatch on the transmitter side. .

  The correction matrices 27 and 18 are obtained in the memory when the RF transmitter / receiver is started up or during non-communication time. During actual communication, correction is performed on both the transmitter 10 side and the receiver 20 side using the correction matrices 27 and 18.

ただし、このときは、Ｑチャンネルローカル信号ＴｘＬｏ＿Ｑの極性を反転させる。このようにしても、送信機１０の出力信号は式（４５）で表されるので、第１の実施例と全く同様に、式（４７）のサブキャリア信号を得ることができる。<Second embodiment>
In the first embodiment described above, in step (5), it was used signal TxI _3, TxQ ₃ shown in equation (44) into a baseband signal of pattern 3, instead of the signal TxI _3, TxQ ₃ Thus, the signal of the pattern 3A shown in the following equation (44 ′) can be used. This signal is the same as the signal of pattern 1 shown in equation (26).

However, at this time, the polarity of the Q channel local signal TxLo_Q is inverted. Even in this case, since the output signal of the transmitter 10 is expressed by Expression (45), the subcarrier signal of Expression (47) can be obtained in the same manner as in the first embodiment.

<Third embodiment>
FIG. 11 shows the configuration of an RF transceiver that implements the first embodiment. The RF transmitting / receiving apparatus of the present embodiment adds a correction unit 19 shown in FIG. 6 and a correction unit 28 shown in FIG. 7 to the configuration of the transmitter 10 and the receiver 20 shown in FIG. A unit 31, a local signal generation unit 32, and a swap unit 33 are added. The control unit 31 includes a correction matrix estimation unit 27 that estimates the correction matrix 27 (FIG. 6) of the receiver 20, a correction matrix 18 (FIG. 7) of the transmitter 10, a correction matrix storage unit that stores the estimated correction matrix, A pattern in which a transmission I, Q channel baseband signal TxI, TxQ output from the discrete inverse Fourier transformer 11 is a signal of pattern 1, 2, 3 or 3A is generated as a subcarrier and input to the transmitter 10 A generation unit and a local signal control unit that controls the local signal generation unit 32 are provided. In the control unit 31, the processes of the steps (1) to (6) described above are performed. The local signal generation unit 32 generates an I channel local signal TxLo_I and a Q channel local signal TxLo_Q to be sent to the transmitter 10, and generates an I channel local signal Rxo_I and a Q channel local signal RxLo_Q to be sent to the receiver 20. When implementing the second embodiment, the polarity of the Q channel local signal TxLo_Q may be reversed. The swap unit 33 exchanges the I-channel local signal TxLo_I and the Q-channel local signal TxLo_Q on the transmitter 10 side during the processing in step (2).

  When implementing the second embodiment, the signal of the pattern 3A is input in step (5), and the polarity of the Q channel local signal TxLo_Q to the transmitter 10 output from the local signal generator 32 is input. Is reversed.

10: Transmitter, 11: Discrete inverse Fourier transformer, 12 (I, Q): DA converter, 13 (I, Q): Low-pass filter, 14 (I, Q): Amplifier, 15 (I, Q): Mixer, 16: adder, 17: amplifier, 18: correction matrix, 19: correction unit 20: receiver, 21: low noise amplifier, 22 (I, Q): mixer, 23 (I, Q): low-pass filter 24 (I, Q): amplifier, 25 (I, Q): AD converter, 26: discrete Fourier transformer, 27: correction matrix, 28: correction unit 31: control unit, 32: local signal generation unit, 33: Swap section

Claims (9)

In the transmitter, a plurality of subcarriers that are orthogonal to each other are inverse Fourier transformed to generate a transmission I channel baseband signal and a transmission Q channel baseband signal that are orthogonal to each other, and the transmission I channel baseband signal and the transmission I channel baseband signal The transmission Q channel baseband signal is quadrature modulated and then added and transmitted as an RF signal,
At the receiver, the RF signal is received and orthogonally demodulated to separate the received I channel baseband signal and the received Q channel baseband signal, and the received I channel baseband signal and the received Q channel baseband signal are Fourier transformed. An IQ mismatch correction method in an OFDM direct conversion RF transceiver that generates a plurality of subcarriers that are orthogonal to each other,
The first transmission of the first transmission baseband signal and the second transmission baseband signal of the transmitter orthogonal to each other obtained when data is input to the kth subcarrier of the transmitter. A pattern 1 signal is generated with a baseband signal as a transmission I channel baseband signal and the second transmission baseband signal as a transmission Q channel baseband signal, and the transmission I channel of the pattern 1 signal is generated The baseband signal is quadrature modulated with the I channel local signal of the transmitter, and the transmission Q channel baseband signal of the pattern 1 signal is quadrature modulated with the Q channel local signal of the transmitter, and both orthogonal modulation is performed. A first step of adding the received signals to the receiver via a loopback as an RF signal;
A pattern 2 signal is generated in which the first transmission baseband signal is a transmission Q channel baseband signal and the second transmission baseband signal is a transmission I channel baseband signal. The transmission I channel baseband signal is orthogonally modulated with the Q channel local signal, and the transmission Q channel baseband signal of the pattern 2 signal is orthogonally modulated with the I channel local signal, and both orthogonal modulation is performed. A second step of adding the signals and inputting them to the receiver via a loopback as an RF signal,
Based on the reception result of the receiver obtained in the first step and the reception result of the receiver obtained in the second step, the gain error and the phase error in the transmitter are canceled. An IQ mismatch characterized in that a gain matrix and a phase error are corrected with respect to the k-th subcarrier data obtained by the receiver based on the obtained correction matrix. Correction method. The IQ mismatch correction method according to claim 1,

To obtain K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes for the receiver, and correct the k-th subcarrier data of the receiver. IQ mismatch correction method.
However,
* Is a complex conjugate.
R _K1 sets the data arranged in the kth subcarrier of the transmitter to d _k , sets the transmission I channel baseband signal to TxI ₁ , sets the transmission Q channel baseband signal to TxQ _1, and sets the signal of the pattern 1 to ,

Then, the RF data output from the transmitter is input to the receiver via a loopback, and is received data obtained on the kth subcarrier.
R- _k1 is received data obtained on the _{-kth subcarrier} by the receiver due to IQ mismatch when the signal of the pattern 1 is used.
R _k2 is a transmission I channel baseband signal is TxI ₂ , a transmission Q channel baseband signal is TxQ _2, and the signal of the pattern 2 is

And the local signal of the transmitter is exchanged between the I channel and the Q channel, and when the RF signal output from the transmitter is input to the receiver via a loopback, the receiver receives the k-th sub-signal. Received data obtained on a carrier.
R- _k2 is received data obtained on the -k-th subcarrier at the receiver due to IQ mismatch when the pattern 2 signal is used and the local signal is switched between the I channel and the Q channel. is there.
h is a gain error in the receiver.
ψ is a phase error in the receiver.
Re [] indicates the real part.
Im [] indicates an imaginary part. The IQ mismatch correction method according to claim 1,
The data of the kth subcarrier of the receiver is corrected by the correction matrix of the receiver, and then the signal of the pattern 1 is generated, and the transmission I channel baseband of the signal of the pattern 1 is generated The signal is quadrature modulated with the I channel local signal of the transmitter, the transmission Q channel baseband signal of the signal of the pattern 1 is quadrature modulated with the Q channel local signal of the transmitter, and the signal subjected to bi-orthogonal modulation And adding to the receiver via a loopback as an RF signal;
After correcting the data of the kth subcarrier of the receiver by the correction matrix of the receiver, the first transmission baseband signal is used as a transmission I channel baseband signal, and the second transmission baseband is used. A pattern 3 signal is generated using a signal whose polarity is inverted as a transmission Q channel baseband signal, and the transmission I channel baseband signal of the pattern 3 signal is orthogonally modulated by the I channel local signal of the transmitter Then, the transmission Q channel baseband signal of the pattern 3 signal is quadrature modulated with the Q channel local signal of the transmitter, and both quadrature modulated signals are added together to receive the RF signal as a RF signal via a loopback. A fourth step to input to the machine,
Based on the reception result of the receiver obtained in the third step and the reception result of the receiver obtained in the fourth step, a correction matrix of the transmitter is obtained, and the obtained correction matrix is obtained. An IQ mismatch correction method, wherein a gain error and a phase error are corrected with respect to data of the k-th subcarrier of the transmitter by a matrix. The IQ mismatch correction method according to claim 2,
Correction for K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes for the receiver is performed on the data of the k-th subcarrier of the receiver;

To obtain J ₁ , J ₂ , J ^* ₁ , and J ^* ₂ as correction matrixes for the transmitter, and perform correction on the subcarrier data of the transmitter. Method.
However,
RC _k1 uses the signal of the pattern 1 at the transmitter, and performs correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes of the receiver and data of each subcarrier of the receiver The received data obtained for the kth subcarrier by the receiver when the above is performed.
RC _k3 is a transmission I channel baseband signal as TxI _3, the transmission Q channel baseband signal and TxQ _3, the signal pattern 3

As described above, an RF signal output from the transmitter is input to the receiver via a loopback, and correction using K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as a correction matrix of the receiver is performed on the receiver. The received data obtained for the k-th subcarrier by the receiver when the data of each subcarrier is obtained.
g is a gain error in the transmitter.
φ is a phase error in the transmitter. The IQ mismatch correction method according to claim 1,
The data of the kth subcarrier of the receiver is corrected by the correction matrix of the receiver, and then the signal of the pattern 1 is generated, and the transmission I channel baseband of the signal of the pattern 1 is generated The signal is quadrature modulated with the I channel local signal of the transmitter, the transmission Q channel baseband signal of the signal of the pattern 1 is quadrature modulated with the Q channel local signal of the transmitter, and the signal subjected to bi-orthogonal modulation And the step 3A for inputting to the receiver via a loopback as an RF signal,
After correcting the data of the kth subcarrier of the receiver by the correction matrix of the receiver, the first transmission baseband signal is used as a transmission I channel baseband signal, and the second transmission baseband is used. A signal of a pattern 3A using a signal as a transmission Q channel baseband signal is generated, and the transmission I channel baseband signal among the signals of the pattern 3A is orthogonally modulated with an I channel local signal of the transmitter, and the pattern 3A The transmission Q channel baseband signal is quadrature modulated with a signal obtained by inverting the polarity of the Q channel local signal of the transmitter, and the signals obtained by bi-orthogonal modulation are added together to obtain the RF signal as a RF signal via a loopback. 4A step to input to the machine,
Based on the reception result of the receiver obtained in the step 3A and the reception result of the receiver obtained in the step 4A, a correction matrix of the transmitter is obtained, and the obtained correction matrix is obtained. An IQ mismatch correction method, wherein a gain error and a phase error are corrected with respect to data of the k-th subcarrier of the transmitter by a matrix. The IQ mismatch correction method according to claim 2,
Correction for K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes for the receiver is performed on the data of the k-th subcarrier of the receiver;

To obtain J ₁ , J ₂ , J ^* ₁ , and J ^* ₂ as correction matrixes for the transmitter, and perform correction on the subcarrier data of the transmitter. Method.
However,
RC _k1 uses the signal of the pattern 1 at the transmitter, and performs correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as correction matrixes of the receiver and data of each subcarrier of the receiver The received data obtained for the kth subcarrier by the receiver when the above is performed.
RC _k3 is a transmission I channel baseband signal as TxI _3, the transmission Q channel baseband signal and TxQ _3, the signal pattern 3A

And the polarity of the local signal of the Q channel among the local signals of the transmitter is inverted, and the RF signal output from the transmitter is input to the receiver via a loopback, for correction of the receiver This is the received data obtained for the kth subcarrier in the receiver when correction by K ₁ , K ₂ , K ^* ₁ , K ^* ₂ as a matrix is performed on the data of each subcarrier of the receiver. .
g is a gain error in the transmitter.
φ is a phase error in the transmitter. Inverse Fourier transform means for inputting a plurality of subcarriers that are orthogonal to each other and performing inverse Fourier transform to generate a transmission I channel baseband signal and a transmission Q channel baseband signal that are orthogonal to each other, and the transmission I channel base A quadrature modulation unit that quadrature modulates a band signal with an I channel local signal and orthogonally modulates the transmission Q channel baseband signal with a Q channel local signal, and adds both signals after the quadrature modulation by the quadrature modulation unit A transmitter comprising adding means for converting to an RF signal;
Orthogonal demodulation means that separates the received I-channel baseband signal and the received Q-channel baseband signal by receiving and orthogonally demodulating the RF signal, and Fourier transforming the received I-channel baseband signal and the received Q-channel baseband signal A receiver comprising Fourier transform means for generating a plurality of subcarriers orthogonal to each other;
In an OFDM direct conversion type RF transceiver having
The first transmission of the first transmission baseband signal and the second transmission baseband signal of the transmitter orthogonal to each other obtained when data is input to the kth subcarrier of the transmitter. A pattern 1 signal is generated with a baseband signal as a transmission I channel baseband signal and the second transmission baseband signal as a transmission Q channel baseband signal, and the transmission I channel of the pattern 1 signal is generated The baseband signal is quadrature modulated with the I channel local signal of the transmitter, and the transmission Q channel baseband signal of the pattern 1 signal is quadrature modulated with the Q channel local signal of the transmitter, and both orthogonal modulation is performed. First control means for adding the received signals and inputting them as RF signals to the receiver via a loopback;
A pattern 2 signal is generated in which the first transmission baseband signal is a transmission Q channel baseband signal and the second transmission baseband signal is a transmission I channel baseband signal. The transmission I channel baseband signal is orthogonally modulated with the Q channel local signal, and the transmission Q channel baseband signal of the pattern 2 signal is orthogonally modulated with the I channel local signal, and both orthogonal modulation is performed. A second control means for adding the signals and inputting them as RF signals to the receiver via a loopback;
Based on the reception result of the receiver obtained by the first control means and the reception result of the receiver obtained by the second control means, the gain error and phase error at the transmitter are canceled. In addition, a correction matrix of the receiver is obtained, and a gain error and a phase error are corrected with respect to the k-th subcarrier data obtained by the receiver using the obtained correction matrix. RF transceiver. The RF transmitting / receiving apparatus according to claim 1.
The data of the kth subcarrier of the receiver is corrected by the correction matrix of the receiver, and then the signal of the pattern 1 is generated, and the transmission I channel baseband of the signal of the pattern 1 is generated The signal is quadrature modulated with the I channel local signal of the transmitter, the transmission Q channel baseband signal of the signal of the pattern 1 is quadrature modulated with the Q channel local signal of the transmitter, and the signal subjected to bi-orthogonal modulation And a third control means for inputting to the receiver via a loopback as an RF signal;
After correcting the data of the kth subcarrier of the receiver by the correction matrix of the receiver, the first transmission baseband signal is used as a transmission I channel baseband signal, and the second transmission baseband is used. A pattern 3 signal is generated using a signal whose polarity is inverted as a transmission Q channel baseband signal, and the transmission I channel baseband signal of the pattern 3 signal is orthogonally modulated by the I channel local signal of the transmitter Then, the transmission Q channel baseband signal of the pattern 3 signal is quadrature modulated with the Q channel local signal of the transmitter, and both quadrature modulated signals are added together to receive the RF signal as a RF signal via a loopback. A fourth control means for inputting to the machine,
Based on the reception result of the receiver obtained by the third control means and the reception result of the receiver obtained by the fourth control means, a correction matrix of the transmitter is obtained and the obtained An RF transmitting / receiving apparatus, wherein a gain error and a phase error are corrected with respect to the k-th subcarrier data of the transmitter by a correction matrix. The RF transmitting / receiving apparatus according to claim 1.
The data of the kth subcarrier of the receiver is corrected by the correction matrix of the receiver, and then the signal of the pattern 1 is generated, and the transmission I channel baseband of the signal of the pattern 1 is generated The signal is quadrature modulated with the I channel local signal of the transmitter, the transmission Q channel baseband signal of the signal of the pattern 1 is quadrature modulated with the Q channel local signal of the transmitter, and the signal subjected to bi-orthogonal modulation 3A control means for adding to the receiver via a loopback as an RF signal;
After correcting the data of the kth subcarrier of the receiver by the correction matrix of the receiver, the first transmission baseband signal is used as a transmission I channel baseband signal, and the second transmission baseband is used. A signal of a pattern 3A using a signal as a transmission Q channel baseband signal is generated, and the transmission I channel baseband signal among the signals of the pattern 3A is orthogonally modulated with an I channel local signal of the transmitter, and the pattern 3A The transmission Q channel baseband signal is quadrature modulated with a signal obtained by inverting the polarity of the Q channel local signal of the transmitter, and the signals obtained by bi-orthogonal modulation are added together to obtain the RF signal as a RF signal via a loopback. 4A control means to be input to the machine,
Based on the reception result of the receiver obtained by the control means of the 3A and the reception result of the receiver obtained by the control means of the 4A, the correction matrix of the transmitter is obtained and the obtained An RF transmitting / receiving apparatus, wherein a gain error and a phase error are corrected with respect to the k-th subcarrier data of the transmitter by a correction matrix.

Images