JP5965372B2 - Communication system and optical signal transmission method - Google Patents

Description

  The present invention relates to a technique for compressing a digital RoF signal transmitted between BBU and RRU.

In a cellular system, in order to improve the degree of freedom of cell configuration, a base station function is physically separated by dividing it into a signal processing unit (BBU: Base Band Unit) and an RF unit (RRU: Remote Radio Unit). It is considered to do. At this time, the radio signal is transmitted between the BBU and the RRU by the RoF technology. RoF technology can be broadly divided into analog RoF technology and digital RoF technology depending on the optical transmission method. Recently, digital RoF technology with excellent transmission quality has been actively studied, and standardization such as CPRI (Common Public Radio Interface) [1]. The specification is being developed under the organization. As a connection medium between BBU and RRU, a coaxial cable, an optical fiber, or the like is used. In particular, the transmission distance can be increased by connecting with an optical fiber.

  One BBU can accommodate a plurality of RRUs. Thereby, BBU required for each RRU can be collected into one, and it becomes possible to reduce an operation cost and an installation cost. As an example of such a form, a form in which BBU-RRU is connected by a PON (Passive Optical Network) system has been proposed. As a PON signal multiplexing method, TDM (Time Division Multiplex), WDM (Wavelength Division Multiplex), FDM (Frequency Division Multiplex), or the like can be selected.

  Hereinafter, the digital RoF transmission technology between BBU and RRU is referred to as related technology. Also, a digital signal (IQ data) for each of the I-axis and Q-axis of the radio signal created by the BBU is converted to an optical signal and transmitted to the RRU, and the optical signal received by the RRU is converted to a radio signal and transmitted to the radio terminal. The link to transmit is called a downlink. On the other hand, a link that receives a radio signal transmitted by a radio terminal by RRU, converts the received radio signal to an optical signal and transmits it to the BBU, converts the optical signal received by the BBU to IQ data, and demodulates the signal Is called uplink.

  FIG. 10 shows an example of the RRU apparatus configuration related to the present invention. For uplink signal processing, the RRU 91 includes an antenna 11 that transmits / receives a radio signal, a transmission / reception switching unit 12 that switches transmission / reception, and an amplifier that amplifies the signal power of the received radio signal to a level that allows signal processing. 13, a down-conversion unit 14 that down-converts a radio signal to baseband, an A / D conversion unit 15 that converts the down-converted analog signal into IQ data, and a baseband filter that performs filtering processing on the IQ data Unit (upstream) 16, a frame conversion unit 17 that multiplexes IQ data and a control signal between BBU 93 and RRU 91, and an electro-optical (E / O) conversion unit 18 that converts an electrical signal into an optical signal and transmits it. . The transmission / reception switching unit 12 is compatible with both FDD (Frequency Division Duplex) and TDD (Time Division Duplex).

  Further, the RRU 91 performs an optical signal received from the BBU 93 to an electrical signal for the downlink signal processing, and an optical-electrical (O / E) conversion unit 28 for converting the optical signal received from the BBU 93 into a control signal and IQ data between the BBU 93 and the RRU 91 from the received signal. A frame conversion unit 27 to be extracted, a baseband filter unit (downlink) 26 that performs filtering processing on IQ data, a D / A conversion unit 25 that converts IQ data into an analog signal, and an up-conversion that upconverts the analog signal Unit 24, amplifier 23 that amplifies the power to a predetermined transmission power, transmission / reception switching unit 12, and antenna 11.

  FIG. 11 shows a device configuration example of BBU related to the present invention. For uplink signal processing, the BBU 93 demodulates the IQ data, an O / E converter 31 that converts an optical signal into an electrical signal, a frame converter 32 that extracts a control signal and IQ data from a received signal, and the IQ data. A modem unit 33 is included. Further, for downlink signal processing, the BBU 93 is a modem 33 that outputs IQ data of a radio signal, a frame converter 34 that multiplexes IQ data and a control signal, and an electrical signal that is converted into an optical signal and transmitted. / O converter 35 is provided.

  Taking an example of transmitting a Long Term Evolution (LTE) signal using CPRI, a sampling frequency of 30.72 MHz is used for a system with a system bandwidth of 20 MHz, and digital sampling for each of the I axis and the Q axis. The number of quantization bits of 4 to 20 bits for the upstream signal and 8 to 20 bits for the downstream signal is applied. The frame converter 34 inserts a control signal into 1/16 of the entire frame, and further transmits the signal after 8B / 10B encoding. Therefore, a very wide bandwidth is required in the optical fiber section for digital RoF transmission.

  As a technique for reducing the required bandwidth for digital RoF transmission, there is a technique for reducing the number of quantization bits using predictive coding. In this method, the amplitude value of the next sample point is predicted from several sample points of the radio signal. FIG. 12 shows an example of prediction operation. In FIG. 12, the value of the next sample point is predicted from the values of the past three sample points. In the linear prediction method, which is one of the prediction methods, a value obtained by multiplying and pasting several past sample points by a coefficient is used as a predicted value of the next sample point. The coefficient at this time is hereinafter referred to as a prediction coefficient. The prediction coefficient is set so that an error signal, which is the difference between the actual sample point value and the prediction value, becomes small. The higher the prediction accuracy, the closer the amplitude value of the error signal is to zero.

  FIG. 13 shows an example of the probability density function of the error signal. In FIG. 13, for the OFDM (Orthogonal Frequency Division Multiplexing) radio signal based on the parameters of the LTE PHY layer, the predicted value of the next sample point is calculated at the past 32 sample points. The prediction coefficient was determined so that the Euclidean norm of the error signal was minimized every 512 sample points. From FIG. 13, it can be seen that the probability value of the amplitude value of the error signal is higher near zero. Therefore, if this error signal is encoded by entropy encoding such as Huffman encoding, an amplitude value having a higher appearance probability can be transmitted with a smaller number of bits, so that the required bandwidth for digital RoF transmission can be reduced.

CPRI, “CPRI Specification V5.0”, Sep. , 2011, http: // www. cpri. info / spec. html

  In the method of entropy encoding and transmitting the related error signal, the number of bits after encoding is not constant, so the required bandwidth for digital RoF transmission is not constant. For this reason, it is necessary to set the communication capacity in the optical fiber section to a value with a margin for the maximum required band for digital RoF transmission and the average required band for digital RoF transmission, leading to an increase in the required communication capacity.

  An object of the present invention is to reduce the number of quantization bits of a digital RoF signal transmitted between BBU and RRU, and to suppress an increase in required communication capacity.

  In order to achieve the above object, in the present invention, the error signal is nonlinearly quantized and transmitted, thereby reducing the number of quantization bits compared to the case of using a fixed quantization step, and the required bandwidth for digital RoF transmission. Keep constant.

Specifically, the communication system according to the present invention is:
A communication system in which an antenna unit that transmits and receives radio signals and a signal processing unit that controls the antenna unit are connected using an optical transmission line,
The antenna unit and the signal processing unit are
A prediction coefficient determination unit that calculates a prediction coefficient using a plurality of sample points of transmission data;
An error prediction unit that generates an error signal from the difference between the value of the next sample point calculated using the prediction coefficient and the actual value of the next sample point of the transmission data;
A non-linear quantization unit that quantizes the error signal with a non-linear quantization step size;
An error signal transmission unit that transmits the error signal quantized by the nonlinear quantization unit to the optical transmission line;
A quantization step restoring unit that converts the quantization step size of the error signal transmitted by the error signal transmission unit into a uniform quantization step size;
A signal restoration unit that restores the error signal whose quantization step size has been restored by the quantization step restoration unit to a sample point of transmission data using the prediction coefficient calculated by the prediction coefficient determination unit;
Equipped with a,
The nonlinear quantization unit calculates a quantization error that occurs when the error signal is quantized and outputs the quantization error to the error prediction unit,
The error prediction unit predicts the value of the next sample point using the value obtained by adding the quantization error to the sample point of the transmission data and the prediction coefficient, and calculates an error signal;
The shall be the feature.

  In the communication system according to the present invention, the nonlinear quantization unit may change a quantization step size according to an average amplitude of the radio signal.

  In the communication system according to the present invention, the nonlinear quantization unit may estimate the characteristics of the radio signal from radio band allocation information and change the quantization step size according to the estimated characteristics of the radio signal.

  In the communication system according to the present invention, the characteristic of the radio signal may be the number of subcarriers used in the radio signal.

  In the communication system according to the present invention, the characteristic of the wireless signal may be the number of wireless terminals that are communicating by transmitting and receiving the wireless signal.

Specifically, the optical signal transmission method according to the present invention includes:
An optical signal transmission method of a communication system in which an antenna unit that transmits and receives a radio signal and a signal processing unit that controls the antenna unit are connected using an optical transmission line,
In the antenna unit and the signal processing unit,
A prediction coefficient determination procedure for calculating a prediction coefficient using a plurality of sample points of transmission data;
A quantization error determination procedure for calculating a quantization error generated when the error signal is quantized;
The quantization error using the value and the prediction coefficients by adding the sample points of the transmitted data, an error prediction step of generating an error signal to predict the value of the next sample point,
A non-linear quantization procedure for quantizing the error signal with a non-linear quantization step size;
An error signal transmission procedure for transmitting the error signal quantized by the nonlinear quantization procedure to the optical transmission line;
Have

Specifically, the optical signal transmission method according to the present invention includes:
An optical signal transmission method of a communication system in which an antenna unit that transmits and receives a radio signal and a signal processing unit that controls the antenna unit are connected using an optical transmission line,
In the antenna unit and the signal processing unit,
A quantization step restoration procedure for returning the quantization step size of the error signal transmitted through the transmission line to a uniform quantization step size;
A signal restoration procedure for restoring the error signal whose quantization step size is restored in the quantization step restoration procedure to a sample point of transmission data using a prediction coefficient received together with the error signal transmitted through the transmission path;
I have a,
The error signal is calculated by predicting the value of the next sample point using the value obtained by adding the quantization error generated when the error signal is quantized to the sample point of the transmission data and the prediction coefficient. That
It is characterized by.

  The above inventions can be combined as much as possible.

  According to the present invention, the number of quantization bits can be reduced as compared with the case where a fixed quantization step is used. Also, since the required bandwidth for digital RoF transmission is constant, the required communication capacity in the optical fiber section is reduced as compared with the related art.

2 shows an example of nonlinear quantization according to the present invention. 2 shows a device configuration example of an RRU 91 according to the first embodiment. 2 shows a device configuration example of an RRU 91 according to the first embodiment. The apparatus structural example of RRU91 of Embodiment 2 is shown. The apparatus structural example of RRU91 of Embodiment 2 is shown. The apparatus structural example of RRU91 of Embodiment 3 is shown. The apparatus structural example of BBU93 of Embodiment 3 is shown. The apparatus structural example of RRU91 of Embodiment 4 is shown. The apparatus structural example of BBU93 of Embodiment 4 is shown. The apparatus structural example of RRU91 relevant to this invention is shown. The apparatus structural example of BBU93 relevant to this invention is shown. An example of prediction operation is shown. An example of the probability density function of an error signal is shown.

  Embodiments of the present invention will be described with reference to the accompanying drawings. The embodiments described below are examples of the present invention, and the present invention is not limited to the following embodiments. In the present specification and drawings, the same reference numerals denote the same components.

  The invention according to the present embodiment includes a nonlinear quantization unit on the transmission side and a quantization step restoration unit on the reception side. The nonlinear quantization unit nonlinearly quantizes the error signal. FIG. 1 shows an example of nonlinear quantization according to the present invention. In the present embodiment, nonlinear quantization means that the error signal is quantized with a quantization step size corresponding to the amplitude of the error signal. The quantization step size varies depending on the amplitude error. Data with a small amplitude error reduces the quantization step size, and data with a large amplitude error increases the quantization step size. Then, the quantization step restoration unit on the reception side restores the step size of the received error signal to the linear quantization step size. That is, the nonlinear step size is converted into a linear quantization step size in which the quantization step is constant regardless of the amplitude of the error signal.

(Embodiment 1)
FIG. 2 shows a device configuration example of the RRU 91 according to the first embodiment. Unlike the related technology, the RRU 91 uses a prediction coefficient extraction unit 43 that extracts a prediction coefficient multiplexed on an error signal, and linearly quantizes the data quantized nonlinearly on the BBU 93 side, unlike the related art. A quantization step restoration unit 41 to be returned and a signal restoration unit 42 to restore the original IQ data by calculating a prediction value of the next sample point based on the notified prediction coefficient and adding an error signal.

  Unlike the related art, the RRU 91 generates a prediction coefficient based on IQ data, calculates an error signal based on the prediction coefficient, and outputs a prediction coefficient and an error signal for uplink signal processing. And an error prediction unit 44 and a nonlinear quantization unit 45 for nonlinearly quantizing the error signal. The prediction coefficient determination and error prediction unit 44 may be separated into a prediction coefficient determination unit that determines a prediction coefficient and an error prediction unit that outputs an error signal. However, in this embodiment, an example of one configuration is used. explain. The prediction coefficient determined by the prediction coefficient determination and error prediction unit 44 is output to the frame conversion unit 17. The frame conversion unit 17 multiplexes the prediction coefficient with the error signal. Thereby, the prediction coefficient is notified to the BBU 93 side.

  An example of the apparatus configuration of the BBU 93 of the first embodiment is shown in FIG. Unlike the related art, the BBU 93 generates a prediction coefficient based on IQ data, calculates an error signal based on the prediction coefficient, and outputs a prediction coefficient and an error signal for downlink signal processing. And an error prediction unit 54 and a nonlinear quantization unit 55 that nonlinearly quantizes the error signal. The prediction coefficient determined by the prediction coefficient determination and error prediction unit 54 is output to the frame conversion unit 34. The frame conversion unit 34 multiplexes the prediction coefficient with the error signal. Thereby, the prediction coefficient is notified to the RRU 91 side.

  The BBU 93 is different from the related art in that it performs uplink signal processing, and a prediction coefficient extraction unit 53 that extracts a prediction coefficient multiplexed on an error signal, and linearly quantizes the data quantized nonlinearly on the RRU 91 side. A quantization step restoration unit 51 to be returned and a signal restoration unit 52 to calculate the predicted value of the next sample point based on the notified prediction coefficient and add the error signal to restore the original IQ data.

  The RRU 91 functions as an antenna unit, and the BBU 93 functions as a signal processing unit. In the downlink, the frame conversion unit 34, the E / O conversion unit 35, the O / E conversion unit 28, and the frame conversion unit 27 function as an error signal transmission unit. In the uplink, the frame conversion unit 17, the E / O conversion unit 18, the O / E conversion unit 31, and the frame conversion unit 32 function as an error signal transmission unit.

In the optical signal transmission method according to the present embodiment, after downlink signal processing, the BBU 93 sequentially executes a prediction coefficient determination procedure, an error prediction procedure, a nonlinear quantization procedure, and an error signal transmission procedure. , RRU 91 sequentially executes a quantization step restoration procedure and a signal restoration procedure.
In the prediction coefficient determination procedure, the prediction coefficient determination and error prediction unit 54 calculates a prediction coefficient using a plurality of sample points of IQ data.
In the error prediction procedure, the prediction coefficient determination and error prediction unit 54 generates an error signal from the difference between the value of the next sample point calculated using the prediction coefficient and the actual value of the next sample point.
In the nonlinear quantization procedure, the nonlinear quantization unit 55 quantizes the error signal with a nonlinear quantization step size corresponding to the amplitude of the error signal.
In the error signal transmission procedure, the frame converter 34 transmits an error signal to the optical fiber, and the frame converter 27 receives the error signal. At this time, the prediction coefficient extraction unit 43 extracts a prediction coefficient from the signal received by the frame conversion unit 27 and outputs the prediction coefficient to the signal restoration unit 42.
In the quantization step restoration procedure, the quantization step restoration unit 41 converts the step size of the error signal into a linear quantization step size in which the quantization step is constant regardless of the amplitude of the error signal.
In the signal restoration procedure, the signal restoration unit 42 restores the error signal whose step size has been restored to the sample point of IQ data using the prediction coefficient calculated by the prediction coefficient determination and error prediction unit 54.

In the optical signal transmission method according to the present embodiment, after uplink signal processing, the RRU 91 sequentially executes a prediction coefficient determination procedure, an error prediction procedure, a nonlinear quantization procedure, and an error signal transmission procedure. , BBU 93 sequentially executes a quantization step restoration procedure and a signal restoration procedure.
In the prediction coefficient determination procedure, the prediction coefficient determination and error prediction unit 44 calculates a prediction coefficient using a plurality of sample points of IQ data.
In the error prediction procedure, the prediction coefficient determination and error prediction unit 44 generates an error signal from the difference between the value of the next sample point calculated using the prediction coefficient and the actual value of the next sample point.
In the nonlinear quantization procedure, the nonlinear quantization unit 45 quantizes the error signal with a nonlinear quantization step size corresponding to the amplitude of the error signal.
In the error signal transmission procedure, the frame conversion unit 17 transmits an error signal to the optical fiber, and the frame conversion unit 32 receives the error signal. At this time, the prediction coefficient extraction unit 53 extracts a prediction coefficient from the signal received by the frame conversion unit 32 and outputs the prediction coefficient to the signal restoration unit 52.
In the quantization step restoration procedure, the quantization step restoration unit 51 converts the step size of the error signal into a linear quantization step size in which the quantization step is constant regardless of the amplitude of the error signal.
In the signal restoration procedure, the signal restoration unit 52 restores the error signal whose step size has been restored to the sample point of the IQ data using the prediction coefficient determined by the prediction coefficient determination and error prediction unit 44.

  Here, the calculation method of the prediction coefficient in the prediction coefficient determination and error prediction units 44 and 54 is not limited. For example, the prediction coefficient can be obtained by the least square method or the like.

  If the number of bits used for nonlinear quantization is kept constant, the required bandwidth for digital RoF transmission is always constant. The prediction coefficient determination and error prediction units 44 and 54 cannot calculate the predicted value of the first sample point. The predicted value of the first sample point may be transmitted as an error signal as it is, or may be transmitted by quantizing only the first sample value with another quantization step / number of quantization bits. .

Prediction coefficients is a _1, the last sample point over prediction coefficients a ₁ as an example the case of calculating a predicted value of the next sample point is described a specific example of an operation of the signal restorer 42 and 52.
Predictive coefficient determination and error prediction unit 44 and 54, the _a 1 _{S i} as a prediction value of _{S i + 1} multiplied by the prediction coefficient _{a 1} in the i th sample point _{S i,} the error signal _{(S i + 1 -a 1 S} i) Is output. By performing non-linear quantization on the error signal, S _{i + 1} −a ₁ S _i is transmitted.

The quantization step restoration unit 41 converts the quantization step of the error signal (S _{i + 1} −a ₁ S _i ) into linear quantization. The signal restoration units 42 and 52 add the error signal by applying the prediction coefficient a ₁ to the i-th sample point S _i ′ restored so far, and output a ₁ S _i ′ + S _{i + 1} −a ₁ S _i . . Assuming that the i-th sample point can be accurately restored, since S _i = S _i ′, S _{i + 1} becomes the i + 1-th sample value, and the sample value can be restored.

  As described above, the invention according to this embodiment transmits an error signal after nonlinear quantization. As a result, the number of quantization bits can be reduced compared to the case where a fixed quantization step is used, and the required bandwidth for digital RoF transmission can be kept constant.

(Embodiment 2)
The smaller the number of bits after nonlinear quantization, the more the required bandwidth for digital RoF transmission can be reduced, but the quantization error of the error signal increases. At this time, the prediction coefficient determination and error prediction units 44 and 54 predict the error based on the sample value of the radio signal not including the quantization error, whereas the signal restoration units 42 and 52 include the quantization error. The error is predicted based on the input signal. In the second embodiment, the prediction coefficient determination and error prediction units 44 and 54 predict errors including nonlinear quantization errors to improve the accuracy of signal restoration.

  FIG. 4 shows a device configuration example of the RRU 91 of the second embodiment. In the present embodiment, a prediction coefficient determination unit 441 and an error prediction unit 442 are provided instead of the prediction coefficient determination and error prediction unit 44. FIG. 5 shows a device configuration example of the BBU 93 of the second embodiment. In the present embodiment, a prediction coefficient determination unit 541 and an error prediction unit 542 are provided instead of the prediction coefficient determination and error prediction unit 54. In either case, after the prediction coefficient is determined by the prediction coefficient determination units 441 and 541, the error prediction unit 442 and 542 calculates an error signal based on the prediction coefficient.

  It is assumed that the error prediction units 442 and 542 predict the value of the next sample point based on the past N sample points. At this time, the non-linear quantization units 45 and 55 calculate the quantization error included in the error signal at the i + 1 to i + Nth (i = 0, 1, 2,...) Sample points and feed back to the error prediction units 442 and 542. The error prediction units 442 and 542 predict errors of the (i + N + 1) th and subsequent sample points using values obtained by adding the quantization errors to the signals of the (i + 1) to i + Nth sample points.

Prediction coefficients is a _1, by multiplying the coefficients a ₁ to a value of the past sampling point as an example a case of predicting a next sample value, describing the operation of the second embodiment.
The error prediction unit 442 and 542 of the first embodiment, the i-th _a 1 sample points obtained by multiplying the coefficient _{S i} of _{S i} as a prediction value of _{S i + 1,} and outputs an error signal _(S i + 1 _-a 1 _{S i)} . Nonlinear quantization is performed on the error signal. At this time, when a quantization error q _i has occurred, S _{i + 1} −a ₁ S _i + q _i is transmitted as an error signal. The signal restoration units 42 and 52 add the error signal by multiplying the i-th sample value S _i ′ restored so far by a coefficient, and output S _{i + 1} −a ₁ S _i + a ₁ S _i ′ + q _i . Assuming that the i-th sample value can be accurately restored, S _i = S _i ′, so that S _{i + 1} + q _i becomes the i + 1-th sample value.

In addition, when a quantization error q _{i + 1} occurs when the error signal of the i + 2nd sample point is nonlinearly quantized, S _{i + 2} −a ₁ S _{i + 1} + q _{i + 1} is output and transmitted. The signal restoration units 42 and 52 output S _{i + 2} −a ₁ S _{i + 1} + a ₁ (S _{i + 1} + q _i ) + q _{i + 1} = S _{i + 2} + q _i + q _{i + 1} as the _{i + 2nd} sample value.

Therefore, when the quantization error q _i occurs, the signal restored by the signal restoration units 42 and 52 of the first embodiment also includes a quantization error when the error signal of the past sample point is nonlinearly quantized. It is listed. That is, each quantization error in nonlinear quantization is added to the signals restored by the signal restoration units 42 and 52 thereafter.

On the other hand, the second embodiment will be described. As before, i-th _S i _{+ 1} was nonlinear quantization error signal sample point -a ₁ S _i + _q i is transmitted, the signal restoration unit 42 and 52 _{S i} + 1 + _{q i} is i + 1-th sample value Is output as In this case, the nonlinear quantization units 45 and 55 output the quantization error q _i to the error prediction units 442 and 542. In the error prediction units 442 and 542, when the error signal of the i + 2 sample point is predicted, the quantization error q _i when the error signal of the i + 1 sample point is nonlinearly quantized is added to the i + 1 sample value. And S _{i + 2} −a ₁ (S _i + q _i ) is output. Then, nonlinear quantization is performed, and S _{i + 2} −a ₁ (S _{i + 1} + q _i ) + q _{i + 1} ′ is transmitted. The signal restoration units 42 and 52 output S _{i + 2} −a ₁ (S _{i + 1} + q _i ) + a ₁ (S _{i + 1} + q _i ) + q _{i + 1} ′ = S _{i + 2} + q _{i + 1} ′ as the _{i +} 2th sample point. Therefore, the signal restored by the signal restoration units 42 and 52 does not include a quantization error when the error of the past sample point is nonlinearly quantized.

  Therefore, in the communication system according to the present embodiment, even when a quantization error occurs, the quantization when the error of the past sample points to the signal restored by the signal restoration units 42 and 52 is nonlinearly quantized. The influence of the conversion error can be eliminated.

(Embodiment 3)
The average amplitude value of the error signal varies according to the average power or average amplitude value of the IQ data. Therefore, the number of bits after nonlinear quantization can be reduced by changing the quantization step size at the time of nonlinear quantization according to the average power or average amplitude value of the radio signal. Thus, in the third embodiment, the signal power of IQ data is estimated, and the quantization step size at the time of nonlinear quantization is changed according to the estimated signal power.

  FIG. 6 shows a device configuration example of the RRU 91 according to the third embodiment. Compared to the first embodiment, the RRU 91 includes a quantization parameter information extraction unit (downlink) 46 that extracts a quantization parameter transmitted from the BBU 93 for downlink signal processing. The quantization parameter obtained by the quantization parameter information extraction unit (downlink) 46 is input to the quantization step restoration unit 41 to change the quantization parameter. Here, the quantization parameter refers to the quantization step size. The uplink apparatus configuration of the RRU 91 is the same as that of the first embodiment.

  FIG. 7 shows a device configuration example of the BBU 93 of the third embodiment. Compared to the first embodiment, the BBU 93 includes a quantization parameter determination unit (downlink) 56 that determines a quantization parameter from the average amplitude value or average power value of IQ data for downlink signal processing. The quantization parameter obtained by the quantization parameter determination unit (downlink) 56 is input to the nonlinear quantization unit 55, and the quantization step size of the nonlinear quantization unit 55 is changed. The quantization parameter determined by the quantization parameter determination unit 56 is transmitted to the RRU 91 side. The uplink apparatus configuration of the BBU 93 is the same as that of the first embodiment.

  The communication system according to the present embodiment can operate in combination with the invention of the second embodiment.

  When the amplifier 13 controls the power value of the radio signal received by the RRU 91 to be constant, the third embodiment cannot be applied to the uplink. If the power value of the radio signal received by the RRU 91 is not constant, the third embodiment can also be applied to the uplink.

(Embodiment 4)
In the third embodiment, the average power or average amplitude value of the radio signal is estimated from the IQ data. In the fourth embodiment, the characteristics of a radio signal are estimated from the radio band allocation information, and the quantization step size at the time of nonlinear quantization is changed according to the estimated characteristics of the radio signal.

  FIG. 8 shows a device configuration example of the RRU of the fourth embodiment. Compared to the first embodiment, the RRU 91 includes a quantization parameter information extraction unit (downlink) 46 that extracts a quantization parameter transmitted from the BBU 93 for downlink signal processing. The quantization parameter obtained by the quantization parameter information extraction unit (downlink) 46 is input to the quantization step restoration unit 41 to adjust the quantization step size.

  Compared to the first embodiment, the RRU 93 includes a quantization parameter information extraction unit (uplink) 47 that extracts a quantization parameter transmitted from the BBU 93 for uplink signal processing. The quantization parameter obtained by the quantization parameter information extraction unit (upstream) 47 is input to the nonlinear quantization unit 45 to change the quantization step size of the nonlinear quantization unit.

  FIG. 9 shows a device configuration example of the BBU 93 of the fourth embodiment. Compared to the first embodiment, the BBU 93 includes a quantization parameter determination unit (downlink) 56 that estimates IQ data power from radio band allocation information (downlink) and determines a quantization parameter for downlink signal processing. The quantization parameter obtained by the quantization parameter determination unit (downlink) 56 is input to the nonlinear quantization unit 55, and the quantization step size of the nonlinear quantization unit 55 is changed. The quantization parameter determined by the quantization parameter determination unit (downlink) 56 is transmitted to the RRU 91 side. In the OFDM cellular system, the average power of IQ data varies depending on the number of subcarriers used in downlink IQ data. Therefore, the number of used subcarriers is calculated from the radio band allocation information (downlink), and the quantization parameter is changed according to the number of used subcarriers.

  Compared with the first embodiment, the BBU 93 includes a quantization parameter determination unit (uplink) 57 that estimates the dispersion of IQ data from the radio band allocation information (uplink) and determines the quantization parameter for uplink signal processing. The quantization parameter determined by the quantization parameter determination unit (upstream) 57 is input to the quantization step restoration unit 51 to adjust the quantization step size. Further, the frame conversion unit 34 transmits the quantization parameter determined by the quantization parameter determination unit (upstream) 57 to the RRU 91 side. The amplitude value of the uplink IQ data is controlled to be substantially constant by the amplifier 13. However, in the OFDMA cellular system, if the number of subcarriers used for a radio signal is different, the PAPR (Peak to Average Power Ratio) of the radio signal is different. Further, even in an SC-FDMA (Single-Carrier Frequency-Multiple Access) cellular system, if the number of radio signals multiplexed differs according to the number of radio terminals, the PAPR of the radio signal differs. Depending on the PAPR, what is necessary is to set the dynamic range for nonlinear quantization. Therefore, PAPR of the radio signal is estimated and the quantization step size is changed according to the radio band allocation information (uplink).

  The communication system according to the present embodiment can operate in combination with the invention of the second embodiment.

  The present invention can be applied to the information communication industry.

11: Antenna 12: Transmission / reception switching unit 13: Amplification unit 14: Down-conversion unit 15: A / D conversion unit 16: Baseband filter unit (upstream)
17: Frame conversion unit 18: Electro-optical (E / O) conversion unit 23: Amplifier 24: Up-conversion unit 25: D / A conversion unit 26: Baseband filter unit (downlink)
27: Frame converter 28: Optical-electrical (O / E) converter 31: O / E converter 32: Frame converter 33: Modulator / demodulator 34: Frame converter 35: E / O converter 41: Quantization step Restoration unit 42: signal restoration unit 43: prediction coefficient extraction unit 44: prediction coefficient determination and error prediction unit 441: prediction coefficient determination unit 442: error prediction unit 45: nonlinear quantization unit 46: quantization parameter information extraction unit (downlink)
47: Quantization parameter information extraction unit (upstream)
51: quantization step restoration unit 52: signal restoration unit 53: prediction coefficient extraction unit 54: prediction coefficient determination and error prediction unit 541: prediction coefficient determination unit 542: error prediction unit 55: nonlinear quantization unit 56: quantization parameter determination Department (down)
57: Quantization parameter determination unit (upstream)
91: RRU
93: BBU

Claims (7)

A communication system in which an antenna unit that transmits and receives radio signals and a signal processing unit that controls the antenna unit are connected using an optical transmission line,
The antenna unit and the signal processing unit are
A prediction coefficient determination unit that calculates a prediction coefficient using a plurality of sample points of transmission data;
An error prediction unit that generates an error signal from the difference between the value of the next sample point calculated using the prediction coefficient and the actual value of the next sample point of the transmission data;
A non-linear quantization unit that quantizes the error signal with a non-linear quantization step size;
An error signal transmission unit that transmits the error signal quantized by the nonlinear quantization unit to the optical transmission line;
A quantization step restoring unit that converts the quantization step size of the error signal transmitted by the error signal transmission unit into a uniform quantization step size;
A signal restoration unit that restores the error signal whose quantization step size has been restored by the quantization step restoration unit to a sample point of transmission data using the prediction coefficient calculated by the prediction coefficient determination unit;
Equipped with a,
The nonlinear quantization unit calculates a quantization error that occurs when the error signal is quantized and outputs the quantization error to the error prediction unit,
The error prediction unit predicts the value of the next sample point using the value obtained by adding the quantization error to the sample point of the transmission data and the prediction coefficient, and calculates an error signal;
Communication system that characterized the. The communication system according to claim 1 , wherein the nonlinear quantization unit changes a quantization step size according to an average amplitude of the radio signal. The nonlinear quantization unit estimates the characteristics of the radio signal from the radio band allocation information, and changes the quantization step size according to the characteristics of the estimated wireless signals, according to claim 1 Communications system. The communication system according to claim 3 , wherein the characteristic of the radio signal is the number of subcarriers used for the radio signal. The communication system according to claim 3 , wherein the characteristic of the wireless signal is the number of wireless terminals that are communicating by transmitting and receiving the wireless signal. An optical signal transmission method of a communication system in which an antenna unit that transmits and receives a radio signal and a signal processing unit that controls the antenna unit are connected using an optical transmission line,
In the antenna unit and the signal processing unit,
A prediction coefficient determination procedure for calculating a prediction coefficient using a plurality of sample points of transmission data;
A quantization error determination procedure for calculating a quantization error generated when the error signal is quantized;
The quantization error using the value and the prediction coefficients by adding the sample points of the transmitted data, an error prediction step of generating an error signal to predict the value of the next sample point,
A non-linear quantization procedure for quantizing the error signal with a non-linear quantization step size;
An error signal transmission procedure for transmitting the error signal quantized by the nonlinear quantization procedure to the optical transmission line;
An optical signal transmission method comprising: An optical signal transmission method of a communication system in which an antenna unit that transmits and receives a radio signal and a signal processing unit that controls the antenna unit are connected using an optical transmission line,
In the antenna unit and the signal processing unit,
A quantization step restoration procedure for returning the quantization step size of the error signal transmitted through the transmission line to a uniform quantization step size;
A signal restoration procedure for restoring the error signal whose quantization step size is restored in the quantization step restoration procedure to a sample point of transmission data using a prediction coefficient received together with the error signal transmitted through the transmission path;
I have a,
The error signal is calculated by predicting the value of the next sample point using the value obtained by adding the quantization error generated when the error signal is quantized to the sample point of the transmission data and the prediction coefficient. That
An optical signal transmission method characterized by the above.

Images