JP2020068504A - Adaptive equalizer and carrier recovery circuit - Google Patents

Abstract

To provide an adaptive equalizer capable of realizing stable high demodulation performance even at a high multi-value.SOLUTION: An adaptive equalizer includes an adaptive equalization unit 31 that compensates a frequency characteristic of an input signal, and a tap update unit 32 that updates a tap coefficient for the adaptive equalization unit 31 by an algorithm based on a difference between a plurality of ideal points arranged on a complex plane and an output signal of the adaptive equalization unit 31, and the tap update unit 32 does not update the tap coefficient when the input signal or the output signal of the adaptive equalization unit 31 is located outside a determination area provided within the outer edges of the plurality of ideal points on the complex plane.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an adaptive equalizer used for reproducing a carrier wave and a received wave in digital wireless transmission, and a carrier wave recovery circuit including the adaptive equalizer.

  In recent years, wireless traffic has been increasing more and more, and from the viewpoint of high efficiency of frequency use, in digital wireless transmission, there is an increasing demand for high-speed transmission by a high multilevel QAM (Quadrature Amplitude Modulation) method. . In this high multi-level QAM system, demodulation performance may be deteriorated due to phase noise (phase error) of the carrier generated in the transmitter and the receiver. Therefore, there is known a carrier recovery circuit that improves demodulation performance (bit error rate) based on the degree of influence of phase noise and thermal noise (see, for example, Patent Document 1).

  In this carrier recovery circuit, the loop filter control unit controls the bandwidth of the loop filter based on the phase error detected by the phase error detector and the amplitude error detected by the amplitude error detector, whereby phase noise and thermal The bandwidth is set appropriately according to the noise, and the demodulation performance is improved.

JP, 2011-101177, A

  By the way, in the multi-level modulation, the phase error detection range of carrier wave / carrier reproduction becomes extremely narrow. That is, in the case of low multi-value, the phase error detection range is wide because the distance between adjacent ideal points is large, but in the case of high multi-value, the phase error detection range is small because the distance between adjacent ideal points is small. Becomes narrower. Since the phase error detection range is significantly narrowed, when the phase jitter (phase fluctuation) increases under the phase noise environment, the carrier recovery may be out of synchronization.

  As a result, when the tap coefficient is updated by the conventional decision-directed algorithm so that the output signal (or the received signal) approaches the ideal point with the closest value, an event occurs in which the signal point arrangement is unintentionally converged. I will end up. For example, as shown in FIG. 13, the output signal (black circles in the figure) is phase-rotated as shown in FIG. Then, it is assumed that the number of cases where the power of the ideal point is smaller than the power of the output signal outside the ideal point frame F1 is increased. In this case, as shown in FIG. 15, the tap coefficient is updated so that the output signal of the protruding portion P1 falls within this frame F1, and as a result, the output signal group converges small as shown in FIG. The average power of the signal at the digitizer output is reduced. Then, by repeating such updating of the tap coefficient, as shown in FIG. 17, for example, the four output signals converge so as to be concentrated around the ideal point (in the state where the signal point arrangement size is small, An event occurs that falls into the optimum state.

  On the other hand, in the carrier recovery circuit described in Patent Document 1, the specifications of the carrier recovery loop at the time of high multi-value are switched depending on whether thermal noise reduction or phase noise reduction is prioritized. Since the unstable operation due to the decrease of the phase error detection range due to the multi-valued conversion is not considered, it is difficult to realize the stable operation of the demodulator in the low C / N environment.

  Therefore, an object of the present invention is to provide an adaptive equalizer and a carrier recovery circuit including the adaptive equalizer, which can realize stable high demodulation performance even at high multivalues.

  In order to solve the above-mentioned problems, the invention according to claim 1 provides an adaptive equalizer for compensating the frequency characteristic of an input signal, a plurality of ideal points arranged on a complex plane, and an output of the adaptive equalizer. An algorithm based on an error from a signal, and a tap updating unit that updates the tap coefficient for the adaptive equalization unit, wherein the tap updating unit is the input signal or the output signal of the adaptive equalization unit is the complex The adaptive equalizer is characterized in that the tap coefficient is not updated when it is located outside a determination area provided within the outer edges of the plurality of ideal points on a plane.

  The invention according to claim 2 is the adaptive equalizer according to claim 1, further comprising: a power calculation unit that calculates an average power of an output signal of the adaptive equalization unit, and the tap update unit includes the input signal. Alternatively, when the output signal of the adaptive equalization unit is located near the outer edge of the determination area, it is determined whether to update the tap coefficient based on the average power calculated by the power calculation unit. And

  According to a third aspect of the present invention, the first phase rotator that rotates the phase of the input signal and the frequency characteristic of the phase rotation signal that is the input signal whose phase is rotated by the first phase rotator are compensated. 3. The adaptive equalizer according to claim 1, a phase error detector for detecting a phase error included in the phase rotation signal compensated by the adaptive equalizer, and a phase rotation control based on the phase error. A rotation signal generation unit that generates a signal, and a second phase rotator that rotates the phase of the input signal based on the phase rotation control signal, wherein the first phase rotator is the phase rotation control. It is a carrier recovery circuit characterized in that the phase of the input signal is rotated based on the signal.

  According to the invention described in claim 1, when the input signal or the output signal of the adaptive equalization unit is located outside the determination area smaller than the outer edges of the plurality of ideal points, the tap coefficient is not updated. . That is, if there is a possibility that the power of the input signal or the output signal of the adaptive equalizer is larger than the power of the ideal point group (the power of the ideal point group is smaller than the power of the input signal or the output signal), tap Since the coefficient is not updated, it is possible to prevent the average power of the output signal of the adaptive equalization unit from being reduced or attenuated (maintaining the average power). As a result, it is possible to obtain a stable output from the adaptive equalizer, and it is possible to realize highly accurate and stable demodulation performance and carrier recovery even in a high multi-value.

  According to the second aspect of the invention, for the input signal located near the outer edge of the determination area or the output signal of the adaptive equalization unit, the tap coefficient is updated based on the average power of the output signal. It is not done. Therefore, by not updating the tap coefficient near the outer edge of the determination area when the average power is small, the average power is increased, or when the average power is high, the tap coefficient near the outer edge of the determination area is updated. As a result, the average power can be further maintained by attenuating the average power.

  According to the third aspect of the invention, the phase rotation control signal is generated and the phase of the input signal is rotated based on the phase error of the phase rotation signal whose frequency characteristic is compensated by the adaptive equalizer. Even if there is waveform distortion due to, it is possible to accurately estimate the phase noise of the carrier wave and realize high demodulation performance and carrier wave reproduction. Moreover, as described above, since a stable adaptive equalizer output can be obtained, it is possible to realize more accurate and stable demodulation performance and carrier recovery. Furthermore, since the frequency characteristic of the phase rotation signal is compensated by the adaptive equalizer, the influence of thermal noise can be reduced.

It is a schematic block diagram which shows the carrier recovery circuit which concerns on Embodiment 1 of this invention. It is a schematic block diagram which shows the microwave radio system provided with the carrier recovery circuit of FIG. 2 is a schematic block diagram showing an adaptive equalizer of the carrier recovery circuit of FIG. 1. FIG. It is a figure which shows the relationship between the ideal point and the determination area in the adaptive equalizer of FIG. FIG. 4 is a diagram showing a state in which a received signal is phase-rotated in the adaptive equalizer of FIG. 3. It is a figure which shows the relationship between a received signal and a determination area in the phase rotation state shown in FIG. 6 is a diagram showing an output of the adaptive equalizer after the phase rotation shown in FIG. 5. FIG. It is a schematic block diagram which shows the adaptive equalizer which concerns on Embodiment 2 of this invention. It is a schematic block diagram which shows the adaptive equalizer which concerns on Embodiment 3 of this invention. In the present invention, it is a diagram showing an error situation (a) of a received signal when C / N is high and an error situation (b) of a received signal when C / N is low. 9A and 9B are explanatory diagrams showing a process of changing to a power attenuation state and a process of changing to a power increasing state in the adaptive equalizer of FIG. FIG. 10 is an explanatory diagram showing signal processing of a central portion in the adaptive equalizer of FIG. 9. It is a figure which shows the received signal in the conventional adaptive equalizer, and the relationship of an ideal point. It is a figure which shows the state which the received signal rotated the phase in the conventional adaptive equalizer. It is a figure explaining the update method of a tap coefficient in the phase rotation state shown in FIG. It is a conceptual diagram which shows the signal state following FIG. It is a conceptual diagram which shows the signal state following FIG.

  Hereinafter, the present invention will be described based on the illustrated embodiments.

(Embodiment 1)
1 to 7 show this embodiment, FIG. 3 is a schematic block diagram showing an adaptive equalizer 3 according to this embodiment, and FIG. 1 includes this adaptive equalizer 3. FIG. 3 is a schematic block diagram showing a carrier recovery circuit 1. The carrier wave reproducing circuit 1 is a circuit for reproducing a carrier wave in digital wireless transmission, and is provided in the receiving device 102 of the microwave wireless system shown in FIG. Here, the microwave wireless system will be briefly described first. The transmission signal mapped and modulated in the transmission device 101 is converted into an analog signal, multiplied by the carrier wave W1, and transmitted from the antenna. When it is received by the antenna of the receiving apparatus 102 through the multipath fading environment, it is multiplied by the carrier wave W2 to be digitally converted, demodulated by the carrier wave reproducing circuit 1 and demapped.

  The carrier wave regenerating circuit 1 mainly includes a first phase rotator 2, an adaptive equalizer 3, a phase error detector 4, an LPF 5, an NCO (rotation signal generator) 6, and a second phase rotator. 7 and an equalizer 8.

  The first phase rotator 2 is a rotator / multiplier that rotates the phase of the input signal, and rotates the phase of the input signal based on a phase rotation control signal of the NCO 6 described later. Specifically, phase rotation is performed on each of the I-channel baseband signal and the Q-channel baseband signal converted into digital signals, based on the sine wave and cosine wave of the phase rotation control signal of the NCO 6. Is.

  The adaptive equalizer 3 compensates the frequency characteristic of the phase rotation signal which is the input signal whose phase is rotated by the first phase rotator 2, that is, equalization for eliminating waveform distortion and data error of the phase rotation signal. It is a vessel. Here, the adaptive equalizer 3 is composed of a decision feedback equalizer (DFE: Decision Feedback Equalizer) or a linear equalizer, and updates the tap coefficient as described later.

  The phase error detector 4 is a detector that detects a phase error included in the phase rotation signal compensated by the adaptive equalizer 3. A specific detection method is a well-known technique. For example, a signal point corresponding to an output signal is selected from the signal point array of the modulation method used between the transmission / reception devices 101 and 102, and the coordinates of the selected signal point are selected. And the coordinates of the input signal point are compared to calculate the phase error value.

  The LPF 5 is a filter that removes a high frequency component of the phase error detected by the phase error detector 4 according to a predetermined bandwidth, and is composed of a low pass filter.

  The NCO 6 is a generation unit that generates a phase rotation control signal based on the phase error from which the high-frequency component has been removed by the LPF 5, and is composed of an NCO (Numerically Controlled Oscillator). Specifically, the sine wave and the cosine wave having opposite phases are generated based on the phase error from the LPF 5 and output to the first phase rotator 2 to control the phase rotation by the first phase rotator 2. To do. Further, the generated phase rotation control signal is output to the second phase rotator 7.

  The second phase rotator 7 is a rotator / multiplier that rotates the phase of the input signal. The second phase rotator 7 rotates the phase of the input signal based on the phase rotation control signal from the NCO 6 to equalize frequency characteristics. Output to the container 8. That is, the phase of the input signal is rotated based on the sine wave and the cosine wave from the NCO 6 based on the phase error detected by the adaptive equalizer 3 after frequency characteristic compensation (waveform distortion is eliminated). Thus, the adaptive equalizer 3 is mounted in the carrier recovery loop (the loop of the first phase rotator 2, the phase error detector 4, the LPF 5 and the NCO 6), and the frequency characteristic is compensated by this. The phase noise of the input signal is canceled based on the phase error value estimated later.

  Next, a method of updating the tap coefficient in the adaptive equalizer 3 will be described. As shown in FIG. 3, the adaptive equalizer 3 includes an adaptive equalizer (equalizer body, filter) 31 for compensating the frequency characteristic of the input signal, and a tap update for updating the tap coefficient for the adaptive equalizer 31. And a part 32.

  The tap updating unit 32 is an algorithm (decision-oriented algorithm) based on the error between the plurality of ideal points (reference signals) R arranged on the complex plane and the output signal of the adaptive equalization unit 31. This is an updating unit that updates the tap coefficient (updates the previous tap coefficient). That is, a plurality of ideal points R are arranged in a grid pattern at predetermined equal intervals, and an output signal and an ideal point R are obtained by using a decision-oriented algorithm based on a minimum mean square error (MMSE: Minimum Mean Square Error). The tap coefficient is calculated and updated so as to minimize the error power of (to bring the output signal closer to the ideal point R having the closest value). As a decision-oriented algorithm, an LMS (Least Mean Square) algorithm or an RLS algorithm is used. (Recursive Least Square) algorithm is adopted.

  The tap updating unit 32 updates the tap coefficient when the received signal (input signal) S is located inside the determination area C provided on the complex plane, and the received signal S is located outside the determination area C. If it is located, the tap coefficient is not updated. Here, the determination area C is an area provided within the outer edge (outer frame F1) of the plurality of ideal points R on the complex plane, and in this embodiment, as shown in FIG. 4, on the complex plane. A plurality of ideal points R are arranged in a grid pattern in a regular square area centered on the origin at predetermined equal intervals. Note that FIG. 4 shows an ideal state in which the received signal S overlaps the ideal point R.

  A circular judgment area C centered on the origin on the complex plane is provided at the outer edge of the ideal point R group, that is, in the area of the outer frame F1 of a square shape, and the diameter of the judgment area C is equal to the ideal point R group. The size is set to be approximately the same as one side of the regular square. That is, a large circular judgment area C is provided so as to enclose as many ideal points R as possible, whereby the constellation (signal point arrangement diagram on the complex plane, signal space diagram) is maintained and the amount of power is increased. And the tap coefficient can be updated for as many received signals S (output signals) as possible.

  Then, the update control unit 321 determines whether the received signal S is located inside or outside the determination area C, and when it is located inside, transmits “updatable” to the LMS updating unit 322. If it is located on the outer side, “update not possible” is transmitted to the LMS updating unit 322. Next, when the LMS update unit 322 receives “updatable”, the LMS update unit 322 calculates and updates the tap coefficient by the decision-oriented algorithm based on the tap update error between the ideal point R and the output signal of the adaptive equalization unit 31. And inputs it to the adaptive equalizer 31. On the other hand, when the LMS updating unit 322 receives “unupdateable”, the LMS updating unit 322 does not calculate or update the tap coefficient based on the received signal S (output signal).

  As a result, as shown in FIG. 5, when the phase of the received signal S group is rotated, the tap coefficient is updated for the received signal S located inside the judgment area C, and the outside of the judgment area C, that is, in FIG. The tap coefficient is not updated for the received signal S located in the protruding portion P2. Therefore, the received signal S of the protruding portion P2 is not tap-updated so as to be within the outer frame F1 of the ideal point R group, and as shown in FIG. 7, the average power of the signal at the equalizer output is There is no reduction.

  As described above, according to the adaptive equalizer 3, the tap coefficient is updated when the received signal (input signal) S is located outside the determination area C that is smaller than the outer edge of the ideal point R group. Absent. That is, when the power of the reception signal S is larger than the power of the ideal point R in the determination area C (the power of the ideal point R in the determination area C is smaller than the power of the reception signal S), the tap coefficient is updated. Therefore, it is possible to prevent the average power of the output signal of the adaptive equalization unit 31 from being reduced or attenuated (maintaining the average power). As a result, it is possible to obtain a stable output from the adaptive equalizer 3, and it is possible to realize highly accurate and stable demodulation performance and carrier recovery even in a high multi-value.

  Further, according to the carrier wave regenerating circuit 1, the phase rotation control signal is generated based on the phase error of the phase rotation signal whose frequency characteristic is compensated (waveform distortion is eliminated) by the adaptive equalizer 3 to generate the input signal. Since the phase is rotated, even if there is waveform distortion due to fading, it is possible to accurately estimate the carrier phase noise (detected by the phase error detector 4) and achieve high demodulation performance and carrier recovery. . Moreover, as described above, since a stable adaptive equalizer output can be obtained, it is possible to realize more accurate and stable demodulation performance and carrier recovery. Further, since the frequency characteristic of the phase rotation signal is compensated by the adaptive equalizer 3, the influence of thermal noise can be reduced.

(Embodiment 2)
FIG. 8 is a schematic block diagram showing the adaptive equalizer 3 according to this embodiment. In this embodiment, when the output signal of the adaptive equalization unit 31 is located inside the determination area C, the tap coefficient is updated, and when it is located outside the determination area C, the tap coefficient is updated. The configuration is different from that of the first embodiment in that there is no point, and the same configurations as those of the first embodiment are denoted by the same reference numerals and the description thereof will be omitted.

  That is, the update control unit 321 determines whether the output signal of the adaptive equalization unit 31 is located inside or outside the determination area C, and if it is located inside, the “updatable” is LMS updated. When it is located outside, the “update not possible” is transmitted to the LMS update unit 322. When the LMS updating unit 322 receives “updatable”, the LMS updating unit 322 calculates and updates the tap coefficient based on the tap updating error between the ideal point R and the output signal of the adaptive equalization unit 31, and the adaptation and the like. Input to the conversion unit 31. On the other hand, when the LMS update unit 322 receives “update disabled”, the LMS update unit 322 does not calculate or update the tap coefficient based on this output signal.

  As described above, according to this embodiment, when the output signal of the adaptive equalization unit 31 is located outside the determination area C smaller than the outer edge of the ideal point R group, the tap coefficient is not updated. Therefore, similarly to the first embodiment, it is possible to prevent the average power of the output signal of the adaptive equalization unit 31 from being reduced or attenuated. As a result, it is possible to obtain a stable output from the adaptive equalizer 3, and it is possible to realize highly accurate and stable demodulation performance and carrier recovery even in a high multi-value. Moreover, since it is determined whether or not to update the tap coefficient based on the position of the output signal of the adaptive equalization unit 31, that is, the power, the average power of the output signal of the adaptive equalization unit 31 is reduced or attenuated. It becomes possible to prevent more appropriately.

(Embodiment 3)
9 to 12 show this embodiment, and FIG. 9 is a schematic block diagram showing the adaptive equalizer 3 according to this embodiment. In this embodiment, when the received signal (input signal) or the output signal of the adaptive equalization unit 31 is located near the outer edge of the determination area C (outer side near the edge), the output signal of the adaptive equalization unit 31 The configuration is different from that of the first embodiment in that it is determined whether or not the tap coefficient is updated based on the average power, and the same components as those of the first embodiment are denoted by the same reference numerals, and the description thereof is omitted. Is omitted.

  In this embodiment, as shown in FIG. 10, the determination area C is the same as the outer edge (outer frame F1) of the ideal point R, and tap update is performed based on the position of the received signal (the circle in the figure is the ideal point R). The case of determining whether or not is described. First, the effect of thermal noise will be described. When C / N is high, as shown in FIG. 10A, the amplitude range E due to thermal noise is small and all received signals are within the outer frame F1 (determination area C). Positioned, the average of the error between the hard decision result and the received signal becomes zero, and the power is maintained. On the other hand, when C / N is low, as shown in FIG. 10B, the amplitude range E due to thermal noise is large and the received signal is located outside F2 of the outer frame F1, and the error between the hard decision result and the received signal is The average becomes 0 or more, and as a result, the adaptive equalization unit 31 updates the tap coefficient so as to attenuate the power.

  On the other hand, in this embodiment, first, a power calculation unit 323 that calculates the average power of the output signal of the adaptive equalization unit 31 is provided. Further, when the received signal is located near the outer edge of the outer frame F1, the update control unit 321 switches between the power attenuation state and the power increase state based on the average power calculated by the power calculation unit 323.

  Specifically, as shown in FIG. 11A, the state in which the tap coefficient is updated with respect to the received signal located near the outer edge (outer side) E2 of the outer frame F1 is the power attenuation state. In this state, compared with the received signal in the adjacent range E1 (small arrow), the received signal in the vicinity of the outer edge E2 (large arrow) is largely tap-updated in the reduction direction, so that the power is attenuated. On the other hand, as shown in FIG. 11B, the state in which the tap coefficient is not updated for the received signal located near the outer edge of the outer frame F1 is the power increase state. In this state, for the ideal point R in the vicinity of the inner edge of the outer frame F1 (inside the vicinity of the edge), only the received signal in the adjacent range E1 is tap-updated in the expansion direction, so that the power increases. Note that, in the central portion of the outer frame F1, as shown in FIG. 12, the update distances (arrows) of the received signals of the adjacent ranges E1 adjacent to each other with respect to the ideal point R are equal (the power fluctuations are balanced). Because it is thought that, tap update as usual.

  Then, when the average power calculated by the power calculation unit 323 exceeds the first threshold value, the update control unit 321 switches to the power attenuation state and performs the tap update of the reception signal located near the outer edge of the outer frame F1. Then, “updatable” is transmitted to the LMS updating unit 322. When the average power calculated by the power calculation unit 323 is smaller than the second threshold value (<first threshold value), the update control unit 321 switches to the power increase state and is located near the outer edge of the outer frame F1. “Unable to update” is transmitted to the LMS updating unit 322 so that the tap update of the received signal is not performed.

  According to this embodiment, for the input signal located near the outer edge of the determination area C (outer frame F1) or the output signal of the adaptive equalization unit 31, the tap coefficient is updated based on the average power of the output signal. May or may not take place. Therefore, when the average power is small, the tap coefficient is not updated in the vicinity of the outer edge of the determination area C to increase the average power, or when the average power is high, the update of the tap coefficient in the vicinity of the outer edge of the determination area C is performed. By performing the above, it becomes possible to further maintain the average power by attenuating the average power.

  Although the embodiments of the present invention have been described above, the specific configuration is not limited to the above-mentioned embodiments, and even if there is a design change or the like within a range not departing from the gist of the present invention, Included in the invention. For example, although the circular determination area C is provided in the above embodiment, the determination areas C having other shapes may be provided. For example, the shape may be substantially the same as the shape of the outer edge / outer frame F1 of the ideal point R group (regular quadrangle), or as shown in FIG. The point R group and the received signal S group may have a substantially regular octagonal shape.

1 Carrier wave reproduction circuit 2 1st phase rotator 3 Adaptive equalizer 31 Adaptive equalization part 32 Tap update part 321 Update control part 322 LMS update part 323 Power calculation part 4 Phase error detector 5 LPF
6 NCO (rotation signal generator)
7 Second phase rotator 8 Equalizer S Received signal (output signal)
R ideal point

Claims (3)

An adaptive equalization unit that compensates for the frequency characteristics of the input signal,
An algorithm based on the error between a plurality of ideal points arranged on a complex plane and the output signal of the adaptive equalization unit, a tap update unit that updates the tap coefficient for the adaptive equalization unit,
Where the input signal or the output signal of the adaptive equalization unit is located outside the determination area provided within the outer edges of the plurality of ideal points on the complex plane. , Do not update the tap coefficient,
An adaptive equalizer characterized by the above. A power calculator for calculating the average power of the output signal of the adaptive equalizer,
Whether the tap update unit updates the tap coefficient based on the average power calculated by the power calculation unit when the input signal or the output signal of the adaptive equalization unit is located near the outer edge of the determination area. Determine whether or not
The adaptive equalizer according to claim 1, wherein: A first phase rotator that rotates the phase of the input signal;
The adaptive equalizer according to claim 1 or 2, which compensates a frequency characteristic of a phase rotation signal which is an input signal whose phase is rotated by the first phase rotator.
A phase error detector for detecting a phase error included in the phase rotation signal compensated by the adaptive equalizer,
A rotation signal generation unit that generates a phase rotation control signal based on the phase error,
A second phase rotator that rotates the phase of the input signal based on the phase rotation control signal;
And the first phase rotator rotates the phase of the input signal based on the phase rotation control signal,
A carrier wave regenerating circuit characterized in that

Images