JP2007129549A - Equalizer and equalizing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an equalizer and an equalizing method adaptively controlling the number of path samples at every detecting path to a propagation environment. <P>SOLUTION: The equalizer has a path-timing detector 101 detecting a path on the basis of a known signal from a receiving signal; and an adaptive path-sample number control unit 102 adaptively determining the number of the path samples as the number of a transmission-line estimate used for generating a transmission-line responding vector, on the basis of a detecting-path timing supplied from the path-timing detector 101 and a path-delay profile. The equalizer further has a weight generator 108 generating an equalizing weight on the basis of the number of the path samples determined by the adaptive path-sample number control section, and an equalizer 109 equalizing the receiving signal. The adaptive path-sample number control unit 102 adaptively determines the number of the path samples at every detecting path to the propagation environment. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an equalization apparatus and an equalization method applied to a receiver of a communication system that performs communication in a multipath environment.

  In a wireless communication system typified by mobile communication and wireless LAN (Local Area Network), a transmission signal (radio wave) from a transmission side is received by a transmission path that differs depending on reflection or scattering from a background object (such as a building). To reach. As an equalization method for achieving excellent equalization characteristics in communication in such a multipath environment, there is a method of performing equalization in the time domain or equalization in the frequency domain. In this equalization method, since equalization weights are generated based on the results of path detection and transmission path estimation, the accuracy of the path detection and transmission path estimation greatly affects the equalization characteristics. Therefore, in the conventional method of generating the transmission path response vector by arranging the transmission path estimation values corresponding to the detection path timings as they are as vector elements, the side lobe component of each path is reduced when the distance between the detection paths is small. Since they affect each other, the accuracy of the transmission path estimation value deteriorates. Also, the equalization characteristics deteriorate when there is a path timing error.

Therefore, in addition to the transmission path estimation value at the detection path timing, the transmission path response vector is generated by arranging the transmission path estimation values at a plurality of timings before and after the detection path timing as vector elements, and this occurs when the path interval is small. A multi-pass sampling method (see Non-Patent Document 1) that reduces the influence of side lobe components between paths and the influence of timing errors and improves equalization characteristics has been proposed. In this multipath sampling method, the number of path samples N _ps (N _ps is a natural number), which is the number of transmission path estimation values arranged for the detected path, is set in advance. The number of path samples N _ps is always the same for all detection paths.

FIG. 13 shows the configuration of a conventional equalizer, and FIG. 14 schematically shows the setting of the number of pass samples when the number of pass samples N _ps is set to 5 in advance. Hereinafter, a conventional equalization method using the multi-pass sampling method will be described with reference to FIGS.

  Referring to FIG. 13, the equalization apparatus 200 is applied to a receiver of a wireless communication system, and includes a path timing detection unit 201, a detection path timing transmission path estimation unit 105, and an adjacent path timing transmission path estimation unit 106. A transmission line response vector generation unit 202, a weight generation unit 108, an equalization unit 109, and a data demodulation unit 110.

The path timing detection unit 201 detects multipath timing using a known signal (pilot signal) from the received signal S _RSIG, and outputs the detected path timing as path detection information S _PDTI . Detection path timing transmission path estimation section 105 receives reception signal S _RSIG and path detection information S _PDTI , performs transmission path estimation for detection path timing, and outputs the estimation result as detection path timing transmission path estimation value S _CEDT. To do. The adjacent path timing transmission path estimation unit 106 receives the received signal S _RSIG and the path detection information S _PDTI , performs transmission path estimation at two timings before and after each of the detected path timings, and uses the estimation result as the adjacent path timing. Output as transmission path estimated value S _CENT .

The transmission path response vector generation unit 202 receives the detected path timing transmission path estimated value S _CEDT and the adjacent path timing transmission path estimated value S _CENT , and uses these transmission path estimated values S _CEDT and S _CENT to transmit a transmission path response vector. Generate and output S _CRV . Specifically, the transmission path response vector generation unit 202 performs transmission path estimation values h ₁ and h _{2 at} each detection path timing (solid line) and two timings before and after each detection path timing as shown in FIG. Transmission path estimated values h _1-2 , h _1-1 , h _{1 + 1} , h _{1 + 2} , h _2-2 , h _2-1 , h _{2 + 1} , h _{2 + 2 at} adjacent path timing (broken line) _Are arranged as vector elements to generate a transmission line response vector S _CRV . The output (transmission path response vector S _CRV ) of the transmission path response vector generation unit 202 is supplied to the weight generation section 108.

The weight generation unit 108 generates and outputs an equalization weight S _EQW using the transmission path response vector S _CRV supplied from the transmission path response vector generation unit 202. The equalization unit 109 receives the reception signal S _RSIG and the equalization weight S _EQW , performs equalization processing on the reception signal, and outputs the equalized reception signal S _EQSIG . The output of the equalization unit 109 (the equalized reception signal S _EQSIG ) is supplied to the data demodulation unit 110. As an equalization method in the equalization unit 109, there is chip equalization that performs filtering in the time domain.

The data demodulation unit 110 demodulates the equalized reception signal S _EQSIG supplied from the equalization unit 109 and outputs an information sequence S _RDAT .
IEICE General Conference 2005 B-5-120, "Examination of high-accuracy channel separation method in MMSE chip equalizer for HSDPA terminal"

  In the conventional equalization method using the multi-pass sampling method described above, the number of path samples in all detection paths is fixed to a preset number regardless of the propagation environment. For this reason, if the path interval is large enough to be hardly affected by sidelobe components, or if the path timing is accurate, adjacent path samples (transmission path estimated values of adjacent path timing) become an extra noise path. This causes a problem that the equalization characteristic is deteriorated.

  An object of the present invention is to provide an equalization apparatus and an equalization method capable of solving the above-described problem and adaptively controlling the number of path samples for each detection path according to a propagation environment.

To achieve the above object, according to a first aspect of the present invention, a reception signal obtained by combining a transmission signal from an external device via a multipath is input, and the multipath timing is determined using a known signal from the reception signal. A path timing detection unit that detects and outputs as path detection information;
A detection path timing transmission path estimation unit that receives the received signal and the path detection information as input, performs transmission path estimation at each of the detected path timings detected by the path timing detection unit, and outputs a detection path timing transmission path estimation value;
The received signal and the path detection information are input, and transmission path estimation is performed for each of a plurality of adjacent path timings adjacent to before and after the detected path timing detected by the path timing detection unit, and an adjacent path timing transmission path estimation value is obtained. An adjacent path timing transmission path estimator for output;
The path detection information is input, an adaptive path sample number control unit that determines the number of path samples, which is the number of transmission path estimation values for each detection path that are arranged in the generation of equalization weights, and outputs as the adaptive path sample number information;
A transmission path response vector generation unit that receives the detected path timing transmission path estimation value, the adjacent path timing transmission path estimation value, and the adaptive path sample number information as input, and generates and outputs a transmission path response vector in which the transmission path estimation values are arranged. When,
An equalization weight generation unit that takes the transmission path vector as input, generates the equalization weight, and outputs it as equalization weight information;
An equalization unit that receives the received signal and the equalization weight information as input, equalizes the received signal, and outputs the received signal after equalization;
The equalized received signal is input, and a data demodulating unit that estimates a transmission signal corresponding to the received signal,
The adaptive path sample number control unit adaptively controls the number of path samples for each detection path.

According to a second aspect of the present invention, a received signal obtained by synthesizing a transmission signal from an external device via a multipath is input, and the multipath timing is detected from the received signal using a known signal as path detection information. A path timing detection unit that outputs a path delay profile obtained from the received signal as profile information, and
A detection path timing transmission path estimation unit that receives the received signal and the path detection information as input, performs transmission path estimation at each of the detected path timings detected by the path timing detection unit, and outputs a detection path timing transmission path estimation value;
The received signal and the path detection information are input, and transmission path estimation is performed for each of a plurality of adjacent path timings adjacent to before and after the detected path timing detected by the path timing detection unit, and an adjacent path timing transmission path estimation value is obtained. An adjacent path timing transmission path estimator for output;
The number of adaptive path samples that are input as the path detection information and the profile information, determine the number of path samples that is the number of transmission path estimation values for each detected path that are arranged in the generation of equalization weights, and output as the number of adaptive path sample information A control unit;
A transmission path response vector generation unit that receives the detected path timing transmission path estimation value, the adjacent path timing transmission path estimation value, and the adaptive path sample number information as input, and generates and outputs a transmission path response vector in which the transmission path estimation values are arranged. When,
An equalization weight generation unit that takes the transmission path vector as input, generates the equalization weight, and outputs it as equalization weight information;
An equalization unit that receives the received signal and the equalization weight information as input, equalizes the received signal, and outputs the received signal after equalization;
The received signal after equalization is input, and is composed of a data demodulator that estimates a transmission signal corresponding to the received signal,
The adaptive path sample number control unit adaptively controls the number of path samples for each detection path.

  In the first and second inventions, the adaptive path sample number control unit determines the number of path samples for each detected path according to the propagation environment, and the transmission path response vector generation unit determines the determined path. A transmission path response vector is generated based on the number of samples. In this way, a transmission line response vector can be generated with an optimal number of path samples according to the propagation environment.

  As an index for determining the number of path samples, an index such as an interval between detected paths (path interval), a level of the detected path, and a level difference between the detected path and adjacent paths before and after the detected path is used.

  When the interval between detection paths (path interval) is used as an index, the number of path samples is controlled according to the propagation environment by setting the number of path samples in inverse proportion to the path interval. When the detected path level is used as an index, the number of path samples is controlled according to the propagation environment by setting the number of path samples in proportion to the detected path level. When the level difference between the detected path and the adjacent paths before and after it is used as an index, the propagation path estimation value is not arranged as a vector element at the adjacent path timing where the level difference from the detected path is greater than or equal to the threshold value. The number of pass samples is controlled according to.

  According to the present invention, since the number of path samples corresponding to the propagation environment is determined for each detection path, when the path interval is large enough to be hardly affected by sidelobe components, or when the path timing is accurate, It is possible to reduce the influence on the equalization characteristic of the adjacent path sample that has become an extra noise path. Furthermore, the equalization characteristic can be improved by setting an appropriate number of path samples for the propagation environment.

  Next, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration of an equalization apparatus according to an embodiment of the present invention. Referring to FIG. 1, the equalization apparatus according to the present embodiment is applied to a receiver of a wireless communication system, and includes a path timing detection unit 101, an adaptive path sample number control unit 102, and a detected path timing transmission path estimation. Unit 105, adjacent path timing transmission path estimation section 106, transmission path response vector generation section 107, weight generation section 108, equalization section 109, and data demodulation section 110.

The path timing detection unit 101 receives the received signal S _RSIG and identifies a multipath based on a known signal (specifically, a pilot signal included in the received signal) from the input received signal S _RSIG. A delay profile is calculated and path timing is detected for each path. The path timing detection unit 101 outputs the calculated path delay profile as profile information S _PDPI and outputs the detected timing (detected path timing) of each multipath as path detection information S _PDTI .

The adaptive path sample number control unit 102 receives the output (path detection information S _PDTI and profile information S _PDPI ) of the path timing detection unit 101, and the output (number of path samples) of the control information calculation unit 103. And a path sample number determination unit 104 that receives the control information S _PSMI ). The control information calculation unit 103 calculates an index for determining the number of path samples based on the profile information S _PDPI and the path detection information S _PDTI supplied from the path timing detection unit 101, and the calculation result is used to control the number of path samples. Output as information S _PSMI . The path sample number determination unit 104 determines the number of path samples for each detection path based on the path sample number control information _SPSMI supplied from the control information calculation unit 103, and adapts the determined number of path samples for each detection path. Output as pass sample number information S _APSI .

Detection path timing transmission path estimation section 105 receives reception signal S _RSIG and path detection information S _PDTI , performs transmission path estimation at each detection path timing based on these inputs, and transmits the estimation result to detection path timing transmission. Output as the estimated path value S _CEDT . The adjacent path timing transmission path estimation unit 106 receives the received signal S _RSIG and the path detection information S _PDTI as inputs, and based on these inputs, transmissions at adjacent (N _psm −1) / 2 timings adjacent to each detected path timing. The path estimation is performed, and the estimation result is output as the adjacent path timing transmission path estimation value S _CENT . Here, N _psm (N _psm is a natural number) is the maximum value of the number of pass samples in the multi-pass sampling method.

The transmission path response vector generation unit 107 receives the adaptive path sample number information S _APSI , the detected path timing transmission path estimation value S _CEDT and the adjacent path timing transmission path estimation value S _CENT , respectively. A transmission path response vector S _{CRV in} which the estimated values are arranged as vector elements is generated and output. The weight generation unit 108 receives the transmission path response vector S _CRV as input, and generates and outputs an equalization weight S _EQW based on this input. The equalization unit 109 receives the reception signal S _RSIG and the equalization weight S _EQW , performs equalization of the reception signal based on these inputs, and outputs the equalized reception signal S _EQSIG . Data demodulator 110 generates a data sequence S _RDAT corresponding to the transmitted data by demodulating the equalized received signal S _EQSIG.

  In the equalizer of the present embodiment configured as described above, the adaptive path sample number control unit 102 determines the number of path samples according to the propagation environment for each detected path, and the transmission path response vector generation unit 107 Since the transmission path response vector is generated based on the determined number of path samples, an optimal equalization weight can be adaptively generated according to the propagation environment.

  The configuration of the adaptive path sample number control unit 102 will be specifically described below.

  FIG. 2 shows the configuration of the control information calculation unit 103. Referring to FIG. 2, the control information calculation unit 103 includes a path interval calculation unit 111, a detected path level calculation unit 112, an adjacent path level calculation unit 113, and a path level difference calculation unit 114.

The path interval calculation unit 111 receives the path detection information S _PDTI , calculates an interval between the detected paths (path interval), and outputs the calculation result as the path interval information S _PDDI . The detected path level calculation unit 112 receives the path detection information S _PDTI and the profile information S _PDPI , calculates the path level for each path (detected path) for which the path timing is detected by the path timing detection unit 101, and A level difference (detection path level difference) between the detection path having the maximum detection level and the other detection paths is calculated, and the calculation result is output as detection path level information S _PLDT . The adjacent path level calculation unit 113 receives the path detection information S _PDTI and the profile information S _PDPI as input, calculates the level of each adjacent path before and after the detection path timing (N _psm −1) / 2 timing, and calculates the level. The result is output as adjacent path level information S _PLNT . The path level difference calculation unit 114 receives the path detection information S _PDTI , the detected path level information S _PLDT, and the adjacent path level information S _PLNT as inputs, and based on these inputs, the level difference between the detected path and the adjacent paths before and after the path. And the calculation result is output as adjacent path level difference information S _PLDI .

In the control information calculation unit 103 configured as described above, the path interval information S _PDDI , the detected path level information S _PLDT , the adjacent path level information S _PLNT , and the adjacent path level difference are used as indices for determining the number of path samples. Information S _PLDI is output. The levels of the detection path and the adjacent path are given by, for example, the amplitude of the signal waveform, power, SIR (Signal to Interference Power Ratio), and the like.

FIG. 3 is a schematic diagram for explaining the path interval. The unit of the path interval is the minimum accuracy (minimum timing) of path detection. In the example shown in FIG. 3, the path interval τ between the first path (amplitude X1) and the second path (amplitude X2) is 4. The path interval calculation unit 111 calculates this path interval and outputs it as path interval information S _PDDI .

FIG. 4 is a schematic diagram for explaining the detected path level difference. As shown in FIG. 4, the detection path level calculation unit 112 calculates a level difference Lv between the first path (amplitude X1) and the second path (amplitude X2), which is the detection path of the maximum level, and detects detection path level information. S Output as _PLDT . The level difference Lv (power difference) between the second path and the first path is:

It becomes. Here, when the SIR is set to the detection path level, the SIR value of each detection path is used.

FIG. 5 is a schematic diagram for explaining the adjacent path level difference. As shown in FIG. 5, when the amplitude of the detection path is y and the amplitudes of the adjacent paths before and after the detection path are z ₋ and z ₊ , respectively, the level difference (power difference) between the detection path and the adjacent paths before and after the detection path is respectively ,

It becomes. The path level difference calculation unit 114 calculates these level differences LvD ₋ and LvD ₊ and outputs them as adjacent path level difference information S _PLDI .

In the adaptive path sample number control unit 102, the path sample number determination unit 104 includes path interval information S _PDDI , detected path level information S _PLDT, and adjacent path level difference information S _PLDI supplied as an index from the control information calculation unit 103. The number of path samples for each detection path is determined using at least one piece of information.

Hereinafter, when the path interval information S _PDDI is used as an index (Example 1), the detected path level information S _PLDT is used as an index (Example 2), and the adjacent path level difference information S _PLDI is used as an index (Embodiment 3) The operation and effect of the case (Example 4) in which three pieces of information of path interval information S _PDDI , detected path level information S _PLDT and adjacent path level difference information S _PLDI are used as indices will be described. This will be specifically described.

In this embodiment, path interval information S _PDDI (path interval) is used as an index for determining the number of path samples. The path sample number determination unit 104 increases the number of path samples when the path interval is small in order to reduce the influence of the side lobe component between the paths and the timing error that occur when the path interval is small. If is large, the number of pass samples is reduced. In order to realize the control of the number of pass samples, the pass sample number determination unit 104 is given a relationship between the pass interval threshold value and the number of pass samples in advance.

FIG. 6 shows an example of the relationship between the path interval threshold and the number of path samples. In this example, the maximum value of the number of path samples is Npsm = 5, and four path interval thresholds ThD ₁ = 0, ThD ₂ = 6, ThD ₃ = 11, ThD ₄ = ∞, and the path interval τ (τ is a positive real number) ) And NpsD ₁ = 5, NpsD ₂ = 3, and NpsD ₃ = 1, respectively, and the path sample number determination unit 104 When the interval τ is ThD _i ≦ τ <ThD _{i + 1} (i = 1, 2, 3), the number of path samples is set to NpsD _i . In this case, when the path interval τ is 0 ≦ τ <6, the number of path samples is 5, and when the path interval τ is 6 ≦ τ <11, the number of path samples is 3, and the path interval τ is 11 ≦ When τ, the number of pass samples is 1.

FIG. 7A schematically shows path samples used for generation of a transmission path response vector in a conventional equalization apparatus having a fixed number of path samples, and FIG. FIG. 6 schematically shows path samples used for generation of transmission path response vectors. In each of the examples shown in FIGS. 7A and 7B, four paths (solid lines) are detected, and the path interval τ ₁ between the first path and the second path is 5, The path interval τ ₂ of the third path is 9, and the path interval τ ₃ of the third path and the fourth path is 13.

  In the conventional equalization apparatus, the number of path samples is fixed to a preset maximum value Npsm = 5 of the number of path samples. Therefore, as shown in FIG. 7A, when four paths (solid lines) are detected by the path timing detection unit 101, each of the first to fourth paths is independent of the path interval. The number of pass samples is the maximum value Npsm = 5 of the number of pass samples.

On the other hand, according to the present embodiment, if the number of path samples from the first path to the fourth path is Nps _j (j = 1, 2, 3, 4), the number of path samples Nps _{1 in} the first path and The number of path samples Nps ₂ in the second path is based on the path interval between the first path and the second path, and the number of path samples Nps ₃ in the third path is based on the path distance between the second path and the third path. The number of path samples Nps ₄ in the fourth path is determined according to the relationship shown in FIG. 6 based on the path interval between the third path and the fourth path. Therefore, in the present embodiment, as shown in FIG. 7B, the path interval τ ₁ between the first path and the second path is 5, so that Nps ₁ and Nps ₂ are both 5. Further, since the path interval τ ₂ between the second path and the third path is 9, Nps ₃ is set to 3. Further, since the path interval τ ₃ between the third path and the fourth path is 13, Nps ₄ is set to 1. Thus, by controlling the number of path samples in each detection path according to the path interval, the number of path samples according to the propagation environment can be determined.

  The number of path samples may be determined for each path detection period, or may be determined once every N periods (N is an arbitrary natural number) of path detection periods.

  Further, the pass sample number determination process by the pass sample number determination unit 104 can be basically divided into the following three processes.

(First pass sample number determination process)
When the maximum number of paths that can be detected by the path timing detector is P (P is a natural number), the p-th (p = 1, 2,..., P−1) detection path and the p + 1-th detection path. (the tau _p positive real number) paths interval tau _p is, Nd + 1 or a preset (Nd is a natural number) is a real number path interval threshold ThD _i (ThD _i of _{the, ThD 1 <ThD 2 <...} <ThD Nd + ₁ , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1, 2,..., Nd) In this case, the number of path samples in the p-th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ), and the number of path samples in the P-th detection path is set to pass. Set based on the interval τ _P-1 .

(Second pass sample number determination process)
When the maximum number of paths that can be detected by the path timing detector is P (P is a natural number), the p-th (p = 1, 2,..., P−1) detection path and the p + 1-th detection path. (the tau _p positive real number) paths interval tau _p is, Nd + 1 or a preset (Nd is a natural number) is a real number path interval threshold ThD _i (ThD _i of _{the, ThD 1 <ThD 2 <...} <ThD Nd + ₁ , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1, 2,..., Nd) In this case, the number of path samples in the (p + 1) th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ), and the number of path samples in the first detection path is set to pass. Set based on the interval τ ₁ .

(Third pass sample number determination process)
When the maximum number of paths that can be detected by the path timing detector is P (P is a natural number), the p-th (p = 1, 2,..., P−1) detection path and the p + 1-th detection path. Path interval τ _p (τ _p is a positive real number) and the path interval τ _{p + 1} between the (p + 1) th detection path and the (p + 2) th detection path, the smaller path interval τ ′ _p is , Preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), when ThD _j ≦ τ ′ _p <ThD _{j + 1} (j = 1, 2,..., Nd), the number of path samples in the p + 1th detection path is NpsD _j (NpsD _j is a natural _{number, NpsD 1> NpsD 2> ...} > NpsD Nd) is set to, path in the first detection path Set based the number of samples in the path interval tau _1, and set based the number of passes samples in the P-th path detected in the path interval tau _P-1.

In this embodiment, since the index is only the path interval information S _PDDI (path interval), in the configuration of the control information calculation 103 shown in FIG. 2, the detected path level calculation unit 111, the adjacent path level calculation unit 113, and The path level difference calculation unit 114 is not necessary.

In this embodiment, detection path level information S _PLDT (level difference between each detection path and the maximum level detection path) is used as an index for determining the number of path samples. The path sample number determination unit 104 increases the number of path samples when the level difference is small and the level difference from the maximum level detection path is large in order to prevent noise increase due to the placement of multiple paths when the path level is small. To reduce the number of pass samples. In order to realize this pass sample number control, a relationship between the level difference and the pass sample number is set in the pass sample number determining unit 104.

FIG. 8 shows an example of the relationship between the level difference and the number of pass samples. In this example, the maximum value of the number of path samples is Npsm = 5, and the level difference Lv (Lv is determined by _four path level thresholds ThL ₁ = 0, ThL ₂ = 1.0, ThL ₃ = 3.0, ThL ₄ = ∞. Is a positive real number), and the number of path samples corresponding to the specified level difference Lv range is NpsL ₁ = 5, NpsL ₂ = 3, and NpsL ₃ = 1, respectively. 104 sets the number of path samples to NpsL _i when the level difference Lv is ThL _i ≦ Lv <ThL _{i + 1} (i = 1, 2, 3). In this case, when the level difference Lv is 0 ≦ Lv <1.0, the number of pass samples is 5, and when the level difference Lv is 1.0 ≦ Lv <3.0, the number of pass samples is 3. When the level difference Lv is 3.0 ≦ τ, the number of pass samples is 1.

FIG. 9 (a) schematically shows path samples used to generate a transmission path response vector in a conventional equalizer with a fixed number of path samples, and FIG. 9 (b) shows the present embodiment. FIG. 6 schematically shows path samples used for generation of transmission path response vectors. In each of the examples shown in FIGS. 9A and 9B, four paths (solid lines) are detected, the first path is the maximum level detection path, and the first and second paths are detected. Level difference Lv ₂ is 2.1 (dB), level difference Lv ₃ between the first pass and the third pass is 1.2 (dB), and level difference Lv ₄ between the first pass and the fourth pass is 4.3 (dB). dB).

  In the conventional equalization apparatus, the number of path samples is fixed to a preset maximum value Npsm = 5 of the number of path samples. For this reason, as shown in FIG. 9A, when four paths (solid lines) are detected by the path timing detection unit 101, the first path is independent of the level difference from the maximum level detection path. The number of pass samples from the first pass to the fourth pass is the maximum number of pass samples Npsm = 5.

On the other hand, according to the present embodiment, if the number of path samples from the first path to the fourth path is Nps _j (j = 1, 2, 3, 4), the number of path samples Nps _{1 in} the first path is The preset maximum number of pass samples is Npsm (= 5). Further, the number of path samples Nps ₂ in the second path is based on the level difference Lv ₂ between the first path and the second path, and the number of path samples Nps ₃ in the third path is the level difference between the first path and the third path. Based on Lv ₃ , the number of path samples Nps ₄ in the fourth pass is determined according to the relationship shown in FIG. 8 based on the level difference Lv ₄ between the first pass and the fourth pass. Accordingly, in the case of the present embodiment, as shown in FIG. 9B, the level difference Lv ₂ is 2.1 (dB), so the number of path samples Nps ₂ is 3. Since the level difference Lv ₃ is 1.2 (dB), the number of pass samples Nps ₃ is 3. Since the level difference Lv ₄ is 4.5 (dB), the number of pass samples Nps ₄ is 1. As described above, by controlling the number of path samples in each detection path according to the level difference Lv, it is possible to determine the number of path samples according to the propagation environment.

The above-described path sample number determination processing according to the present embodiment is basically p-th (p = 1, 2,..., P) where P is the maximum number of paths that can be detected by the path timing detection unit. The level difference Lv _p (Lv _p is a real number) between the detection path level and the maximum level detection path is a preset Nl + 1 (Nl is a natural number) path level threshold ThL _k (ThL _k is a real number, ThL ₁ < For ThL ₂ <... <ThL _{Nl + 1} , ThL ₁ = 0, ThL _{Nl + 1} = ∞, k = 1, 2,..., Nl + 1), ThL _l ≦ Lv _p <ThL _{l + 1} (l = 1, 2,..., Nl), the processing is to set the number of path samples in the p-th detection path to NpsL _l (NpsL _l is a natural number, NpsL ₁ <NpsL ₂ <... <NpsL _Nl ).

  The number of path samples may be determined for each path detection period, or may be determined once every N periods (N is an arbitrary natural number) of path detection periods.

Further, in this embodiment, since the index is only the detected path level information S _PLDT (detected path level difference), the path interval calculation unit 111, the adjacent path level calculation unit in the configuration of the control information calculation 103 shown in FIG. 113 and the path level difference calculation unit 114 are not necessary.

In the present embodiment, adjacent path level difference information S _PLDI (adjacent path level difference) is used as an index for determining the number of path samples. The path sample number determination unit 104 considers the level difference between adjacent adjacent paths when there is a path timing error, so that the number of path samples is not arranged for adjacent path timings having a large adjacent level difference. Control. In order to realize this path sample number control, a threshold value α [dB] of the adjacent path level difference is set in the path sample number determining unit 104, and transmission is performed for the adjacent path timing at which the adjacent path level difference is equal to or greater than this threshold value. The number of path samples is controlled so that the path estimation value is not used as a vector element.

FIG. 10A schematically shows path samples used for generation of a transmission path response vector in a conventional equalization apparatus having a fixed number of path samples, and FIG. FIG. 6 schematically shows path samples used for generation of transmission path response vectors. In each of the examples shown in FIGS. 10A and 10B, four paths (solid lines) are detected. The level differences LvD _{1, -1} and LvD _{1, + 1} between the first path and the adjacent paths before and after it are 1.4 (dB) and 1.2 (dB), respectively. The level differences LvD _{2, -1} and LvD _{2, + 1} between the second path and the adjacent paths before and after the second path are 1.3 (dB) and 2.6 (dB), respectively. The level differences LvD _{3, -1} and LvD _{3, + 1} between the third path and the adjacent paths before and after the third path are 2.1 (dB) and 1.4 (dB), respectively. The level differences LvD _{4, -1} and LvD _{4, + 1} between the fourth path and the adjacent paths before and after the fourth path are 1.8 (dB) and 2.3 (dB), respectively. Here, the threshold α of the adjacent path level difference is 2. The number of path samples from the first pass to the fourth pass is Nps _j (j = 1, 2, 3, 4).

In the conventional equalization apparatus, the number of path samples is fixed to a preset maximum value Npsm (= 3) of the number of path samples. Therefore, as shown in FIG. 10A, when four paths (solid lines) are detected by the path timing detection unit 101, the first path to the fourth path are independent of the adjacent path level difference. Each of the pass sample numbers Nps _{1 to} Nps ₄ is the maximum number of pass samples Npsm (= 5).

On the other hand, according to this embodiment, the level differences LvD _{1, -1} and LvD _{1, + 1} between the first path and the adjacent paths before and after the first path are 1.4 (dB) and 1.2 (dB), respectively. Since both are less than the threshold value α (= 2), the number Nps ₁ of pass samples in the first pass is 3. The level differences LvD _{2, -1} and LvD _{2, + 1} between the second path and the adjacent paths before and after it are 1.3 (dB) and 2.6 (dB), respectively, and the level difference LvD _{2, + 1} is Since the threshold value α (= 2) is exceeded, the number of path samples Nps _{2 in} the second path is 2. The level differences LvD _{3, -1} and LvD _{3, + 1} between the third path and the adjacent paths before and after it are 2.1 (dB) and 1.4 (dB), respectively, and the level difference LvD _{3, -1} is the threshold α is greater than (= 2), the path number of samples Nps ₃ of the third path is 2. The level differences LvD _{4, -1} and LvD _{4, + 1} between the fourth path and the adjacent paths before and after it are 1.8 (dB) and 2.3 (dB), respectively, and the level difference LvD _{4, + 1} is Since the threshold value α (= 2) is exceeded, the number of pass samples Nps _{4 in} the fourth pass is 2. Thus, by controlling the number of path samples in each detection path according to the adjacent path level difference LvD, the number of path samples according to the propagation environment can be determined.

  The number of path samples may be determined for each path detection period, or may be determined once every N periods (N is an arbitrary natural number) of path detection periods.

In this embodiment, since the index is only adjacent path level difference information S _PLDI (adjacent path level difference), the path interval calculation unit 111 is not necessary in the configuration of the control information calculation 103 shown in FIG.

In this embodiment, as an index for determining the number of path samples, path interval information S _PDDI (path interval), detected path level information S _PLDT (level difference between each detected path and the maximum level detected path), and adjacent path level difference Three pieces of information S _PLDI (adjacent path level difference) are used. The path sample number determination unit 104 determines the first number of path samples for each detected path by using the path interval information S _PDDI in the same procedure as in the first embodiment, and the same as in the second embodiment. The second path sample number is determined by using the detected path level information S _PLDT in the procedure of the above, and the third path sample number is determined by using the adjacent path level difference information S _PLDI in the same procedure as in the third embodiment. And the number of path samples having the smallest value among the first to third number of path samples is set as the number of path samples of the detection path. Thereby, it is possible to reduce the influence of the side lobe component between the paths and the influence of the timing error. In this embodiment, the adjacent path level threshold value α is 2 [dB], and the maximum value of the number of path samples is Npsm = 5.

FIG. 11A schematically shows path samples used for generation of a transmission path response vector in a conventional equalization apparatus having a fixed number of path samples, and FIG. FIG. 6 schematically shows path samples used for generation of transmission path response vectors. In each of the examples shown in FIGS. 10A and 10B, four paths (solid lines) are detected. The path interval τ ₁ between the first path and the second path is 5, and the path interval τ ₂ between the second path and the third path is 8. The second path is the detection path of the maximum level, the level difference Lv ₁ between the first path and the second path is 1.1 (dB), and the level difference Lv ₃ between the second path and the third path is 2.1 (dB). ). The level differences LvD _{1, -1} , LvD _{1, + 1} , LvD _{1, -2} , LvD _{1, + 2} between the first path and the adjacent paths before and after it are 1.8 (dB) and 1.2 (dB), respectively. ) 3.8 (dB), 1.9 (dB). The level differences LvD _{2, -1} , LvD _{2, + 1} , LvD _{2, -2} , LvD _{2, + 2} between the second path and the adjacent paths before and after it are 1.0 (dB) and 1.3 (dB), respectively. ), 1.8 (dB), and 3.2 (dB). The level differences LvD _{3, -1} , LvD _{3, + 1} , LvD _{3, -2} , LvD _{3, + 2} between the third path and the adjacent paths before and after it are 1.8 (dB) and 2.2 (dB), respectively. ) 4.0 (dB), 4.0 (dB). The number of path samples from the first pass to the fourth pass is Nps _j (j = 1, 2, 3, 4).

In the conventional equalization apparatus, the number of path samples is fixed to a preset maximum value Npsm (= 5) of the number of path samples. For this reason, as shown in FIG. 11A, when four paths (solid lines) are detected by the path timing detection unit 101, the information such as the path interval, the detected path level difference, and the adjacent path level difference is related. without the respective paths sample number Nps ₁ ～Nps ₄ is the maximum value of either the number of passes the samples from the first pass to the fourth pass Npsm (= 5).

On the other hand, according to the present embodiment, the path sample number determination unit 104 performs the first path sample number determination process using the path interval information S _PDDI and the detected path level information for each of the first to third paths. A second pass sample number determination process using S _PLDT and a third pass sample number determination process using adjacent path level difference information S _PLDI are sequentially performed. In the first pass sample number determination process, the range of the path interval τ (τ is a positive real number) is defined by four path interval thresholds ThD ₁ = 0, ThD ₂ = 6, ThD ₃ = 11, and ThD ₄ = ∞. The number of path samples corresponding to the defined range of path interval τ is NpsD ₁ = 5, NpsD ₂ = 3, and NpsD ₃ = 1, respectively, and the path interval τ is ThD _i ≦ τ <ThD _{i + 1} (i = if the 1, 2, 3), sets the number of paths samples NpsD _i. In the second pass sample number determination process, four path level threshold values ThL ₁ = 0, ThL ₂ = 1.0, ThL ₃ = 3.0, and ThL ₄ = ∞, so that the level difference Lv (Lv is a positive real number) The range is defined, and the number of path samples corresponding to the defined level difference Lv is NpsL ₁ = 5, NpsL ₂ = 3, NpsL ₃ = 1, and the level difference Lv is ThL _i ≦ Lv <ThL _{i +} if a ₁ (i = 1,2,3), sets the number of paths samples NpsL _i. In the third path sample number determination process, the path estimation value is not used as a vector element for the adjacent path timing whose adjacent path level difference is equal to or greater than a predetermined threshold α (= 2 [dB]). Control the number of samples. Among the first to third pass sample number candidate values obtained by the first to third pass sample number determination processes, the smallest candidate value is determined as the pass sample number.

  FIG. 12 shows the results of the first to third pass sample number determination processing for the first to third passes shown in FIG.

According to the result shown in FIG. 12, in the first pass, 5 is obtained as the pass sample number candidate value in the first pass sample number determining process, and 3 is obtained as the pass sample number candidate value in the second pass sample number determining process. And 4 is obtained as the pass sample number candidate value in the third pass sample number determination process. In this case, since the pass sample number candidate value 3 obtained by the second pass sample number determination process is the smallest, the pass sample number Nps ₁ of the first pass is determined to be this candidate value “3”.

In the second pass, 5 is obtained as the pass sample number candidate value in the first pass sample number determining process, 5 is obtained as the pass sample number candidate value in the second pass sample number determining process, and the third In the pass sample number determination process, 4 is obtained as the pass sample number candidate value. In this case, since the path sample number candidate value 4 obtained by the third path sample number determination process is the smallest, the second pass path sample number Nps ₂ is determined to be this candidate value “4”.

Also, in the third pass, 3 is obtained as the pass sample number candidate value in the first pass sample number determination process, 3 is obtained as the pass sample number candidate value in the second pass sample number determination process, In the pass sample number determination process, 2 is obtained as the pass sample number candidate value. In this case, since the path sample number candidate value 2 obtained by the third pass sample number determination process is the smallest, the pass sample number Nps ₃ of the third pass is determined to be this candidate value “2”.

  The number of path samples may be determined for each path detection period, or may be determined once every N periods (N is an arbitrary natural number) of path detection periods.

Further, in this embodiment, all three indexes of the path interval information S _PDDI , the detected path level information S _PLDT and the adjacent path level difference information S _PLDI are used, but two of these indexes may be used. .

  In addition, among the pass sample number candidate values obtained from two or three indexes, the smallest pass sample number candidate value is set as the pass sample number. On the contrary, the largest pass sample number candidate value is obtained. May be determined as the number of pass samples.

  In addition, priorities may be set for a plurality of indices, and indices to be used may be determined based on the priorities.

  The equalization method after determining the number of path samples in the present invention described above includes, for example, chip equalization that performs filtering in the time domain and frequency domain equalization that performs equalization processing in the frequency domain.

  The present invention can be applied not only to a wireless communication system but also to all communication systems that cause a problem of multipath interference. For example, since the same multipath interference problem occurs in a transmission line used in cable TV, the present invention can also be applied to a communication system using such a transmission line.

It is a block diagram which shows the structure of the equalization apparatus which is one Embodiment of this invention. It is a block diagram which shows the structure of the control information calculation part shown in FIG. It is a schematic diagram for demonstrating the detection of the path | pass space | interval by the control information calculation part shown in FIG. It is a schematic diagram for demonstrating the detection of the detection path level difference by the control information calculation part shown in FIG. It is a schematic diagram for demonstrating the detection of the adjacent path level difference by the control information calculation part shown in FIG. It is a figure which shows an example of the relationship between a path space | interval threshold value and the number of path samples. (A) is a schematic diagram for explaining the number of path samples used for generation of a transmission path response vector in a conventional equalizer with a fixed number of path samples, and (b) is a first embodiment of the present invention. It is a schematic diagram for demonstrating the number of path samples used for the production | generation of the transmission line response vector in an example. It is a figure which shows an example of the relationship between a level difference and the number of pass samples. (A) is a schematic diagram for explaining the number of path samples used for generation of a transmission path response vector in a conventional equalizer with a fixed number of path samples, and (b) is a second embodiment of the present invention. It is a schematic diagram for demonstrating the number of path samples used for the production | generation of the transmission line response vector in an example. (A) is a schematic diagram for explaining the number of path samples used for generation of a transmission path response vector in a conventional equalizer with a fixed number of path samples, and (b) is a third embodiment of the present invention. It is a schematic diagram for demonstrating the number of path samples used for the production | generation of the transmission line response vector in an example. (A) is a schematic diagram for explaining the number of path samples used for generation of a transmission path response vector in a conventional equalizer with a fixed number of path samples, and (b) is a fourth embodiment of the present invention. It is a schematic diagram for demonstrating the number of path samples used for the production | generation of the transmission line response vector in an example. It is a figure which shows the result of having performed the pass sample number determination process about the 1st thru | or 3rd path | pass shown in FIG. It is a block diagram which shows the structure of the conventional equalization apparatus. It is a schematic diagram for demonstrating the conventional multipass sampling method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Equalizer 101 Path timing detection part 102 Adaptive path sample number control part 103 Control information calculation part 104 Path sample number determination part 105 Detection path timing transmission path estimation part 106 Adjacent path timing transmission path estimation part 107 Transmission path response vector generation part 108 Weight generator 109 Equalizer 110 Data demodulator 111 Path interval calculator 112 Detected path level calculator 113 Adjacent path level calculator 114 Path level difference calculator

Claims (32)

A path timing detection unit that receives a reception signal obtained by synthesizing a transmission signal from an external device via a multipath, detects a timing of the multipath using a known signal from the reception signal, and outputs it as path detection information When,
A detection path timing transmission path estimation unit that receives the received signal and the path detection information as input, performs transmission path estimation at each of the detected path timings detected by the path timing detection unit, and outputs a detection path timing transmission path estimation value;
The received signal and the path detection information are input, and transmission path estimation is performed for each of a plurality of adjacent path timings adjacent to before and after the detected path timing detected by the path timing detection unit, and an adjacent path timing transmission path estimation value is obtained. An adjacent path timing transmission path estimator for output;
The path detection information is input, an adaptive path sample number control unit that determines the number of path samples, which is the number of transmission path estimation values for each detection path that are arranged in the generation of equalization weights, and outputs as the adaptive path sample number information;
A transmission path response vector generation unit that receives the detected path timing transmission path estimation value, the adjacent path timing transmission path estimation value, and the adaptive path sample number information as input, and generates and outputs a transmission path response vector in which the transmission path estimation values are arranged. When,
An equalization weight generation unit that takes the transmission path vector as input, generates the equalization weight, and outputs it as equalization weight information;
An equalization unit that receives the received signal and the equalization weight information as input, equalizes the received signal, and outputs the received signal after equalization;
The equalized received signal is input, and a data demodulating unit that estimates a transmission signal corresponding to the received signal,
The equalizing apparatus, wherein the adaptive path sample number control unit adaptively controls the number of path samples for each detection path. The adaptive path sample number control unit includes:
A control information calculation unit that takes the path detection information as input, calculates a path interval between the paths for two detection paths that are continuous in time, and outputs the path interval as control information for the number of path samples;
The equalization apparatus according to claim 1, further comprising: a path sample number determining unit that receives the path sample number control information as input, determines the number of path samples for each detected path, and outputs the path sample number information as the adaptive path sample number information.   3. The path sample number determination unit according to claim 2, wherein when the path interval is larger than a threshold value, the path sample number is decreased, and when the path interval is smaller than the threshold value, the path sample number is increased. Equalization device. The path sample number determination unit is p-th (p = 1, 2,..., P−1), where P (P is a natural number) is the maximum number of paths that can be detected by the path timing detection unit. The path interval τ _p (τ _p is a positive real number) between the detection path and the (p + 1) -th detection path is a preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1 , 2,..., Nd), the number of path samples in the p-th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ), and the Pth The equalization apparatus according to claim 2, wherein the number of path samples in the detection path is set based on the path interval τ _P−1 . The path sample number determination unit is p-th (p = 1, 2,..., P−1), where P (P is a natural number) is the maximum number of paths that can be detected by the path timing detection unit. The path interval τ _p (τ _p is a positive real number) between the detection path and the (p + 1) -th detection path is a preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1 , 2,..., Nd), the number of path samples in the (p + 1) th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ). The equalization apparatus according to claim 2, wherein the number of path samples in the detection path is set based on the path interval τ _1. . The path sample number determination unit is p-th (p = 1, 2,..., P−1), where P (P is a natural number) is the maximum number of paths that can be detected by the path timing detection unit. Of the path interval τ _p (τ _p is a positive real number) between the detection path and the (p + 1) th detection path and the path interval τ _{p + 1} between the (p + 1) th detection path and the (p + 2) th detection path, The smaller path interval τ ′ _p is a preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0 , ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), when ThD _j ≦ τ ′ _p <ThD _{j + 1} (j = 1, 2,..., Nd), th NpsD _j path number of samples in the detection path (the NpsD _j are natural _{numbers, NpsD 1> NpsD 2> ...} > NpsD Nd) set to , Set based on the number of passes samples in the first detection path to the path interval tau _1, and set based the number of passes samples in the P-th path detected in the path interval tau _P-1, etc. according to claim 2 Device. A reception signal obtained by combining a transmission signal from an external device via a multipath is input, and the multipath timing is detected using the known signal from the reception signal and output as path detection information. A path timing detection unit that outputs a path delay profile obtained from the signal as profile information;
A detection path timing transmission path estimation unit that receives the received signal and the path detection information as input, performs transmission path estimation at each of the detected path timings detected by the path timing detection unit, and outputs a detection path timing transmission path estimation value;
The received signal and the path detection information are input, and transmission path estimation is performed for each of a plurality of adjacent path timings adjacent to before and after the detected path timing detected by the path timing detection unit, and an adjacent path timing transmission path estimation value is obtained. An adjacent path timing transmission path estimator for output;
The number of adaptive path samples that are input as the path detection information and the profile information, determine the number of path samples that is the number of transmission path estimation values for each detected path that are arranged in the generation of equalization weights, and output as the number of adaptive path sample information A control unit;
A transmission path response vector generation unit that receives the detected path timing transmission path estimation value, the adjacent path timing transmission path estimation value, and the adaptive path sample number information as input, and generates and outputs a transmission path response vector in which the transmission path estimation values are arranged. When,
An equalization weight generation unit that takes the transmission path vector as input, generates the equalization weight, and outputs it as equalization weight information;
An equalization unit that receives the received signal and the equalization weight information as input, equalizes the received signal, and outputs the received signal after equalization;
The received signal after equalization is input, and is composed of a data demodulator that estimates a transmission signal corresponding to the received signal,
The equalizing apparatus, wherein the adaptive path sample number control unit adaptively controls the number of path samples for each detection path. The adaptive path sample number control unit includes:
Using the path detection information as an input, calculate one or more of a path interval between two consecutive detection paths in time, a level of the detection path, and a level difference between the detection path and a plurality of adjacent paths before and after the detection path. A control information calculation unit that outputs the path sample number control information;
The equalization apparatus according to claim 7, further comprising: a path sample number determination unit that receives the path sample number control information, determines a path sample number for each detected path, and outputs the path sample number information as the adaptive path sample number information. The control information calculation unit is configured to calculate a path interval between the two detection paths in which detection timings in the path timing detection unit are temporally continuous,
The equalization apparatus according to claim 8, wherein the path sample number determination unit determines the number of path samples based on the path interval calculated by the control information calculation unit.   10. The path sample number determination unit according to claim 9, wherein when the path interval is larger than a threshold value, the path sample number is decreased, and when the path interval is smaller than the threshold value, the path sample number is increased. Equalization device. The path sample number determination unit is p-th (p = 1, 2,..., P−1), where P (P is a natural number) is the maximum number of paths that can be detected by the path timing detection unit. The path interval τ _p (τ _p is a positive real number) between the detection path and the (p + 1) -th detection path is a preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1 , 2,..., Nd), the number of path samples in the p-th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ), and the Pth The equalization apparatus according to claim 9, wherein the number of path samples in the detection path is set based on the path interval τ _P−1 . The path sample number determination unit is p-th (p = 1, 2,..., P−1), where P (P is a natural number) is the maximum number of paths that can be detected by the path timing detection unit. The path interval τ _p (τ _p is a positive real number) between the detection path and the (p + 1) -th detection path is a preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1 , 2,..., Nd), the number of path samples in the (p + 1) th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ). The equalization apparatus according to claim 9, wherein the number of path samples in the detection path is set based on the path interval τ _1. . The path sample number determination unit is p-th (p = 1, 2,..., P−1), where P (P is a natural number) is the maximum number of paths that can be detected by the path timing detection unit. Of the path interval τ _p (τ _p is a positive real number) between the detection path and the (p + 1) th detection path and the path interval τ _{p + 1} between the (p + 1) th detection path and the (p + 2) th detection path, The smaller path interval τ ′ _p is a preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0 , ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), when ThD _j ≦ τ ′ _p <ThD _{j + 1} (j = 1, 2,..., Nd), th NpsD _j path number of samples in the detection path (the NpsD _j are natural _{numbers, NpsD 1> NpsD 2> ...} > NpsD Nd) set to , Set based on the number of passes samples in the first detection path to the path interval tau _1, and set based the number of passes samples in the P-th path detected in the path interval tau _P-1, etc. according to claim 9 Device. The control information calculation unit is configured to calculate a level difference between the detection path of the maximum level and other detection paths among the detection paths detected by the path timing detection unit,
When the level difference calculated by the control information calculation unit is greater than a threshold, the path sample number determination unit reduces the number of path samples in the other detection paths, and the level difference calculated by the control information calculation unit is less than the threshold. The equalization apparatus according to claim 8, wherein when the number is small, the number of path samples in the other detection path is increased. The path sample number determination unit detects the level and maximum level of the p-th (p = 1, 2,..., P) detection path, where P is the maximum number of paths that can be detected by the path timing detection unit. The level difference Lv _{p with} respect to the path (Lv _p is a real number) is the Nl + 1 (Nl is a natural number) path level threshold value ThL _k (ThL _k is a real number, ThL ₁ <ThL ₂ <... _{Nl + 1} , ThL ₁ = 0, ThL _{Nl + 1} = ∞, k = 1, 2,..., Nl + 1), ThL _l ≦ Lv _p <ThL _{l + 1} (l = 1, 2,..., Nl) 15. The equalization apparatus according to claim 14, wherein the number of path samples in the p-th detection path is set to NpsL ₁ (NpsL ₁ is a natural number, NpsL ₁ <NpsL ₂ <... <NpsL _Nl ). The control information calculation unit is configured to calculate a level difference between each of the plurality of adjacent paths adjacent to the detection path detected by the path timing detection unit before and after the detection timing of the detection path. And
The path sample number determining unit excludes the transmission path estimation value at the detection timing of the adjacent path having the level difference larger than the threshold from the transmission path estimation value used for generating the transmission path response vector. 9. Equalizer according to claim 8, wherein the number is determined. The control information calculation unit includes at least two path samples among an interval between the detection paths, a level difference between the detection path of the maximum level and another detection path, and a level difference between the detection path and the plurality of adjacent paths. Configured to calculate number control information,
The path sample number determination unit selects the smallest number of path samples from among the number of path samples calculated based on each of the at least two path sample number control information calculated by the control information calculation unit. The equalization apparatus described in 1. The control information calculation unit includes at least two path samples among an interval between the detection paths, a level difference between the detection path of the maximum level and another detection path, and a level difference between the detection path and the plurality of adjacent paths. Configured to calculate number control information,
9. The path sample number determination unit selects the largest number of pass samples among the number of path samples calculated based on each of the at least two path sample number control information calculated by the control information calculation unit. The equalization apparatus described in 1.   The equalization apparatus according to claim 1, wherein the process for equalizing the received signal in the equalization unit is an equalization process in a time domain or a frequency domain. A timing of the multipath is detected using a known signal from a received signal obtained by synthesizing a transmission signal from an external device via a multipath, and a plurality of adjacent timings before and after the detection path timing are detected. In an equalization method for generating an equalization weight for estimating the transmission signal from the received signal, using a transmission path response vector generated based on a transmission path estimation value in each of adjacent path timings,
An equalization method including a path sample number control step of adaptively controlling the number of path samples, which is the number of transmission path estimation values for each detected path arranged in the generation of the transmission path response vector.   The path sample number control step includes, among detection paths detected from the received signal, a path interval between detection paths, a detection path level, and a level difference between a detection path and a plurality of adjacent paths adjacent to the detection path. 21. The equalization method according to claim 20, which is a step of controlling the number of path samples for each detection path based on one or more pieces of information. The pass sample number control step includes:
A control information calculation step for calculating a path interval between the paths for two detection paths in which the detection path timing is continuous in time;
21. The equalization method according to claim 20, further comprising: a path sample number determination step that determines the number of path samples for each detected path based on the path interval calculated in the control information calculation step.   The step of determining the number of path samples includes a step of decreasing the number of path samples when the path interval is larger than a threshold value and increasing the number of path samples when the path interval is smaller than a threshold value. 22. The equalization method according to 22. In the step of determining the number of path samples, when the maximum number of paths that can be detected is P (P is a natural number), the pth (p = 1, 2,..., P−1) detection path and the p + 1 th A path interval τ _p (τ _p is a positive real number) with a detection path of Nd + 1 (Nd is a natural number) is a preset path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... < For ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1, 2,..., Nd ), The number of path samples in the p-th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ), and the path samples in the P-th detection path The equalization method according to claim 22, comprising the step of setting the number based on a path interval τ _P-1 . In the step of determining the number of path samples, when the maximum number of paths that can be detected is P (P is a natural number), the p-th detection path (p = 1, 2,..., P−1) and the p + 1-th detection path. A path interval τ _p (τ _p is a positive real number) with a detection path of Nd + 1 (Nd is a natural number) is a preset path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... < For ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), ThD _j ≦ τ _p <ThD _{j + 1} (j = 1, 2,..., Nd ), The number of path samples in the (p + 1) th detection path is set to NpsD _j (NpsD _j is a natural number, NpsD ₁ > NpsD ₂ >...> NpsD _Nd ), and the path samples in the first detection path are set. 23. The equalization method according to claim 22, comprising the step of setting the number based on the path interval [tau] ₁ . In the step of determining the number of path samples, when the maximum number of paths that can be detected is P (P is a natural number), the p-th detection path (p = 1, 2,..., P−1) and the p + 1-th detection path. Path interval τ _p (τ _p is a positive real number) and the path interval τ _{p + 1} between the (p + 1) th detection path and the (p + 2) th detection path. ' _p is a preset Nd + 1 (Nd is a natural number) path interval threshold ThD _i (ThD _i is a real number, ThD ₁ <ThD ₂ <... <ThD _{Nd + 1} , ThD ₁ = 0, ThD _{Nd + 1} = ∞, i = 1, 2,..., Nd + 1), when ThD _j ≦ τ ′ _p <ThD _{j + 1} (j = 1, 2,..., Nd), the path in the p + 1th detection path the number of samples NpsD _j (NpsD _j is a natural _{number, NpsD 1> NpsD 2> ...} > NpsD Nd) is set to, the first detection path Set based path number of samples in the path interval tau ₁ at, and including the step of setting, based the number of passes samples in the P-th path detected in the path interval tau _P-1, the equalization method of claim 22. The pass sample number control step includes:
A control information calculation step of calculating a level difference between the detection path of the maximum level and the other detection paths among the detection paths detected from the received signal;
When the level difference calculated in the control information calculation step is larger than a threshold, the number of path samples in the other detection path is small, and when the level difference calculated in the control information calculation unit is smaller than the threshold, 21. The equalization method according to claim 20, further comprising a step of determining the number of path samples that increases the number of path samples in the detection path. The path sample number determination step detects the level and maximum level of the p-th (p = 1, 2,..., P) detection path, where P is the maximum number of paths that can be detected in the first step. The level difference Lv _{p with} respect to the path (Lv _p is a real number) is the Nl + 1 (Nl is a natural number) path level threshold value ThL _k (ThL _k is a real number, ThL ₁ <ThL ₂ <... _{Nl + 1} , ThL ₁ = 0, ThL _{Nl + 1} = ∞, k = 1, 2,..., Nl + 1), ThL _l ≦ Lv _p <ThL _{l + 1} (l = 1, 2,..., Nl) And the number of path samples in the p-th detection path is set to NpsL ₁ (NpsL ₁ is a natural number and NpsL ₁ <NpsL ₂ <... <NpsL _Nl ). Method. The pass sample number control step includes:
A control information calculation step for calculating a level difference between each of the plurality of adjacent paths adjacent to the detection path detected from the received signal before and after the detection path timing of the detection path;
The path sample number determining step excludes the transmission path estimation value at the detection path timing of the adjacent path whose level difference is greater than a threshold from the transmission path estimation value used for generating the transmission path response vector. 21. The equalization method according to claim 20, further comprising a pass sample number determination step for determining a number. The pass sample number control step includes:
Control information calculation step of calculating at least two path sample number control information among the detection path interval, the level difference between the detection path of the maximum level and other detection paths, and the detection path and the plurality of adjacent paths When,
The path sample number determining step determines the number of path samples for selecting the smallest number of path samples from among the number of path samples calculated based on each of the at least two path sample number control information calculated by the control information calculation unit. The equalization method according to claim 20, comprising: steps. The pass sample number control step includes:
Control information calculation step of calculating at least two path sample number control information among the detection path interval, the level difference between the detection path of the maximum level and other detection paths, and the detection path and the plurality of adjacent paths When,
The path sample number determination step determines the number of path samples for selecting the largest number of path samples among the number of path samples calculated based on each of the at least two path sample number control information calculated by the control information calculation unit. The equalization method according to claim 20, comprising: steps. 32. The equalization method according to claim 20, wherein an equalization process in a time domain or a frequency domain is used as the equalization process for estimating the transmission signal.

Images