JP2001251278A - Interference canceler - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suitably separate and extract demodulation signals concerning each user with few number of stages. SOLUTION: A weight coefficient α is multiplied by the output of an adder 50 and outputted as a corrected interference residual estimated signal e'n,m about ICU14c-n-m (n>=2). Then the weight coefficient α is multiplied by the output of an adder 46 placed at a next-stage finger 42d, and a corrected interference replica h'n-1,m,k generated at the preceding stage is added to the result of multiplication and then outputted as a corrected interference replica h'n,m,k of the present stage.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an interference canceling apparatus, for example, a direct sequence code division multiple access (DS-CDMA).
The present invention relates to an interference canceling device that separates and extracts each user signal from a received spread signal while reducing mutual interference between users in a base station that employs an (ltiple Access) scheme.

[0002]

2. Description of the Related Art In the DS-CDMA system, a user signal transmitted from each mobile station is spread over a wide frequency band by a unique spreading code and transmitted to a transmission line. On the other hand, the base station performs a despreading process using the same spreading code as the mobile station side on the received spread signal on which a plurality of user signals are superimposed, and further performs a transmission channel estimation process.
Separately extract the original user signal. At this time, due to the cross-correlation between the spreading codes, the user signals actually separated and extracted include mutual interference. In order to reduce such mutual interference between users, various types of multi-stage interference canceling apparatuses have been conventionally proposed.

FIG. 8 is a diagram showing a functional configuration of a conventional multi-stage interference canceling apparatus. The interference cancellation apparatus 10 shown in FIG. 1 is provided in a DS-CDMA base station. The interference cancellation apparatus 10 is configured to include a first stage 12-1 to an N-th stage 12-N. The interference cancellation device 10 converts a received spread signal Rb (baseband signal) in which a plurality of user signals are code-division multiplexed into a signal between users. And outputs demodulated signals R _{1 to} R _M for the first to M-th users while reducing mutual interference. here,
N and M are each an integer of 1 or more.

The first stage 12-1 includes an ICU (Interference Canceling Uniform) corresponding to each of M users.
t; interference canceling unit) 14-1-1 to 14-1
-M is provided. Also, ICU14-1-1 to ICU14-1-1
For each of 4-1-M, delay elements 16-1-1 to 16-1-1
16-1-M and adders 18-1-1 to 18-1-M are provided.

[0005] Similarly, the second stage 12-2 to the (N-
1) The stage 12- (N-1) is provided with ICUs 14-n-1 to 14-n-M corresponding to each of M users (2≤n≤N-1). Also, ICU14-n
-1 to 14-n-M, the delay element 16-
n-1 to 16-n-M and adders 18-n-1 to 18-
n−M (2 ≦ n ≦ N−1).

[0006] Also, the final stage, the Nth stage 1
2-N also has an ICU 14 corresponding to each of the M users.
-N-1 to 14-NM are provided. Also, IC
For each of U14-N-1 to 14-N- (M-1), delay elements 16-N-1 to 16-N- (M-1)
And adders 18-N-1 to 18-N- (M-1). No delay element and adder corresponding to ICU14-NM are required.

The interference canceling device 10 is further provided with a level ranking circuit 20 for the first user to
The signal levels of the user signals related to the M-th user are ranked, and the ICU 14-nm is determined according to the ranking.
Is assigned to the user. Specifically, in each stage, ICUs 14-n-1 to ICU 14-n-1 to 14
-N-M are assigned.

When the interference cancellation apparatus 10 is realized by hardware, it is not necessary to actually prepare hardware for all stages. For example, only one stage may be implemented as hardware, and the hardware may be sequentially used as the next stage hardware.

In the first stage 12-1, the reception spread signal
Rb is input to ICU 14-1-1, where
Wataru replica h_{1,1, k}And the interference residual estimation signal e_1,1Is generated
(K is a subscript identifying the path
Is an integer of 1 or more). Interference replica h_{1,1, k}Is I
CU14-2-1, the interference residual estimation signal e_1,1
Is supplied to the adder 18-1-1. After that,
Prika h_{n, m, k}Is estimated by ICU14-nm
In addition, the m-th user and the k-th user included in the received spread signal Rb
This is a signal component (despread signal) related to the signal. On the other hand,
Difference estimation signal e_{n, m}Is input to the ICU14-nm
Spread signal (received spread signal Rb or error signal r _{n, m-1},
r_{n-1, M}) Contained in the m-th user (spreading)
Signal).

The adder 18-1-1 is connected to the ICU 14-1-1.
1 subtracts the interference residual estimation signal e _1,1 from the received spread signal Rb passed through the delay element 16-1-1, and outputs the result as an error signal r _1,1 . This error signal r _1,1
Is supplied to the ICU 14-1-2 and the delay element 16-1-2. The error signal r _{n, m} is a signal obtained by subtracting all the interference residual signals e _{n, m} already output from the received spread signal Rb.

The ICU 14-1-m has an error signal r
_{1, m-1} is input, and an interference replica h _{1, m, k} and an interference residual estimation signal e _{1, m} are generated there (2 ≦ m ≦ M). The interference replica h _{1, m, k} is the ICU 14-2-m provided in the next stage corresponding to the m-th user
Is supplied to The adder 18-1-m is connected to the ICU14
−1−m, the interference residual estimation signal e _{1, m} is subtracted from the error signal r _{1, m−1} passed through the delay element 16-1-m, and is output as the error signal r _{1, m.} I do. Error signal r
_{1, m} is ICU14-1- (m + 1) and delay element 16-1
− (M + 1) (2 ≦ m ≦ M−1).
Only the error signals r1 and M output from the CU 14-1-M are the I signals connected to the highest level of the second stage 12-2.
It is supplied to the CU 14-2-1 and the delay element 16-2-1.

In the second stage 12-2 to the (N-1) th stage 12- (N-1), the ICU 14-n-1 is an IC provided for the same user in the previous stage.
Interference replica h supplied from U14- (n-1) -1
_{n-1,1, k} and the IC connected to the end of the previous stage
The interference residual estimation signal e _{n-1, M} supplied from U14- (n-1) -M is input, where the interference replica h _{n, 1, k} and the interference residual estimation signal e _{n, M} are input _{. 1} is generated (2 ≦ n
≤N-1). The interference replica hn _{, 1, k} is provided in the next stage by the ICU 14-
(N + 1) -1 while the interference residual estimation signal e
_{n, 1} is supplied to the adder 18-n-1.

The adder 18-n-1 is connected to the ICU 14-n-
1 subtracts the interference residual estimation signal en _{, 1} output from the error signal rn _{-1, M} via the delay element 16-n-1, and outputs it as the error signal rn _{, 1} . This error signal r _{n, 1}
Is supplied to the ICU 14-n-2 and the delay element 16-n-2 (2 ≦ n ≦ N−1).

The ICU 14-nm receives the error signal r _{n, m-1}
And an interference replica h _{n-1, m, k} output from the ICU 14- (n-1) -m provided in association with the same user in the previous stage _, and are input there. h _{n, m, k} and an interference residual estimation signal en _{, m} are generated (2 ≦ n ≦ N−1; 2 ≦ m ≦ M). The interference replica hn _{, m, k} is supplied to the ICU 14- (n + 1) -m provided in association with the m-th user in the next stage, while the interference residual estimation signal en _{, m} is added to the adder 18
-N-m.

The adder 18-nm is connected to the ICU 14-n-
The interference residual estimation signal en _{, m} output from m is subtracted from the error signal rn _{, m-1} that has passed through the delay element 16- _{nm, and} is output as the error signal rn _{, m} . The error signal r _{n, m} is I
CU14-n- (m + 1) and the delay element 16-n- (m +
1) and (2 ≦ n ≦ N−1; 2 ≦ m ≦ M−
1), error signal r output from ICU 14-n-M
Only _{n and M} are supplied to the ICU 14- (n + 1) -1 and the delay element 16- (n + 1) -1 which are connected to the highest position of the (n + 1) th stage 12- (n + 1).

Finally, in the N-th stage 12-N, the ICU 14- (N-1) -1 provided for the same user in the previous stage is provided in the uppermost ICU 14-N-1.
And interference replica h _{N-1, 1} to be supplied from, and is connected to the end of the previous stage ICU14- (N-1) and the interference residual estimation signal supplied from -M e _{N-1, M,} Is input, where a demodulated signal R ₁ and an interference residual estimation signal e _{N, 1} for the first user are generated. Demodulated signal R ₁ is supplied to the base station host apparatus (not shown). Further, the interference residual estimation signal e _{N, 1} is supplied to the adder 18-N-1.

The adder 18-N-1 is connected to the ICU 14-N-
1 subtracts the interference residual estimation signal e _{N, 1} output from the error signal r _{N-1, M} passed through the delay element 16-N-1, and outputs it as an error signal r _{N, 1} . This error signal r _{N, 1}
Is supplied to the ICU 14-N-2 and the delay element 16-N-2.

The ICU 14-N-m includes an ICU 14-Nm provided in association with the same user in the previous stage.
Interference replica h _{N−1, m, k} output from (N−1) −m
And an error signal r _{N, m-1} are input, where a demodulated signal R _m for the m-th user and an interference residual estimation signal e _{N, m} are generated (2 ≦ m). ≤M-1). Demodulated signal R _m is supplied to the base station host apparatus (not shown). The interference residual estimation signal e _{N, m} is supplied to the adder 18-N-m.

The adder 18-N-m is connected to the ICU 14-N-
m, subtracts the interference residual estimation signal e _{N, m} output from the error signal r _{N, m-1} via the delay element 16-N-m, and outputs it as the error signal r _{N, m} (2 ≤m≤M-
1). The error signal r _{N, m} is (2 ≦ m ≦ M−
2), ICU14-N- (m + 1) and delay element 16-N
− ( _M + 1), and the error signal r _{N, M-1} is
14-N-M only.

In the ICU 14-NM, the ICU 14-NM provided in the previous stage in association with the M-th user is provided.
Interference replica h output from (N-1) -M
_{N-1, M, k} and ICU14-N-
The error signal r _{N, M-1} is input from (M-1). Then, where the demodulated signal R _M according to the M user are generated. This demodulated signal _{RM is} also supplied to a base station host device (not shown).

FIG. 9 is a diagram showing the configuration of the ICU 14-nm (1≤n≤N; 1≤m≤M). ICU14-n
The −m ICU main body 22 includes an error signal input terminal 64, an interference replica input terminal 66, an interference replica output terminal 68, and an interference residual estimation signal output terminal 70. ICU1
In principle, the error signal rn _{, m-1} is input to the error signal input terminal 64 provided for 4-nm, but when both n and m are 1, the received spread signal Rb is input. Also, n
Is greater than or equal to 2 and less than N and m is 1, the error signal r
_{n-1 and M} are input. Also, the interference replica hn- _{1, m, k} is input to the interference replica input terminal 66 provided in the ICU 14-nm in principle, but nothing is input when n is 1. Further, from the interference replica output terminal 68 provided in the ICU 14-nm, the interference replica h
_{n, m, k} are output, and an interference residual estimation signal output terminal 70 outputs an interference residual estimation signal en _{, m} . Also, ICU
14-n-m but applies to the interference cancellation device 10 shown in FIG. 7, this time, particularly in the N-th stage 12-N, the output of the hard decision unit 40 is externally output as a demodulated signal R _m (Not shown).

The ICU body 22 includes a front finger 28 and
The RAKE combining unit 38, the hard decision unit 40, and the latter-stage finger 42 are included. Here, a plurality of front-stage fingers 28 and rear-stage fingers 42 are provided in accordance with the multipath. Each front finger 28 includes a despreading unit 30,
It is configured to include an adder 32, a transmission path estimating unit 34, and a multiplier 36. On the other hand, the latter finger 42 includes a multiplier 44, an adder 46, and a spreading unit 48.

A spread signal such as an error signal r _{n, m-1} input from the error signal input terminal 64 is input in parallel to each of the pre-stage fingers 28, and a despreading unit 30 provided in each of them provides a pass signal for each pass. Despread. The spreading code used in the despreading unit 30 and the spreading unit 48 is provided from the level ranking circuit 20. The despread signal output from the despreading unit 30 is added by the adder 32 to the interference replica h _{n-1, m} (or null signal) supplied from the interference replica input terminal 66. The output of the adder 32 is supplied to a transmission path estimating unit 34 and a multiplier 36. The transmission path estimation unit 34 detects a phase difference and a signal level difference generated in the transmission path by using, for example, a pilot signal portion included in an input signal,
The transmission path estimation signal including the information is supplied to the multiplier 36 and the multiplier 44 included in the post-finger 42. The multiplier 36 multiplies the conjugate signal of the transmission path estimation signal by the output of the adder 32, thereby generating a signal in which the phase difference and the signal level difference generated in the transmission path are canceled.

In the RAKE synthesizing unit 38, each of the first-stage fingers 2
The signals generated by the eight multipliers 36 are in-phase and added. Then, the hard decision unit 40 makes a hard decision on the output signal of the RAKE combining unit 38. The result of the hard decision is again for the latter finger 4.
2 are input in parallel to each other and included in each of the multipliers 4
In step 4, the signal is multiplied by the transmission path estimation signal supplied from the transmission path estimation unit 34. Thereby, the phase difference and the signal level difference generated in the transmission path are added again to the hard decision result. The output of the multiplier 44 is output from the interference replica output terminal 68 as an interference replica hn _{, m, k} . That is, the interference replica h _{n, m, k} is an estimated signal (despread signal) related to the user signal of the m-th user included in the k-th path component of the received spread code Rb. This interference replica hn _{, m, k} is added to the adder 46.
Is also supplied.

The adder 46 is also supplied with an interference replica h _{n-1, m, k} given a negative sign, where both are added. In short, the adder 46 uses the interference replica h
A signal is generated by subtracting the interference replica hn _{-1 , m, k} from _{n, m, k} . The output of the adder 46 is supplied to a spreading unit 48, where the output is spread again with a spreading code corresponding to the m-th user. The signals output from the spreading section 48 of each subsequent finger 42 are added together in an adder 50 to generate an interference residual estimation signal en _{, m} . Interference residual estimation signal e _{n, m}
Is a signal _obtained by subtracting the interference replica h _{n−1 , m, k} from the interference replica h _{n, m, k} , frequency-spread again with a spreading code corresponding to the m-th user, and path-synthesizing it. But,
In short, the interference residual estimation signal en _{, m} is an estimation signal (spread signal) for the signal component of the m-th user included in the spread signal input from the error signal input terminal 64.

The interference canceling system according to the prior art described above.
In the device 10, the I stage provided at the top of the first stage
When the received spread signal Rb is input to the CU 14-1-1,
Interference replica h while stacking stages_{n, m, k}The accuracy of
Improve the interference residual estimation signal e_{n , m}And the error signal r_{n, m}Is zero
Approach. Then, in the final Nth stage, an interference replica
H_{n, m, k}Instead of the demodulated signal R_mIs each ICU14-N-
m, the demodulated signal R_mThe phase between users
Mutual interference can be reduced.

Japanese Patent Laid-Open Publication No. Hei 10-51353 discloses a successive processing type symbol replica interference canceling apparatus having a configuration similar to that of the interference canceling apparatus 10, but this apparatus is shown in FIG. Thus, the interference residual estimation signal e _{n, m} output from the adder 50
To the interference residual signal output terminal 70.
Output from. That is, the ICU 14a-nm shown in the same figure is used in place of the ICU 14-nm in the interference canceling apparatus 10 shown in FIG.
2, where the weight coefficient α of 1 or less supplied from the weight coefficient supply unit 50 is used as the interference residual estimation signal en _{, m.}
Is multiplied. The interference residual estimation signal e _{n, m} multiplied by the weight coefficient α is used as the corrected interference residual estimation signal e ′ _{n, m} as the interference residual estimation signal output terminal 7.
It is output from 0.

Multiplying the interference residual estimation signal en _{, m} by a weight coefficient α of 1 or less in this manner is considered to be a measure taking the following circumstances into consideration.

That is, when generating the interference replicas h _{n, m, k} at each ICU 14a-nm, transmission path estimation is performed for each user signal for each path. The replica h _{n, m, k and} , consequently, the interference residual estimation signal e _{n, m} generated based on the replica h _{n, m, k} have low reliability. Then, the interference replicas of the previous stage h _{n-1, m,} interference replica h _n of _k and the present _{stage, m,} and again diffuse the difference between _k and those RAKE combining, as interference residual estimation signal e _{n, When the} signal is output as _m , even if the reliability of the interference residual estimation signal e _{n, m} is low, the signal is received as the received spread signal Rb, the error signal rn _{-1, M} , or the error signal rn, _{m. It is} subtracted from _m-1 and supplied to another ICU 14a as a new error signal r _{n, m} . As a result, the interference replica and the interference residual estimation signal generated by the ICU 14a also have low reliability.

[0030] In this regard, Japanese Patent Application Laid-Open No.
In the interference canceling device according to the publication, the ICU 14a-
mn, the interference replica h _{n-1, m, k} of the previous stage
And the RAKE synthesis by spreading again the difference between the interference replica hn _{, m, k} of the current stage and a weight coefficient α of 1 or less.
And outputs it instead of the interference residual estimation signal e _{n, m} . As a result, even if the reliability of the interference replica generated by a certain ICU 14a is low,
The effect (interference) on the signal processing in U14a can be reduced.

[0031]

However, the interference canceling device according to the above-mentioned Japanese Patent Application Laid-Open No. 10-51353 can certainly reduce the applied interference, but has the following problems. That is, the interference canceling device according to the publication is based on the interference canceling device 10 shown in FIGS. 8 and 9, but the normal operation of the interference canceling device 10 is based on the input / output signals of the ICU 14 -nm. in the interference residual estimation signal e _n, component e _n of the k-th path _{m, m}
and ^{(k), ICU14- (n-} 1) interference replica outputted from -m h _{n-1, m,} and _k, interference replica h _n output from _{ICU14-n-m, m,} and _k, of It is premised that the relationship shown in the following equation (5) is established. FIG. 11 is a signal point arrangement diagram showing the relationship between these signals.

[0032]

[Number 2] ^{_{e n, m (k) =}} h n, m, k -h n-1, m, k ... (5) However, the ICU14a-n-m, serving input signal interference replica shown in FIG. 10 hn _{-1, m, k} and the interference replicas h _{n, m, k} and e ′ _{n, m} that are output signals do not have the same relationship as the above equation (5). That is, the following equation (6)
As shown in the figure, the interference replica h _{n, m, k} which is the output signal
And the difference between the input signal and the interference replica h _{n−1, m, k} does not match the ^k- _th path component e _{n, m} ^(k) of the output signal e ′ _{n, m} . FIG. 12 is a signal point arrangement diagram showing the relationship between these signals.

[0033]

Equation 3] ^{_{e 'n, m (k)}} = αe n, m (k) ≠ h n, m, k -h n-1, m, k ... (6) Therefore, the above-mentioned Japanese Unexamined Patent Publication No. 10-51353 Although the interference canceling device according to the publication can reduce the applied interference, when the weight coefficient α is set to a small value, the difference between the left side and the right side of the above equation (6) becomes large, and the convergence of the signal processing even when the stage advances May not proceed quickly. For this reason, there is a problem that the range which can be actually adopted as the value of the weight coefficient α is limited.

In this regard, the interference canceling device according to Japanese Patent Application Laid-Open No. H11-168408 solves such a problem by devising the multiplication position of the weight coefficient α in the ICU. FIG. 13 is a diagram showing a configuration of an ICU provided in the interference cancellation device according to the publication. In the ICU 14b-nm shown in FIG.
b, a multiplier 62b is provided at a stage subsequent to the hard decision section 40, and the weight coefficient supply section 6 is provided by the multiplier 62b.
The weight coefficient α supplied from 0b is multiplied by the hard decision result. The result of the multiplication is supplied to the second-stage finger 42. Thus, the required relationship between the input and output signals as described above is maintained. FIG. 14 is a signal point arrangement diagram showing a relationship between input and output signals of the ICU 14b-nm.

That is, in the ICU 14b-nm,
Interference replica h to be output_{n, m, k}Weights for
The signal multiplied by the number α is the corrected interference replica h ′
_{n, m, k}Output from the interference replica output terminal 68 as
From the interference residual estimation signal output terminal 70,
Interference replica h '_{n, m, k}From the previous stage
Modified interference replica generated in (n-1) stage
h '_{n-1, m, k}-Corrected interference based on subtracted signal
Residual estimation signal e '_{n, m}Is output. For this reason,
Corrected interference replica h '_{n, m, k}And a modified interference replica
H '_{n-1, m, k}And the corrected interference residual estimation signal e ′_{n, m}of
Component e 'related to the k-th pass_{n, m} ^(k)And the following equation (7)
The relationship shown in is satisfied.
You.

[0036]

## EQU4 ## However, e ' _{n, m} ^(k) = h' _{n, m, k} -h ' _{n-1, m, k} (7) However, in the interference canceling apparatus according to Japanese Patent Application Laid-Open No. H11-168408, However, there are the following problems.
That is, in this interference canceling device, the ICU 14b
The interference replica h that should have been originally created at -nm
A product obtained by multiplying _{n, m, k} by a weighting coefficient α is output as a modified interference replica h ′ _{n, m, k} . When the weighting coefficient α is set to a small value, the interference replica h _{n, m , k} changes greatly. As a result, the error signal r _{n, m} may not converge sufficiently with a limited number of stages, and a demodulated signal for each user may not be separated and extracted with sufficient BER characteristics.

The present invention has been made in view of the above problems, and an object of the present invention is to provide an interference canceling apparatus capable of suitably separating and extracting a demodulated signal for each user even with a small number of stages. .

[0038]

In order to solve the above-mentioned problems, an interference cancellation apparatus according to the present invention receives an error signal and a previous stage interference replica, and based on the input signals, an interference residual estimation signal and a current interference replica signal. In a multi-stage interference cancellation apparatus configured to include a plurality of interference cancellation units that output a stage interference replica,
The interference residual estimation signal is directly or indirectly corrected so that the signal level is reduced by the first attenuation, and the corrected interference residual estimation signal is output instead of the interference residual signal. An interference residual estimation signal modifying means for directly or indirectly modifying the current stage interference replica so that the signal level is reduced by an amount corresponding to the first attenuation, and Correction means for correcting the preceding stage interference replica so as to reduce the amount by an amount, and outputting a signal obtained by adding these signals in place of the current stage interference replica.

According to the present invention, in the interference cancellation unit, the interference residual estimation signal is directly or indirectly applied to the signal to reduce the signal level by the first attenuation, and the corrected interference residual estimation signal is originally used. Is output instead of the interference residual estimation signal. Also, the current stage interference replica is directly or indirectly corrected so that the signal level is reduced by the amount corresponding to the first attenuation, and the previous stage is corrected so that the signal level is reduced by the second attenuation. The interference replica is corrected, and a signal obtained by adding the corrected current stage interference replica and the previous stage interference replica is output instead of the current stage interference replica. According to the present invention, a signal in which the influence of the previous stage interference replica remains is output instead of the current stage interference replica, so that it is possible to prevent the interference replica from largely changing. The demodulated signal relating to the user can be suitably separated and extracted.

In one aspect of the present invention, the second attenuation is set so as to decrease as the first attenuation increases. By doing so, the level of the signal output in place of the current stage interference replica can be kept within a certain range.

In one aspect of the present invention, at least one of the first attenuation and the second attenuation is variably set according to the reliability of signal processing in the interference cancellation unit. This makes it possible to correct the interference replica and the interference residual estimation signal within a necessary range, and to appropriately separate and extract the demodulated signal for each user even with a smaller number of stages. .

Further, the interference canceling device according to the present invention
Is a receiving spread signal or an error signal,
To the first stage m-th user based on the signal or error signal
Such an interference residual estimation signal e_{1, m}And the first stage m you
The interference replica h according to the k-th path_{1, m, k}And output
The interference cancel unit is set to the first scan corresponding to the m-th user.
And the error signal, the (n−
1) Interference replica h related to stage mth user kth path
_{n-1, m, k}And are input, and based on the input signals,
Interference residual estimation signal e for n-stage m-th user
_{n, m}And the interference level related to the nth stage, the mth user, and the kth pass.
Prika h_{n, m, k}And output the interference cancellation unit
Is provided on the n-th stage corresponding to the m-th user.
In the multistage interference canceling apparatus, the first stage
The interference cancellation unit provided in the page,
Interference residual estimation signal e_{1, m}Modified interference residual estimation instead of
Signal e '_{1, m}-Stage interference residual estimation signal that outputs
Correction output means, and the interference replica h_{1, m, k}Repair instead of
Corrected interference replica h '_{1, m, k}First stage to output
Interference replica correction output means;
The interference canceling unit provided on the
Interference residual estimation signal e_{n, m}Instead of the modified interference residual estimation signal
No. e '_{n, m}N-th stage interference residual estimation signal correction
Positive output means, and the interference replica h_{n, m, k}Modified instead of
Interference replica h '_{n, m, k}Output the n-th stage
Said first stage comprising:
The interference residual estimation signal correction output means includes the corrected interference residual
Difference estimation signal e '_{1, m}A component e ′ related to the k-th pass_{1, m} ^(k)When
The interference residual estimation signal e_{1, m}The component e related to the k-th pass of_{1, m}
^(k)So that the above expression satisfies the following equation (8).
Negotiation residual estimation signal e '_{1, m}And generating the n-th stage
The interference residual estimation signal correction output means, wherein the corrected interference residual
Estimated signal e ' _{n, m}A component e ′ related to the k-th pass_{n, m} ^(k)And before
The interference residual estimation signal e_{n, m}The component e related to the k-th pass of_{n, m}
^(k)So that the modified dryness satisfies the following equation (9).
Negotiation residual estimation signal e '_{n, m}And the first stage is dried.
The correction replica correction output means satisfies the following expression (10).
The modified interference replica h '_{1, m, k}Produces
The n-th stage interference replica correction output means includes:
Said modified interference replica to satisfy equation (11)
h '_{n, m, k}(0 ≦ α)_{1, m, k}
≤1,0≤α_{n, m, k}≤1,0 <β_{n, m, k}≦ 1, ≦ n ≦
N, 2 ≦ m ≦ M, 1 ≦ k ≦ K, N ≧ 2, M ≧ 2, K ≧
1).

[0043]

[Mathematical formula-see original document] e '_{1, m} ^(k)= Α_{1, m, k}e_{1, m} ^(k) ... (8) e '_{n, m} ^(k)= Α_{n, m, k}e_{n, m} ^(k) ... (9) h '_{1, m, k}= Α_{1, m, k}h_{1, m, k} ... (10) h '_{n, m, k}= Α_{n, m, k}h_{n, m, k}+ Β_{n, m, k}h '_{n-1, m, k} (11) According to the present invention, the original interference replica h of the current stage
_{n, m, k}The coefficient α_{n, m , k}And the previous stay
Modified interference replica h '_{n-1, m, k}To the coefficient β_{n, m, k}
And the signal obtained by adding
Modified interference replica h ′_{n, m, k}Output as This
, The modified interference replica h ′ at the previous stage
_{n-1, m, k}Impact of the current stage of the modified interference replica
h '_{n, m, k}Interference cancellation
Modified interference replica h 'output from the unit
_{n, m, k}Can be prevented from greatly changing,
Demodulated signals for each user can be separated even if there are no stages.
It can be separated and extracted.

In one embodiment of the present invention, β _{n, m, k} = 1−α
_{n, m, k} are satisfied. This makes it possible to satisfy the required relationship between the input and output signals of the interference canceling unit, and to appropriately separate and extract the demodulated signal for each user even with a smaller number of stages.

In one embodiment of the present invention, at least one of α _{n, m, k} or β _{n, m, k} is variably set according to the reliability of signal processing in the interference cancellation unit.
This makes it possible to correct the interference replica and the interference residual estimation signal within a necessary range, and to appropriately separate and extract the demodulated signal for each user even with a smaller number of stages. .

[0046]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the drawings.

Embodiment 1 FIGS. 1 and 2 show I and I provided in interference cancellation apparatus according to Embodiment 1 of the present invention.
It is a figure showing CU. The basic configuration of the interference canceling apparatus according to the first embodiment is the same as that of the related art, and the ICU 14c-1-m shown in FIG. 1 is different from the ICU 14-1-m in the interference canceling apparatus 10 shown in FIG. It is used instead (1 ≦ m ≦ M). FIG.
Are used in place of the ICU 14-nm in the interference canceling device 10 shown in FIG. 8 (2 ≦ n ≦ N; 1 ≦ m ≦
M).

The ICU 14c-1-m shown in FIG.
Compared to the ICU 14-nm shown in FIG. 9, the point that the interference replica is not input to the front finger 28c provided in the ICU body 22c, the interference replica generated in the previous stage is supplied to the rear finger 42c. The difference is that there is no configuration and that a configuration for correcting the output signal is added. The other configuration is ICU1 shown in FIG.
Since this is the same as 4-nm, the same components are denoted by the same reference numerals, and detailed description thereof will be omitted.

First, in the ICU 14c-1-m, a multiplier 75 is provided after the adder 50. Multiplier 75
Is connected to a weighting factor supply unit 71, from which a weighting factor α is supplied. Then, the multiplier 75 multiplies the output of the adder 50, that is, the original interference residual estimation signal e _{1, m} by a weight coefficient α, and divides the multiplication result into the corrected interference residual estimation signal e ′ _{1, m} From the interference residual estimation signal output terminal 70.

The output of the multiplier 44 is supplied to a multiplier group 73. The multiplier group 73 includes a plurality of multipliers corresponding to each path, and each multiplier has a weight coefficient supply unit 6.
9 is supplied with a weighting coefficient α. Then, each multiplier belonging to the multiplier group 73 multiplies the output of the multiplier 44 relating to each path, that is, the original interference replica h _{1, m, k} by the weight coefficient α supplied from the weight coefficient supply unit 69. It is supposed to. The result of this multiplication is output from the interference replica output terminal 68 as a corrected interference replica h ′ _{1, m, k} .

On the other hand, the ICU 14c-n- shown in FIG.
m (n ≧ 2) is different from ICU14-nm shown in FIG. 9 in that the output of the adder 46 is extracted to generate an output signal at the subsequent finger 42d of the ICU main body 22d. And a configuration for correcting the output signal is added. Other configurations are the same as those in FIG. 9, and the same components are denoted by the same reference numerals, and detailed description thereof will not be repeated.

First, in the ICU 14c-nm (n ≧ 2), a multiplier 75 is provided after the adder 50.
Here, the original interference residual estimation signal e _{n, m} is multiplied by the weight coefficient α supplied from the weight coefficient supply unit 71. The result of the multiplication is output from the interference residual estimation signal output terminal 70 as a corrected interference residual estimation signal e ′ _{n, m} . Also, the multiplier 4
The output of 6 is taken out and supplied to each multiplier belonging to the multiplier group 73. The multiplier group 73 includes multipliers corresponding to each path, and each multiplier is supplied with a weight coefficient α from a weight coefficient supply unit 69. And
Each multiplier multiplies the output of the adder 46 by a weight coefficient α. The result of the multiplication is supplied to each adder belonging to the adder group 77. The adder group 77 includes adders corresponding to each path, and each adder is supplied with the corrected interference replica h ′ _{n−1, m, k} generated in the previous stage. Then, in each adder belonging to the adder group 77, the corrected interference replica h ′ _{n−1, m, k} is added to the signal output from the multiplier of each path belonging to the multiplier group 73. . The result of this addition is the corrected interference replica h ' _{n, m, k}
Is output from the interference replica output terminal 68.

The output of the adder 46 is spread again by the spreading section 48 and then path-combined by the adder 50, thereby generating an interference residual estimation signal en _{, m} . Therefore, the output of the adder 46 becomes each path component en _{, m} ^(k) of the interference residual estimation signal en _{, m} . In the second and subsequent stages, each of the multipliers belonging to the multiplier group 73 multiplies the path component en _{, m} ^(k) by a weighting factor α.

As described above, in the interference cancellation apparatus according to the first embodiment described above, the corrected interference residual estimation signal e ′ _{n, m as} the output signal and the corrected interference replica h ′ are output.
_{n, m, and k} satisfy the following expressions (12) to (14). FIG. 3 is a signal point arrangement diagram for explaining the relationship between these signals.

[0055]

E ′ _{n, m} ^(k) = αen _{, m} ^(k) (n ≧ 1) (12) h ′ _{1, m, k} = αh _{1, m, k} (13) h ′ _{n , m, k = αe n,} m (k) + h 'n-1, m, k = α (h n, m, k -h' n-1, m, k) + h 'n-1, m, k = Αh _{n, m, k} + (1−α) h ′ _{n−1, m, k} (n ≧ 2) (14) That is, after the second stage, the original interference replica h _{n, m, k} A signal obtained by adding a product obtained by multiplying the weighted coefficient α and a product obtained by multiplying the corrected interference replica h ′ _{n−1, m, k} of the previous stage by (1−α) is added to the corrected interference replica h ′ _{n , m, k} are output. That is, the influence of the corrected interference replica h ′ _{n−1, m, k} of the previous stage remains on the corrected interference replica h ′ _{n, m, k} which is the output signal. From the above equation (14), it can be seen that the input / output signals of the ICU 14c-nm satisfy the following equation (15). Where e '
_{n, m} ^(k) is a component of the interference residual estimation signal e ′ _{n, m} ^(k) related to the k-th path.

[0056]

[Mathematical formula-see original document] h '_{n, m, k}-H '_{n-1, m, k}= Αe_{n, m} ^(k)= E '_{n, m} ^(k) (15) Therefore, the interference canceling apparatus according to the first embodiment is provided.
According to the relationship required for input and output signals is satisfied,
Converges relatively quickly, suitable for small number of stages
A demodulated signal for each user can be generated. Ma
The corrected interference replica h 'as an output signal_{n, m, k}Before
Stage modified interference replica h '_{n-1, m , k}The effect of
Modified interference replica because it is left to survive
h '_{n, m, k}Can be prevented from changing greatly,
Suitable for each user even with a small number of stages compared to conventional technology
Such a demodulated signal can be generated.

In the above description, the weight coefficient supply unit 71
And the same weighting factor α is supplied by the weighting factor supply unit 69, but different values may be set as long as the convergence of the signal processing is ensured. That is, in the above expression (14), in the second term on the right side of the third expression, the corrected previous stage interference replica h ′ _{n−1, m, k} is _represented by (1−1).
α), this (1−α) represents the survival rate of the previous stage interference replica h ′ _{n−1, m, k in} the corrected current stage interference replica h ′ _{n, m, k} . Therefore, another value (β _{n, m, k} ) may be adopted as long as the above expression (15) is not satisfied.

In the above description, the weight coefficient supply unit 71 and the weight coefficient supply unit 69 are shared by all paths. However, the weight coefficient supply unit 71 and the weight coefficient supply unit 69 may be provided for each path. That is, the weighting factor supply unit 71 and the weighting factor supply unit 69 may output the weighting factors α _{n, m, k} . Furthermore, the weighting factor α supplied from the weighting factor supply unit 71 or the weighting factor supply unit 69 may be fixedly set, or may be set according to the reliability of signal processing in the ICU 14c-nm, etc. It may be set dynamically.
Further, the weight coefficient α does not need to be common to all stages, and for example, the value of the weight coefficient α may increase as the stages progress.

Embodiment 2 FIG. 4 is a diagram showing a configuration of an ICU provided in the interference cancellation apparatus according to Embodiment 2 of the present invention. The basic configuration of the interference canceling apparatus according to the second embodiment is also the same as that of the related art, and the ICU 14e-nm shown in the same drawing is different from the ICU 14-nm in the interference canceling apparatus 10 shown in FIG. It is used instead (1 ≦ m ≦ M).

The ICU 14e-nm shown in FIG.
The difference from the ICU 14-nm shown in FIG. 9 lies in the configuration of the subsequent finger 42e included in the ICU main body 22e and the point that an adder group 77 is provided before the interference replica output terminal 68. Other configurations are the same, and the same reference numerals are given to the same reference numerals, and the detailed description is omitted here. In FIG.
In, the output of the adder 46 is supplied to the multiplier 81.
The weighting coefficient α is supplied from the weighting coefficient supply unit 79 to the multiplier 81 of each path, and the weighting coefficient α is output to the output of the adder 46.
Is multiplied. The result of the multiplication is taken out from the latter-stage finger 42e and supplied to each adder belonging to the adder group 77. Each adder belonging to the adder group 77 has a corrected interference replica h ′ of the previous stage related to each path.
_{n, m, and k} are input, and are added to the multiplication result in the multiplier 81 here. This addition result is output from the interference replica output terminal 68 to the corrected interference replica h ′.
Output as _{n, m, k} .

Also in this case, similar to the first embodiment,
Equations (12) to (15) can be satisfied, and
The relationship required for input / output signals can be satisfied. In addition, _since the effect of the corrected interference replica h ′ _{n−1, m, k} of the previous stage can remain in the corrected interference replica h ′ _{n, m, k} which is the output signal, the corrected interference replica h ′ _{n , m, and k} are prevented from greatly changing, and a demodulated signal for each user can be suitably generated with a smaller number of stages as compared with the related art.

Embodiment 3 FIGS. 5 and 6 show I and I provided in interference cancellation apparatus according to Embodiment 3 of the present invention.
It is a figure showing composition of CU. 5 and 6 show each ICU.
The weight coefficient α is dynamically generated in 14g-nm,
The output signal is modified to the extent necessary.

The ICU 14f-1-m shown in FIG.
The ICU 14f-nm (n ≧ 2) shown in FIG. 6 is used in place of the ICU 14c-1-m shown in FIG.
Is used in place of 14c-nm (n ≧ 2) shown in FIG. This ICU 14f-nm (n
If ≧ 1), the output of the subsequent finger 42c is provided to the pass-by-pass correction processing unit 72.

In this embodiment, the rear finger 4
2 (herein, referred to as “path-specific interference residual estimation signal e _{n, m} ^(k) ”) are supplied to adders 74 and multipliers 84 provided in corresponding path-specific correction processing units 72. Is done. The signal ^obtained by performing path synthesis on the path-based interference residual estimation signal e _{n, m} ^(k) becomes the original interference residual estimation signal en _{, m} . The interference residual estimation signal e _{n, m} (i) for each path is given a negative sign and supplied to the adder 74. The error signal r _{n, m-1 and the} like input from the error signal input terminal 64 are
6, the delay is given by the processing time in the ICU main body 22, and distributed to each adder 74 included in the path-specific correction processing unit 72. In the adder 74, as shown in the following equations (16) to (18), the signal d _{n, m} corresponding to the difference between the two signals is obtained.
^(k) (i) is generated.

[0065]

(Equation 8) d_{n, m} ^(k)(I) = r_{n, m-1}(I) -e_{n, m} ^(k)(I) ... (16)-When m is 1 d_{n, m} ^(k)(I) = r_{n-1, M}(I) -e_{n, m} ^(k)(I) ... (17) ・ When both n and m are 1 d_{n, m} ^(k)(I) = Rb (i) -e_{n, m} ^(k)(I) ... (18) Here, i is the time measured in a chip unit. This message
No. d_{n, m} ^(k)(I) is supplied to the adder 78, where
And the signal d in each chip for the symbol time
_{n, m} ^(k)The power levels of (i) are sequentially added. For example,
When the dispersion is 64, the signal d for 64 chips_{n, m}(I)
The power level is added. The value is supplied to the evaluation unit 80.
It is.

The error signal r supplied from the delay element 26
_{n, m-1 and the} like are supplied to the adder 78 and the adder 7
6 is also supplied. Here, the power levels at each chip time such as the error signal r _{n, m−1} are sequentially added over one symbol time. For example, when the spreading factor is 64, the power levels of the error signals r _{n, m-1} (i) for 64 chips are added. The value is supplied to the evaluation unit 80.

The evaluation unit 80 calculates the following equations (19) to (21)
Is evaluated, the reliability of the per-path interference residual estimation signal e _{n, m} ^{(k) (} the reliability of signal processing performed by the ICU 14f-nm) is evaluated. Here, Σ _i means an operation of sequentially adding the values of each chip over one symbol.

[0068]

９ _i | d _{n, m} ^(k) (i) | ² > Σ _i | r _{n, m-1} (i) | ² (19) • When m is 1, Σ _i | d _{n, m} ^(k) (i) | ² > Σ _i | r _{n−1, M} (i) | ² (20) • When both n and m are 1 Σ _i | d _{n, m} ^(k) (i) | ² > Σ _i | Rb (i) | ² (21) In the evaluation section 80, when the equations (19) to (21) are not satisfied, the interference residual estimation signal e _{n, m} ^{(k) for each} path. Is determined to be highly reliable, and the weighting factor supply unit 82 is instructed to output a weighting factor α that is 1 or close to 1. On the other hand, if equations (19) to (21) are satisfied, it is determined that the reliability of the per-path interference residual estimation signal en _{, m} ^(k) is low,
It instructs the weight coefficient supply unit 82 to output a weight coefficient α smaller than 1 and smaller than the previous weight coefficient α. The weight coefficient α according to the instruction of the evaluation section 80 is output from the weight coefficient supply section 82, and is supplied to the multiplier 84. The multiplier 84 multiplies the per-path interference residual estimation signal e _{n, m} ^(k) by the chip and the weighting coefficient α, and generates a corrected per-path interference residual estimation signal e ′ _{n, m} ^(k). Is done.
The corrected path-based interference residual estimation signal e ′ _{n, m} ^(k) is supplied to the adder 86, where all the path-based correction processing units 72
Modified interference residual estimation signal e ′ for each path output from
_{n, m} ^(k) are added to generate a modified interference residual estimation signal e ′ _{n, m which} is an output signal.

Each of the multipliers included in the multiplier group 73 includes a weight coefficient supply unit 8 included in the path-specific correction processing unit 72.
The same as 2 is connected. The evaluation result of the evaluation unit 80 for the same path is input to the multiplier corresponding to each path of the multiplier group 73 (control lines are not shown), and the same value as the multiplier 84 for the same path. Are supplied. Thus, the weighting factor α is dynamically set according to the reliability of the signal processing in the ICU 14f-nm, and the corrected interference residual estimation signal e ′ _{n, m} as the output signal and the corrected interference replica h ′ _{n , m, k} are generated.

The ICU 14f-n- according to the third embodiment
In m, since the weighting coefficient α is dynamically set according to the reliability of the signal processing, the output signal can be corrected within a necessary range, and the demodulation for each user can be suitably performed with a smaller number of stages. it becomes possible to obtain a signal R _m.
The reliability is evaluated for each path-based interference residual estimation signal e ′ _{n, m} ^(k) , and the weight coefficient α is set according to the evaluation result. Therefore, even when the reliability of the transmission path estimation varies between paths, an appropriate corrected interference residual estimation signal e ′ _{n, m} can be output. As a result, it is possible to accelerate error signal r _n, the convergence of _m such, B of the demodulated signal R _m
ER characteristics can be improved.

Embodiment 4 Embodiments 1 to 3 above
Although the interference canceling device according to (1) employs the same basic configuration as the interference canceling device 10 shown in FIG. 8 (sequential processing type), a parallel processing type may be adopted. FIG. 7 is a diagram illustrating a configuration of a parallel processing type interference cancellation device. The interference cancellation apparatus 100 shown in FIG. 1 is provided in a DS-CDMA base station, and the ICU according to Embodiments 1 to 3 can be applied to this interference cancellation apparatus 100 as well.

The interference canceling apparatus 100 includes a first stage 102-1 to an N-th stage 102-N. While reducing interference
And outputs a demodulated signal R ₁ to R _M of the first user, second M user. Here, N and M are both integers of 1 or more.

In the first stage 102-1, ICUs 104-1-1 to ICU 104-1-1 to MCU 104-1-1 correspond to M users, respectively.
4-1-M are connected in parallel. The first stage 102-1 has a delay element 106-1 and an adder 10
8-1.

Similarly, in each of the second stage 102-2 to the (N-1) th stage 102- (N-1), the ICU 104-n- corresponding to each of the M users is similarly provided.
1 to 104-nM are connected in parallel (2 ≦ n ≦
N-1). The n-th stage 102-n includes a delay element 106-n and an adder 108-n (2 ≦ n ≦ N−1).

The Nth stage 102-which is the final stage
N also has ICUs 104-N-1 to 104-N-M connected in parallel corresponding to each of the M users, but does not require delay elements and adders. IC
The demodulated signal R ₁ is output from U104-N- ₁ to 104-NM.
To R _M are output.

The ICU 104-n provided for each stage
For -m, the ICU described in Embodiment Modes 1 to 3 can be used. Further, when the interference cancellation apparatus 100 is realized by hardware, it is not necessary to actually prepare hardware for all stages. For example, hardware may be provided for only one stage, and the hardware may be sequentially used as hardware for the next stage.

IC provided on first stage 102-1
The received spread signal Rb is input to each of U104-1-1 to 104-1-M. That is, all the ICUs 104-1-1 to 1-1-1 provided in the first stage 102-1
In 104-1-M, the received spread signal Rb is input to both error signal input terminals 64 (see FIG. 2) provided therein. ICU104-1-1-10
4-1-M outputs interference residual signals e _{1, m} (1 ≦ m ≦ M).
08-1. The received spread signal Rb passed through the delay element 106-1 is also input to the adder 108-1, and a signal obtained by subtracting all the interference residual estimation signals e _{1, m} from the received spread signal Rb is sent to the next stage. supplied as an error signal r _1. The delay element 106-1 is connected to the ICU 104-
This circuit delays the input signal by the time required for 1-m signal processing. Further, interference replicas h _{1, m} are output from ICU 104-1-m, respectively (1 ≦ m ≦
M), and they are supplied to the corresponding ICU 104-2-m provided in the next stage.

In the n-th stage 102-n, the ICU 10
4-nm, the error signal r generated in the previous stage
_n-1 is input (2 ≦ n ≦ N−1; 1 ≦ m ≦
M). In other words, in all the ICUs 104-nm provided in the n-th stage 102-n, (2 ≦ n ≦ N−
1; 1 ≦ m ≦ M), and the error signal r _n−1 generated in the previous stage is input to the error signal input terminal 64 provided therein.

In the ICU 104-nm, the error signal r
_n-1 and the interference replica h _{n-1, m} supplied from the previous stage
And the interference residual signal e _{n, m} is generated based on (2 ≦
n ≦ N−1; 1 ≦ m ≦ M). They are provided with a negative sign and are provided to adder 108-n. Adder 108-
An error signal rn _-1 via the delay element 106-n is also input to _n, and a signal obtained by subtracting all the interference residual estimation signals en _{, m} from the error signal rn _-1 is sent to the next stage. supplied as an error signal r _n. The delay element 106-n is an ICU
This is a circuit for delaying an input signal by a time required for 104-nm signal processing. Further, ICU104-n-
m output interference replicas h _{n, m} (2 ≦ n ≦ N−1; 1 ≦ m ≦ M), respectively, which are provided in the corresponding ICU 104- (n + 1) in the next stage.
-M.

In the N-th stage 102-N, the ICU 10
4-N-m, the error signal r generated in the previous stage
_N-1 is input (1 ≦ m ≦ M). In other words, all the ICUs 104 provided in the N-th stage 102-N
At −N−m (1 ≦ m ≦ M), the error signal r _N−1 generated in the previous stage is input to the error signal input terminal 64 provided therein.

In the ICU 104-N-m, the error signal r
_{Based on N−1} and the interference replica h _{N−1, m} supplied from the corresponding ICU 104- (N−1) -m of the (N−1) th stage 102- (N−1), A demodulated signal R _m for the user is generated (1 ≦ m ≦ M). After the demodulated signal R _m is an externally output, it is sent to the base station host apparatus (not shown).

In the interference canceling apparatus 100 described above, the use of the ICU described in the first to third embodiments can satisfy the required relationship for the input / output signals of the ICU 104. In addition, _since the effect of the corrected interference replica h ′ _{n−1, m, k} of the previous stage can remain in the corrected interference replica h ′ _{n, m, k} which is the output signal, the corrected interference replica h ′ _{n , m, and k} are prevented from greatly changing, and a demodulated signal for each user can be suitably generated with a smaller number of stages as compared with the related art.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of an ICU (for a first stage) used in an interference cancellation device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of an ICU (for the second and subsequent stages) used in the interference cancellation device according to the first embodiment of the present invention.

FIG. 3 is a signal point arrangement diagram illustrating a relationship between input and output signals in an ICU used in the interference cancellation device according to the first embodiment of the present invention.

FIG. 4 is a diagram showing a configuration of an ICU used in an interference cancellation apparatus according to Embodiment 2 of the present invention.

FIG. 5 is a diagram showing a configuration of an ICU (for a first stage) used in an interference cancellation device according to a third embodiment of the present invention.

FIG. 6 is a diagram showing a configuration of an ICU (for the second and subsequent stages) used in the interference cancellation device according to the third embodiment of the present invention.

FIG. 7 is a diagram illustrating a general configuration of an interference cancellation apparatus (parallel processing type).

FIG. 8 is a diagram illustrating a general configuration of an interference cancellation apparatus (sequential processing type).

FIG. 9 is a diagram showing a configuration of an ICU according to the related art.

FIG. 10 is a diagram showing a configuration of an ICU (with an output signal correcting function) according to the related art.

11 is a signal point arrangement diagram illustrating a relationship between input and output signals in the ICU shown in FIG. 9;

12 is a signal point arrangement diagram illustrating a relationship between input and output signals in the ICU shown in FIG. 10;

FIG. 13 is a diagram showing a configuration of an ICU (with a function of correcting an output signal) used in an interference canceling device according to the related art.

14 is a signal point arrangement diagram illustrating a relationship between input and output signals in the ICU shown in FIG. 13;

[Explanation of symbols]

10, 100 interference canceling device, 12, 102 stage, 14, 14a, 104 ICU (interference canceling unit), 16, 26, 106, 205 delay element,
18, 32, 46, 50, 74, 76, 78, 86, 1
08, 208, 308 Adder, 20 level ranking circuit, 22, 22a, 22b ICU main body, 28, 2
8c pre-stage finger, 30 despreading unit, 34 channel estimation unit, 36, 44, 73, 75, 81, 84 multiplier,
38 RAKE combining section, 40 hard decision section, 42, 42c,
42d, 42e latter finger, 48 diffuser, 60,
69, 71, 79, 82 weight coefficient supply unit, 64 error signal input terminal, 66 interference replica input terminal, 68 interference replica output terminal, 70 interference residual estimation signal output terminal,
72 path-specific correction processing unit, 80 evaluation unit.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5K022 EE02 EE35 5K052 AA01 BB01 CC00 CC06 DD04 FF32 GG19 GG20 5K059 CC07 DD31 EE02

Claims (6)

[Claims] 1. A multi-stage interference canceling apparatus configured to include a plurality of interference cancellation units that receive an error signal and a previous-stage interference replica and output an interference residual estimation signal and a current-stage interference replica based on the input signals. In the above, the interference residual estimation signal is directly or indirectly corrected so that the signal level is reduced by the first attenuation, and the corrected interference residual estimation signal is replaced with the interference residual signal. Means for correcting an interference residual estimation signal to be output; and directly or indirectly correcting the current stage interference replica so that the signal level is reduced by an amount corresponding to the first attenuation, and The previous stage interference replica is modified so as to be reduced by the amount of attenuation of Interference canceling apparatus characterized by including, an interference replica modifying means for outputting instead of the replica. 2. The interference cancellation apparatus according to claim 1, wherein the second attenuation is set so as to decrease as the first attenuation increases. 3. The interference cancellation device according to claim 1, wherein at least one of the first attenuation amount and the second attenuation amount is determined according to reliability of signal processing in the interference cancellation unit. An interference canceling device variably set. 4. A receiving spread signal or an error signal is input,
Based on the received spread signal or error signal, an interference residual estimation signal e _{1, m} for the first stage m-th user, an interference replica h _{1, m, k} for the first stage m-th user k-th path,
Is provided on the first stage corresponding to the m-th user, and an error signal;
The interference replica h _{n−1, m, k} related to the (n−1) th stage, the mth user, and the kth path is input, and based on the input signals, an interference residual estimation signal for the nth stage, the mth user. a multi-stage interference unit in which an interference canceling unit that outputs an e _{n, m} and an interference replica h _{n, m, k} relating to an n-th stage and an m-th user and a k-th path is provided in the n-th stage corresponding to the m-th user in cancellation device, wherein the interference canceling unit provided in the first stage, the interference residual estimation signal e _1, modified interference residual estimation signal e in place of _{m '1,} the first stage interference residual of outputting the _m wherein the difference estimation signal modifier outputting means, the interference replica h _{1, m,} corrected interference replica h instead of _{k '1, m,} a first stage interference replica corrected output means for outputting a _k, a, the n-th The interference cap provided on the stage Cell unit, the interference residual estimation signal e _n, and the n-stage interference residual estimation signal corrected output means for outputting the corrected interference residual estimation signal e _{'n, m} instead of _m, the interference replica h _{n , m, k} in place of an n-th stage interference replica correction output means for outputting a corrected interference replica h ′ _{n, m, k} , wherein the first stage interference residual estimation signal correction output means includes component e ₁ according to the finished interference residual estimation signal e _'1, component e according to the k paths of _{m' 1, m} ^(k) and the interference residual estimation signal e _1, k-th path _{m, m} ^{(k )} And the modified interference residual estimation signal e ′ _{1, m} are generated such that the following equation (1) is satisfied. estimate signal e _'n, component e according to the k paths of _{m' n, m} ^(k) and the interference residual estimation signal e _n, component e _n of the k-th path _{m, m} ^(k) and the following Expression ( ) The modified interference residual estimation signal e so as to satisfy the 'generates _{n, m,} the first stage interference replica modified output means, the following equation (3) the modified interference replica h to satisfy'
_{1, m, k,} and the n-th stage interference replica correction output means outputs the corrected interference replica h ′ so as to satisfy the following equation (4).
generating an _{n, m, k} interference canceller (0 ≦ α _{1, m, k} ≦
1,0 ≦ α _{n, m, k} ≦ 1,0 <β _{n, m, k} ≦ 1,2 ≦ n ≦ N,
2 ≦ m ≦ M, 1 ≦ k ≦ K, N ≧ 2, M ≧ 2, K ≧ 1). E ′ _{1, m} ^(k) = α _{1, m, k} e _{1, m} ^(k) (1) e ′ _{n, m} ^(k) = α _{n, m, k} en _{, m} ^{( k)} … (2) h ′ _{1, m, k} = α _{1, m, k} hn _{, m, k} … (3) h ′ _{n, m, k} = α _{n, m, k} hn _{, m, k} + Βn _{, m, k} h'n _{-1, m, k} ... (4) 5. The interference cancellation apparatus according to claim 1, wherein β _{n, m, k} = 1−α _{n, m, k} is satisfied. 6. The interference cancellation apparatus according to claim 1 _, wherein at least one of α _{n, m, k} or β _{n, m, k} is variably set according to reliability of signal processing in the interference cancellation unit. An interference cancellation device characterized by being performed.