Method of double weighting parallel interference cancellation

• US 20040246927A1
• Filed: 04/08/2004
• Published: 12/09/2004
• Est. Priority Date: 10/09/2001
• Status: Active Grant

• Current Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Inventors: Jin, Yu, Wei, Limei, Wang, Guolin
• Granted Patent: US 7,321,581 B2
• Time in Patent Office: Days
• US Class Current: 370/335
• CPC Class Codes: H04B 1/71075 Parallel interference cancellation
• CPC Class Codes: H04B 2001/71077 Partial interference cancellation

Abstract

A method for double weighting parallel interference cancellation in CDMA mobile communication system, integrates ideas of both partial weighting method and weighting method based on Bayes rule. When making a decision on each symbol of the user, calculating reliability coefficient of a decision result on the symbol of the user according to the weighting algorithm based on Bayes rule; weighting regenerated signal of the symbol of the user with the reliability coefficient. During process of MAI estimating and removing, obtaining the MAI on the expected user from weighted regenerated signal of other users in chip level; and then setting a weight value, weighting the MAI with the weight value. Finally, removing weighted MAI from received signal, which means partially removing MAI produced by other users on the expected user. At the same time, the present invention also discloses a double weighting PIC method with simplified algorithm, which transfers weighting in chip level to weighting in symbol level, replacing a Hyperbolic tangent decision with a piecewise linear decision method or a look-up table method.

Claims

1. A double weighting parallel interference cancellation method that can be used in a CDMA mobile communication system, comprises:

(a) Making a multi-path despreading, channel estimation and a multi-path combining on an input signal of a user by a RAKE receiver;

(b) Making a hard decision on a result of the multi-path combining;

(c) Calculating a reliability coefficient for a result of the hard decision of each symbol by the multi-path combining result and a value of the channel estimation;

(d) Regenerating a weighted signal in chip level for the user by the hard decision result, the reliability coefficient and the channel estimation value;

(e) Estimating a multiple access interference on each user by the weighted regenerated signal in chip level inputted in parallel of each user, and setting a weight value for a interference cancellation in different stages, weighting the multiple access interference on a expected user;

(f) Subtracting the weighted multiple access interference on the expected user from a base band signal of a received signal, and obtaining an output signal of the said expected user, which is also an input signal of the said expected user in the next stage parallel interference cancellation.

2. The double weighting parallel interference cancellation method according to claim 1, comprises, if the said input signal of user i in the kth stage parallel interference cancellation is expressed as r_i^(k)(t),

In step (a), calculating the said multi-path combining result of user i by a formula Y_i^(m)(k)=Re{y_i^(m)(k)}, which can be expressed as Y_i^(m)(k)=μ

_ia_i^(m)+n_i, where, n_i is a Gaussian White Noise, which is subject to a normal distribution of N(0,σ

_i²), σ

² is a noise power of n_i, a_i^(m) is the mth symbol of user i, whose value is +1 or −

1, μ

_i is a real number that is related with channel fading;

In step (b), calculating the said hard decision result of the mth symbol of user i by a formula â

_i^(m)(k)=sgn(Y_i^(m)(k));

In step (c), calculating the said reliability coefficient of the said decision result on the mth symbol of user i by

$f_i^{\left(m\right)<br><br>\left(k\right)}=\tanh <br><br>\left\{w<br><br>\frac{{\mathrm{\mu <br><br>}}_i<br><br><br><br>Y_i^{\left(m\right)<br><br>\left(k\right)}<br><br>}{{\mathrm{\sigma <br><br>}}_i^2}\right\},$

where, w is a positive real number;

In step (d), calculating the said weighted regenerated signal of user i in chip level by a formula

$g_i^{\left(k\right)}<br><br>\left(t\right)=\underset{l=1}{\overset{L}{\sum <br><br>}}<br><br>A_{\mathrm{il}}<br><br>\underset{n=-\mathrm{\infty <br><br>}}{\overset{\mathrm{\infty <br><br>}}{\sum <br><br>}}<br><br>f_i^{\left(n\right)<br><br>\left(k\right)}<br><br>{\hat{a}}_i^{\left(n\right)<br><br>\left(k\right)}<br><br>p<br><br>\left(t-n<br><br>\text{\hspace{1em}<br><br>}<br><br>T_b-{\mathrm{\tau <br><br>}}_{\mathrm{il}}\right)<br><br>c_i<br><br>\left(t-{\mathrm{\tau <br><br>}}_{\mathrm{il}}\right),$

where, A_il is a estimation value of a_il{square root}{square root over (P_i)}, a_il refers to a channel fading value of user i in the path l, p_i is a power of user i;

In step (e), calculating the said multiple access interference estimation of user i by a formula

${\hat{I}}_i^{\left(k\right)}=\underset{j=1,j\ne <br><br>i}{\overset{N}{\sum <br><br>}}<br><br>g_j^{\left(k\right)}<br><br>\left(t\right);<br><br>$

Assuming the said weight value of the kth stage parallel interference cancellation is p^(k), then calculating the said output signal of user i in step (f) by a formula r_i^(k+1)(t)=r(t)−

p^(k)Î

_i^(k), where, r(t) is the said base band signal of the said received signal.

3. The double weighting parallel interference cancellation method according to claim 2, wherein step (c) of calculating the said reliability coefficient of the said decision result on the mth symbol of user i comprises, replacing the said Hyperbolic tangent decision with a piecewise linear decision, which means approaching a Hyperbolic tangent function tanh(x) with a piecewise linear decision function L(x), a expression of the piecewise linear decision function L(x) is:

4. The double weighting parallel interference cancellation method according to claim 2, wherein step (c) of calculating the said reliability coefficient of the said decision result on the mth symbol of user i comprises, replacing the said Hyperbolic tangent decision with a look-up table method, which means approaching a Hyperbolic tangent function tanh(x) with a decision function T(x) of the look-up table method, a expression of the said decision function T(x) of the look-up table method is:

5. The double weighting parallel interference cancellation method according to claim 1, wherein step (e) comprises, setting the different weight value p^(k) for the interference cancellation in different stages, wherein p⁽¹⁾<

p⁽²⁾ . . . <

p^(k) . . . <

p^(S), where, k is the kth stage of the interference cancellation, and S is a stage number of the interference cancellation.

6. A double weighting parallel interference cancellation method that can be used in CDMA mobile communication system, comprises:

(a) Making a multi-path despreading, channel estimation and a multi-path combining on an input signal of a user by a RAKE receiver;

(b) Making a soft decision on a result of the multi-path combining and a value of the channel estimation;

(c) Setting a weight value for the soft decision in different stages, and weighting the soft decision in symbol level;

(d) Regenerating a weighted signal of the user in chip level by the weighted soft decision result and the channel estimation value;

(e) Estimating multiple access interference on each user by the weighted regenerated signal in chip level inputted in parallel of each user;

(f) Subtracting the multiple access interference on an expected user from a base band signal of a received signal, and obtaining an output signal of the expected user, which is also an input signal of the said expected user in the next stage of parallel interference cancellation.

7. The double weighting parallel interference cancellation method according to claim 6, comprises, if the said input signal of user i in the kth stage parallel interference cancellation is expressed as r_i^(k)(t),

In step (a), calculating the said multi-path combining result of user i by a formula Y_i^(m)(k)=Re{y_i^(m)(k)}, which can be expressed as Y_i^(m)(k)=μ

_ia_i^(m)+n_i, where, n_i is a Gaussian White Noise, which is subject to a normal distribution of N(0,σ

_i²), σ

² is a noise power of n_i, a_i^(m) is the mth symbol of user i, whose value is +1 or −

1, μ

_i is a real number that is related with channel fading;

In step (b), expressing the said soft decision of the said multi-path combining result for user i as ζ

_i^(m)(k)=ƒ

_i^(m)(k)â

_i^(m)(k), calculating the said soft decision by a formula

$f_i^{\left(m\right)<br><br>\left(k\right)}<br><br>{\hat{a}}_i^{\left(m\right)<br><br>\left(k\right)}=\tanh <br><br>\left\{w<br><br>\frac{{\mathrm{\mu <br><br>}}_i<br><br>Y_i^{\left(m\right)<br><br>\left(k\right)}}{{\mathrm{\sigma <br><br>}}_i^2}\right\},$

where, w is a positive real number, â

_i^(m)(k) is a soft decision result for the mth symbol of user i, f_i^(m)(k) is a reliability coefficient of â

_i^(m)(k);

Assuming the said weight value of the said soft decision in the kth stage is p^(k), and then in step (c) calculating the said weighted soft decision of user i in symbol level by a formula ρ

_i^(m)(k)=ζ

_i^(m)(k)·

p^(k);

In step (d), calculating the said weighted regenerated signal in chip level of user i by a formula

$g_i^{\left(k\right)}<br><br>\left(t\right)=\underset{l=1}{\overset{L}{\sum <br><br>}}<br><br>A_{\mathrm{il}}<br><br>\underset{n=-\mathrm{\infty <br><br>}}{\overset{\mathrm{\infty <br><br>}}{\sum <br><br>}}<br><br>{\mathrm{\rho <br><br>}}_i^{\left(n\right)<br><br>\left(k\right)}<br><br>{\hat{a}}_i^{\left(n\right)<br><br>\left(k\right)}<br><br>p<br><br>\left(t-n<br><br>\text{\hspace{1em}<br><br>}<br><br>T_b-{\mathrm{\tau <br><br>}}_{\mathrm{il}}\right)<br><br>c_i<br><br>\left(t-{\mathrm{\tau <br><br>}}_{\mathrm{il}}\right),$

where, A_il is a estimation value of a_il{square root}{square root over (P_i)}, a_il refers to a channel fading value of user i in the path l, and p_i is a power of user i;

In step (e), calculating the said multiple access interference on user i by a

${\hat{I}}_i^{\left(k\right)}=\underset{j=1,j\ne <br><br>i}{\overset{N}{\sum <br><br>}}<br><br>g_j^{\left(k\right)}<br><br>\left(t\right);<br><br>$

In step (f), calculating the said output signal of user i by a formula r_i^(k+1)(t)=r(t)−

Î

_i^(k), where, r(t) is the said base band signal of the said received signal.

8. The double weighting parallel interference cancellation method according to claim 7, wherein step (b) of making the said soft decision of the said multi-path combining result for user i comprises, replacing the said Hyperbolic tangent decision with a piecewise linear decision, which means approaching a Hyperbolic tangent function tanh(x) with a piecewise linear decision function L(x), a expression of the piecewise linear decision function L(x) is:

9. The double weighting parallel interference cancellation method according to claim 7, wherein step (b) of making the said soft decision of the said multi-path combining result for user i comprises, replacing the said Hyperbolic tangent decision with a look-up table method, which means approaching a Hyperbolic tangent function tanh(x) with a decision function T(x) of the look-up table method, a expression of the said decision function T(x) of the said look-up table method is:

10. The double weighting parallel interference cancellation method according to claim 6, wherein step (c) comprises, setting the different weight value p^(k) for the soft decision in different stages, wherein p⁽¹⁾<

p⁽²⁾ . . . <

p^(k) . . . <

p^(S), where, k is the kth stage of the interference cancellation, and S is a stage number of the interference cancellation.