WO2002049301A1

WO2002049301A1 - A channel estimating method

Info

Publication number: WO2002049301A1
Application number: PCT/CN2000/000576
Authority: WO
Inventors: Michael Li
Original assignee: Linkair Communications, Inc.
Priority date: 2000-12-14
Filing date: 2000-12-14
Publication date: 2002-06-20
Also published as: AU2001221413A1; CN1152537C; HK1050783A1; CN1385017A

Abstract

The present invention discloses a channel estimating method using in the cellular mobile communication system, which transmits different pilot symbols having the correlation in both interference and fading in the same subframe at the transmitter side, at the receiver side, performs the difference operation on the received adjacent different pilot symbols, so as to eliminate interference and obtain the exact channel estimating value. The channel estimating method provided by the present invention can effectively eliminate the affect of the interference, give accurate channel parameter estimating value, avoid complex multi-users detection and reduce the complexity of the system.

Description

Channel estimation method

Technical field

The present invention relates to a cellular mobile communication system, and in particular, to a channel estimation method for effectively reducing or eliminating interference in a cellular mobile communication system. Background of the invention

In modern cellular mobile communications, in order to eliminate transmission effects such as channel transmission characteristics and channel fading, and to more accurately receive user service data, channel estimation techniques are usually used. For example, the continuous pilot channel estimation proposed by American Qualcomm (Qualcomm) in the third generation mobile communication standard IS-2000, and the combination of the continuous pilot channel and the dedicated pilot channel proposed by Nokia and Errisson in Europe in WCDMA. Estimation (for the specific scheme of Qualcomm continuous pilot, please refer to the Physical Layer Standard for ANSI / TIA / EIA-95-B; for the joint pilot channel estimation scheme in WCDMA, please refer to 3GPP TS25.211). Since these channel estimation methods are estimates with interference, multi-user detection must be performed; otherwise, accurate channel estimation results will not be given, which will affect the performance of the communication system. However, multi-user detection has greatly increased the complexity of the system. Summary of the Invention

The present invention provides a channel estimation method applied to a cellular mobile communication system. In a multi-cell cellular environment, without the use of multi-user detection, the influence of interference can still be effectively eliminated, and an accurate channel estimation result is given. .

In the channel estimation method of the present invention, different pilot symbols are transmitted in the same subframe at the transmitting end. The interference between the different pilot symbols is correlated and the fading is also correlated. The receiving end performs the following steps: (a) performing despreading, in-phase demodulation, and serial-to-parallel conversion and despreading, orthogonal demodulation, and serial-to-parallel conversion on the received different pilot symbols;

(b) performing subtraction operations on the adjacent different pilot symbols after serial-to-parallel conversion to eliminate interference, and performing average operations on the subtraction calculation results to further eliminate interference;

(c) Divide the calculation result obtained in step b to obtain an estimated value of the channel phase offset; sum the squared result of the calculation result in step b to obtain an estimated value of the channel amplitude fading.

According to the above technical solution of the present invention, the different pilot symbols may be composed of K different pilot symbols as a group, and 2N / K groups of 2N pilot symbols are transmitted, where K and N are both positive integers.

When K is 2, and a total of 2N pilot symbols are transmitted, the subtraction operation of adjacent different pilot symbols after serial-parallel conversion described in step b refers to calculating the odd number of pilot symbols and the adjacent ones. Difference between even-bit pilot symbols.

When K is 2N and a group of 2N different pilot symbols is transmitted, the subtraction operation of the adjacent different pilot symbols after serial-parallel conversion described in step b refers to calculating the M-th pilot symbol and The difference between the M-th bit of the 1st pilot symbol, where M is a positive even number less than or equal to 2N.

According to the above technical solution of the present invention, the different pilot symbols may also be composed of K identical pilot symbols as a group, and a total of 2N pilot symbols of different 2N / K groups are transmitted, where, N Both are positive integers.

When K is 2, and different N sets of 2N pilot symbols are transmitted, the subtraction operation of adjacent different pilot symbols after serial-parallel conversion as described in step b refers to the calculation of odd-numbered adjacent pilots. The difference between symbols and the difference between adjacent pilot symbols of even bits.

When K is N, and different 2 groups of 2N pilot symbols are transmitted, the subtraction operation of adjacent different pilot symbols after serial-parallel conversion described in step ¾ 13 refers to calculating the M-th pilot The difference between the sign and the (N + M) -th pilot symbol, where M is less than or equal to A positive integer of N.

The above-mentioned different pilot symbols are best to take their correlation characteristic values to the minimum.

A channel estimation device applied to a cellular communication system provided by the method of the present invention includes a despread in-phase demodulation device, a despread quadrature demodulation device, a serial-to-parallel converter, a subtracter, an equalizer, a divider, and a squarer. And operator. The received signals are despread and demodulated by a despread in-phase demodulation device and a despread quadrature demodulation device; the despread and demodulated signals are each serial-to-parallel converted by a serial-to-parallel converter; and subtractors are used respectively. Perform difference operation on adjacent different pilot symbols after serial-to-parallel conversion to eliminate interference; use the equalizer to accumulate and average the difference operation results to further eliminate interference; and perform the output of the equalizer in a divider A division operation is performed to obtain an estimated phase offset of the channel; a square sum operation is performed on the output of the equalizer in a square sum operator to obtain an estimated value of the amplitude fading of the channel.

The channel estimation method of the present invention transmits different forms of pilot symbols with correlation between interference and fading at the transmitting end, and eliminates interference by performing operations on adjacent pilot forms with different types of pilot symbols at the receiving end to enable multi-cell Under conditions such as interference, without the use of multi-user detection, the influence of interference can still be effectively eliminated, and accurate channel parameter estimation information is given to reduce the complexity of the system. Brief description of the drawings

Figure 1 is a basic schematic block diagram of a wireless communication system.

Figure 2 is a schematic diagram of a frame structure.

FIG. 3 is a schematic diagram of an implementation example of four pilot symbol designs according to the present invention.

FIG. 4 is a schematic block diagram of a channel estimation device according to the method of the present invention. Mode of Carrying Out the Invention

The present invention is described in detail below through the expression of formulas in conjunction with the drawings. Referring to FIG. 1, at the k-th sampling time, the signal received by the receiving device 105 is: r _k = a _k s _k + I _k + n _k (1)

Where r _t , ^, / _A. , Respectively represent the signal received by the receiving device 105, the signal transmitted by the transmitting device 101, the fading factor of the fading channel 102, and the interference signal from the same cell and other cells at the k-th sampling time. 106, and the white noise signal 107.

Signal or channel parameters are estimated through pilot symbols, and the estimation results of the pilot symbols are used as a reference for subsequent data symbols, and these estimation results remain unchanged within a frame. At this time, the frame structure of the system is shown in FIG. 2.

The core of implementing the method of the present invention lies in the design of the pilot structure. The traditional pilot structure design all transmits the same pilot symbol, and the present invention transmits different pilot symbols in the same subframe. The so-called difference may refer to different positions on the signal constellation diagram, and the constellation diagram The larger the distance, the better. It is best when the correlation characteristic value between two different pilot symbols is the smallest. The interference between different pilot symbols in the same subframe is in the relevant area, and the fading is also in the relevant area.

Referring to FIG. 3, FIG. 3 shows structural designs of several pilot symbols of the present invention. The pilot symbols shown in (a) and (b) are composed of multiple sets of pilot symbols repeatedly, and the pilot symbols of each group are different from each other. (A) shows a group consisting of 2 different pilot symbols, P ₂ , and a total of N groups forming 2 N pilot symbols. (B) shown by a group of 2N mutually different pilot symbols _{_{_{Ρ 7, Ρ 2, Ρ 3}}} , Ρ 4 "- · 'Ρ2Ν-ι, Ρ 2Ν composition herein should satisfy the condition that the value of N different. Interference and fading between pilot symbols are correlated.

The pilot symbols shown in (c) and (d) are composed of multiple groups of pilot symbols, each group is composed of the same pilot symbol, and different pilot symbols are used between groups. (C) shows that two identical pilot symbols and ^, P ₂ and P ₂ , ······ P _Ν -ι and P _N- !, P _N and PN respectively, and share N groups constitute 2N pilot symbols, P ₂ , ·····

P _N is! ^ Mutually different pilot symbols. (D) shows a group consisting of N identical pilot symbols P ₇ and A respectively, a total of 2 groups constituting 2 N pilot symbols, and ₂ are two different pilots. Frequency symbol. Here, the value of N should also satisfy the interference and fading between different pilot symbols.

The following uses Figure 3- (a) as an example to explain the great convenience of channel estimation and the effective elimination of interference caused by the transformation of this pilot structure.

A total of 2N pilot symbols are set in a frame, and a signal is transmitted in the odd-numbered pilot symbols. A signal is transmitted in the even-numbered pilot symbols, that is, the k-th pilot symbol is:

{( ₂ )

among them:

Pi = cos (fi) t + ₀ ) -Q _k sinOt + ₀ ) = A _k cos (6¾ +. + ^) () P ₂ = I _k cos (ot + + sm (ot + φ ₀ ) = Α _{! (} cos (+ φ ₀ -φ ^ After fading the channel, the interference of fading factor and white noise is added. The form of the received signal is somewhat similar to the expression of formula (1). At this time:

, j '' k cos (6) t + φ ₀ + φ _ά1ι + φ _{Ι [} ) + Ι _Ιζ + n _k k = l, 3,-·, 2N-1 _(Λ , r _k = _k p _kd + I _k + n _k = < _t ,, (4)

_{[a k cos (iyt + φ} 0 + φ Λ - ^) + / / c + n k k = 2,4, '· ·, fading is introduced phase offset 2Ν formula.

Because the interval between adjacent pilot symbols is short, and the interference between the different pilot symbols is correlated, and the fading is also correlated, there is a _k = a _k between the adjacent pilot symbols. _ _x , p _dk = (p _d w, I _k = I _k _ _x (k = l, 2, ···, N). Therefore, for the 2k-l, 2k pilot symbols:

r _2k =

+ ⁿ 2k ( ⁶ ) k = l, 2, ... ', N

Referring to FIG. 4, at this time, the signal received by the receiving end is subjected to in-phase demodulation through the despreading and demodulation device 401 and output through the serial-to-parallel converter 403; Alternate demodulation, through the output of the serial-to-parallel converter 404, the following signals are obtained:

(7) d (2k-l) o =

+ (24-1) + ⁿ ₀ (2k-l) (8) diks = a _2k A _2k COSOQ + φ _21ί一_φά2Ι ) + I _s2k + n _s2k (9) dike = ^ 2k ^A 2k sin (^ ₀ + u― <Pd2 _k ) + I. u + n _olk (10) k = l, 2, ...

Arithmetic unit 405 is used to complete (7)-(9).

C _h = -2α, Λ sin + φ _& ) sin () + n, _sk k = l, 3, 5 ··, 2N- 1 (11) The operator 406 is used to complete (8)-(10), Have:

C _ks = 2a _k A _k cos (^ ₀ + ^) sin (^) + n, _ok k = l, 3, 5 ··, 2N-1 (12) where:

k = l, 3,5, 2N-1 (13) k = ⁿ ok- ^{κ =} 1 '3, 5, ……, 2N-1 (14) n _k , n _k are all Gaussian white noise with zero mean.

It can be seen from the expressions of (11) and (12) that the noise has been basically eliminated. In order to accurately estimate the energy of the signal, the influence of white noise has to be removed. To do this, we need to average the formulas (11) and (12).

The equalizers 407 and 408 average the expressions (11) and (12), respectively. Since there are 2N pilot symbols in total, through the above derivation process, the subtraction operation is performed between adjacent pilot symbols.

There are only N terms left in the formulas (11) and (12), so the N terms can be averaged for the formulas (11) and (12): f _sk (15)

c _ko = — ∑ ² ^ cos (^ ₀ + φ _{ά (ί} ) sin (^,) + —∑ n _k (16) After the average operation of (15), (16), noise is largely eliminated Impact, And because the reception interference of P ₂ has the same correlation with the fading through the transmission channel, it has:

c _fa = 2 _k A _k sin + ^) sin (^) + n (17) c _ko = 2 _k A _k cosO ₀ + ¾) sin (^) + n ' ₀ (18) ^ at this time, after averaging It has become very small since then.

Calculate (17) * (17) + (18) * (18) by the square sum operator 409, and get:

Since the pilot symbol is a known signal, 4, sitin are both known signals, and the value of n "is very small and can be ignored here. Therefore, dividing formula (19) by the constant term 4 sin ² (^), The estimated value c _{k of the} signal amplitude fading factor can be obtained.

The divider 410 is used to perform the operations of formulas (17) / (18), and obtains:

tg9 _sp = tg + (20) Similarly, because the pilot symbol is a known signal, so for known information, the value of n "is small and negligible, so we can get an estimate of the phase offset value ^.

Since the influence of interference and noise has been largely eliminated when estimating the channel parameters ^, + ^, the accuracy of the estimation is much higher than the estimated value in the presence of interference and noise in the prior art. An effective channel estimation method. In the present invention, different pilot symbols with correlation between interference and fading are transmitted in the same subframe at the transmitting end, and the receiving end performs operations on the received different pilot symbols to eliminate interference to obtain an accurate value of channel estimation . By using the channel estimation method provided by the present invention, accurate channel estimation values can be given in the case of multi-cell multi-user interference, and complex multi-user detection is avoided, and the complexity of the system is reduced. Effective channel estimation method for mobile communication or interference communication system. The above is only a preferred embodiment of the present invention. The description is not intended to limit the present invention, and any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the scope of the claims of the present invention.

Claims

Claim

1. A channel estimation method applied to a cellular mobile communication system, characterized in that the method includes:

Different pilot symbols are transmitted in the same subframe at the transmitting end, and interference between the different pilot symbols is correlated, and fading is also correlated. The following steps are performed at the receiving end:

(a) Perform despreading, in-phase demodulation and serial-to-parallel conversion and despreading, orthogonal demodulation and serial-to-parallel conversion on the received different pilot symbols; remove interference, and average the calculated results after subtraction. To further eliminate interference;

2. The channel estimation method according to claim 1, wherein: the different pilot symbols are composed of K different pilot symbols as a group, and transmitted.

2N pilot symbols in the 2N / K group, where K and NR are positive integers.

3. The channel estimation method according to claim 2, wherein: K is 2, and 2 N pilot symbols are transmitted in the N group;

The subtraction operation of adjacent different pilot symbols after serial-parallel conversion refers to calculating a difference between pilot symbols of odd bits and pilot symbols of adjacent even bits.

4. The channel estimation method according to claim 2, wherein: κ is 2N, and a group of 2N different pilot symbols is transmitted;

The subtraction operation of adjacent different pilot symbols after serial-parallel conversion refers to calculating the difference between the Mth pilot symbol and the M-1th pilot symbol, where M is less than or equal to 2N Positive even number.

5. The channel estimation method according to claim 1, wherein: The different pilot symbols refer to a group consisting of K identical pilot symbols and transmitting a total of 2N pilot symbols of different 2N / K groups, where K and N are positive integers.

6. The channel estimation method according to claim 5, wherein: κ is 2, and a total of 2N pilot symbols are transmitted in different N groups;

The subtraction operation on serially-parallel converted adjacent different pilot symbols refers to calculating a difference between odd-numbered adjacent pilot symbols and an even-numbered adjacent pilot symbol.

7. The channel estimation method according to claim 5, wherein: κ is N, and different 2 groups of 2N pilot symbols are transmitted;

The performing a subtraction operation on adjacent different pilot symbols after serial-parallel conversion refers to calculating a difference between a pilot symbol of the 笫 M bit and a pilot symbol of the (N + M) th bit, where M is A positive integer less than or equal to N.

8. The channel estimation method according to any one of claims 1 to 7, characterized in that: the optimal setting of the different pilot symbols is such that the correlation characteristic value thereof is the smallest.

9. A channel estimation device applied to a cellular communication system, characterized in that the device includes a despread in-phase demodulation device, a despread quadrature demodulation device, a serial-to-parallel converter, a subtracter, an equalizer, and a divider. And square sum operator;

The received signals are despread and demodulated by a despread in-phase demodulation device and a despread quadrature demodulation device; the despread and demodulated signals are each serial-to-parallel converted by a serial-to-parallel converter; and subtractors are used respectively. Perform difference operation on adjacent different pilot symbols after serial-to-parallel conversion to eliminate interference; use the equalizer to accumulate and average the difference operation results to further eliminate interference; and perform the output of the equalizer in a divider A division operation is performed to obtain an estimated phase offset of the channel; a square sum operation is performed on the output of the equalizer in a square sum operator to obtain an estimated value of the amplitude fading of the channel.