CN101312359A - Apparatus and method for multi-cell combined channel estimation and multi-cell combined detection - Google Patents

Abstract

The invention provides a multi-cell joint channel estimation algorithm and a device thereof, a multi-cell joint detection method and a device thereof, which mainly comprises steps of estimating channel impulse response of each cell and an adjacent cell, selecting N strong signal paths from relative multi-path signals of all the adjacent cell users according to the channel estimation results of each adjacent cell, selecting M strong signal paths from relative multi-path signals of users of the cell according to the channel estimation results of the cell, utilizing the selected N+M strong signal paths to perform multi-cell joint channel estimation and finally obtaining multi-cell joint CIR column vector. And further multi-cell joint detection can be performed by utilizing the above multi-cell joint channel estimation results. Because the multi-path signals selected for multi-cell joint channel estimation are strong signal paths with strong power, the multi-cell joint channel estimation and the multi-cell joint detection can obtain fine CIR estimated performance and large joint detection gain.

Description

Multi-plot joint channel estimating, multi-plot joint checkout gear and method

Technical field

The present invention relates to wireless communication technology, particularly the multi-plot joint channel estimating apparatus in the wireless communication system, multiple cell combined channel estimation method and multi-plot joint checkout gear and multi-cell joint detection method.

Background technology

At present, in low spreading rate of third generation partner program (3GPP) time division duplex and high spreading rate (3GPP TDD LCR and HCR) system, use single community associated detection technique to improve the performance that system up-link signal detects usually.Yet the associated detection technique of single sub-district only can be used to eliminate interference in the sub-district, and can not be used to eliminate presence of intercell interference.The multi-plot joint detection technique is exactly the uplink signal detection technology in order to eliminate interference and presence of intercell interference in the sub-district simultaneously and to propose.Existing multi-plot joint detects to be needed at first to select in neighbor cell some this cell signal to be disturbed stronger interference user, and then utilizing all multipath signals of user in all multipath signals of interference user in the selected neighbor cell and this sub-district to carry out uplink signal detection, signal detection process mainly comprises pilot tone (midamble) territory multi-plot joint channel estimating and two processes of data field multi-plot joint detection.

In the application of reality, still there is following shortcoming in existing multi-plot joint detection technique:

At first, if all multipath signals of user have taken all skews (shift) of midamble sign indicating number in this sub-district, then in the multi-plot joint channel estimation process of midamble territory, the multipath signal of interference user in the neighbor cell can not have been introduced again, in this case, midamble territory multi-plot joint channel estimating will deteriorate to single cell channel to be estimated, thereby the forfeiture multi-plot joint detects the detection gain that is brought;

Secondly, known channel impulse response (CIR each user, Channel Impulse Response) in the window, no matter be user or the interior user of neighbor cell in this sub-district, be not that its all multipath signals all are strong signal footpaths, and because the gain that the weak signal footpath is brought in midamble territory multi-plot joint channel estimating and the detection of data field multi-plot joint is little more a lot of than strong signal footpath, and computational resource that weak signal directly consumes and strong signal footpath are quite, thereby all multipath signals of selected interference user carry out multi-plot joint detection and unreasonable in all multipath signals of this community user of use and the neighbor cell in midamble territory multi-plot joint channel estimating and data field multi-plot joint testing process;

At last, because weak jamming user's strong signal footpath probably directly has higher power than the weak signal of user in interference user in the neighbor cell or this sub-district in the neighbor cell, therefore, if in midamble territory multi-plot joint channel estimating and the detection of data field multi-plot joint, utilize the strong signal footpath of weak jamming user in the neighbor cell will obtain bigger gain, but existing multi-cell joint detection method is not considered the multipath signal of the weak jamming user in the neighbor cell.

Summary of the invention

In order to solve the problems of the prior art, the invention provides multi-plot joint channel estimating apparatus and method to obtain better channel estimating performance.

The present invention also provides multi-plot joint checkout gear and method, to obtain bigger detection gain.

Multiple cell combined channel estimation method of the present invention comprises:

Estimate the channel impulse response CIR of each neighbor cell and this sub-district;

CIR according to each neighbor cell selects N strong signal footpath from the multipath signal of all neighbor cell user correspondences, according to the CIR of this sub-district, select M strong signal footpath from the multipath signal of this community user correspondence, and wherein, N and M are natural number;

Utilize selected N+M strong signal directly to carry out the multi-plot joint channel estimating, obtain multi-plot joint channel impulse response CIR column vector.

Wherein, according to a preferred version of the present invention, from the multipath signal of all neighbor cell user correspondences, select the step in N strong signal footpath to comprise: to calculate the power of each multipath signal in all neighbor cells, therefrom select N the multipath signal that power is the strongest as described N strong signal footpath.

According to another preferred version of the present invention, from the multipath signal of all neighbor cell user correspondences, select the step in N strong signal footpath to comprise: to calculate the power of each CIR, and calculate power average value and the variance of all CIR; Mean value and variance according to all CIR power that calculate are determined a threshold value, select the CIR of power greater than described thresholding from all CIR of described neighbor cell; From the stronger CIR of selected power, select N the multipath signal that power is the strongest as described N strong signal footpath.

According to another preferred version of the present invention, from the multipath signal of all neighbor cell user correspondences, select the step in N strong signal footpath to comprise: to determine the number in strong signal footpath among the CIR, and determine the number N of strong jamming CIR to be selected according to the number in strong signal footpath among each CIR that determines _ShiftAccording to the power of each CIR, from all CIR of neighbor cell, select the strongest N of power _ShiftIndividual CIR; From selected N _ShiftSelect N the multipath signal that power is the strongest as described N strong signal footpath in pairing all multipath signals of individual CIR, wherein, N _ShiftBe natural number.

In addition, from the multipath signal of this community user correspondence, select the step in M strong signal footpath to comprise: to calculate the power of each multipath signal in this sub-district, therefrom select M the multipath signal that power is the strongest as described M strong signal footpath.

M multipath signal that power is the strongest of described selection comprises as the step in described M strong signal footpath:

Order

Wherein, K _{M, C+1}Midamble skew for this sub-district;

If N _{Local, path}＞W _C+1, then make N _{Local, path}=W _C+1, W wherein _C+1For the CIR window of this sub-district long;

If N _{Local, shift}=M-K _{M, C+1}N _{Local, path}＞0 and N _{Local, path}＜W _C+1, then at k _m=0 ..., N _{Local, shift}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}}+1,$ And at k _m=N _{Local, shift}..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$ Otherwise, at k _m=0 ..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$

According to the power of all multipath signals that calculate, from being offset all multipath signals of pairing CIR, each midamble of this sub-district selects

The strongest multipath signal of individual power;

Utilize formula $M=\underset{k_m=0}{\overset{K_{m,C+1}-1}{\mathrm{\Σ}}}{N}_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}$ Upgrade M.

The step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises: generate the required multi-plot joint channel estimate matrix of multi-plot joint channel estimating, and utilize the multi-plot joint channel estimate matrix that generates to carry out midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector.

According to another preferred version of the present invention, the step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises: generate the required multi-plot joint channel estimate matrix of multi-plot joint channel estimating, therefrom select L ' row vector to regenerate the multi-plot joint channel estimate matrix of simplification, utilize the multi-plot joint channel estimate matrix of described simplification to carry out midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector, wherein, N+M≤L '≤L _Mid-W _{Max, c}+ 1, L _MidThe length of representing each midamble piece, W _{Max, c}The maximal possible length of expression channel impulse response window.

Multi-plot joint channel estimating apparatus of the present invention comprises: the strong signal of neighbor cell directly selects module, the strong signal in this sub-district directly to select module, multi-plot joint channel estimation module and at least one channel estimation module; Wherein,

Each channel estimation module is respectively applied for the channel impulse response CIR that estimates designated cell in neighbor cell and this sub-district, obtains the CIR of each neighbor cell and this sub-district;

The strong signal of described neighbor cell directly selects module to be used for selecting N strong signals footpaths according to the CIR of the channel estimation module of corresponding neighbor cell output all multipath signals of all users in the neighbor cell;

The CIR that the strong signal in described this sub-district directly selects module to be used for exporting according to the channel estimation module of corresponding this sub-district selects M strong signal footpaths from all multipath signals of all users of this sub-district;

N+M the strong signal that described multi-plot joint channel estimation module is used for directly selecting module and the strong signal in this sub-district directly to select module to select according to the strong signal of neighbor cell directly carries out the multi-plot joint channel estimating.

Multi-cell joint detection method of the present invention comprises:

Estimate the channel impulse response CIR of each neighbor cell and this sub-district;

CIR according to each neighbor cell selects N strong signal footpath from the multipath signal of all neighbor cell user correspondences, according to the CIR of this sub-district, select M strong signal footpath from the multipath signal of this community user correspondence, and wherein, N and M are natural number;

Utilize selected N+M strong signal directly to carry out the multi-plot joint channel estimating, obtain multi-plot joint CIR column vector;

Utilize the matrix of generation system as a result of multi-plot joint channel estimating;

The first and second data block signals according to sytem matrix that is generated and reception carry out the detection of data field multi-plot joint, obtain the first and second data block data carried by data.

Wherein, according to a preferred version of the present invention, from the multipath signal of all neighbor cell user correspondences, select the step in N strong signal footpath to comprise: to calculate the power of each multipath signal in all neighbor cells, therefrom select N the multipath signal that power is the strongest as described N strong signal footpath.

According to another preferred version of the present invention, from the multipath signal of all neighbor cell user correspondences, select the step in N strong signal footpath to comprise: to calculate the power of each CIR, and calculate power average value and the variance of all CIR; Mean value and variance according to all CIR power that calculate are determined a threshold value, select the CIR of power greater than described thresholding from all CIR of described neighbor cell; From the stronger CIR of selected power, select N the multipath signal that power is the strongest as described N strong signal footpath.

According to another preferred version of the present invention, from the multipath signal of all neighbor cell user correspondences, select the step in N strong signal footpath to comprise: to determine the number in strong signal footpath among the CIR, and determine the number N of strong jamming CIR to be selected according to the number in strong signal footpath among each CIR that determines _ShiftAccording to the power of each CIR, from all CIR of neighbor cell, select the strongest N of power _ShiftIndividual CIR; From selected N _ShiftSelect N the multipath signal that power is the strongest as described N strong signal footpath in pairing all multipath signals of individual CIR, wherein, N _ShiftBe natural number.

In addition, from the multipath signal of this community user correspondence, select the step in M strong signal footpath to comprise: to calculate the power of each multipath signal in this sub-district, therefrom select M the multipath signal that power is the strongest as described M strong signal footpath.

M multipath signal that power is the strongest of described selection comprises as the step in described M strong signal footpath:

Order Wherein, K _{M, C+1}Midamble skew for this sub-district;

If N _{Local, path}＞W _C+1, then make N _{Local, path}=W _C+1, W wherein _C+1For the CIR window of this sub-district long;

If N _{Local, shift}=M-K _{M, C+1}N _{Local, path}＞0 and N _{Local, path}＜W _C+1, then at k _m=0 ..., N _{Local, shift}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}}+1,$ And at k _m=N _{Local, shift}..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$ Otherwise, at k _m=0 ..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$

According to the power of all multipath signals that calculate, from being offset all multipath signals of pairing CIR, each midamble of this sub-district selects

The strongest multipath signal of individual power;

Utilize formula $M=\underset{k_m=0}{\overset{K_{m,C+1}-1}{\mathrm{\Σ}}}{N}_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}$ Upgrade M.

The step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises: calculate the required multi-plot joint channel estimate matrix parameter of multi-plot joint channel estimating, and utilize the multi-plot joint channel estimate matrix that generates to carry out pilot tone midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector.

According to another preferred version of the present invention, the described step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises: generate the required multi-plot joint channel estimate matrix of multi-plot joint channel estimating, therefrom select L ' row vector to regenerate the multi-plot joint channel estimate matrix of simplification, utilize the multi-plot joint channel estimate matrix of described simplification to carry out midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector, wherein, N+M≤L '≤L _Mid-W _{Max, c}+ 1, L _MidThe length of representing each midamble piece, W _{Max, c}The maximal possible length of expression channel impulse response window.

Utilize the step of the matrix of generation system as a result of multi-plot joint channel estimating to comprise: to upgrade corresponding community associated CIR according to the multi-plot joint channel estimation results; Calculate the specific CIR of user equipment (UE) according to the community associated CIR after upgrading, and carry out the channel estimating reprocessing; Calculate the CIR of particular virtual Resource Unit VRU according to the CIR of the corresponding particular UE after the channel estimating reprocessing; Determine the quantity of the VRU that the participation multi-plot joint detects, and select the VRU of respective numbers; Calculate the associating CIR of selected VRU in corresponding each local VRU and the special neighbourhood sub-district, and according to the associating CIR generation system matrix of the corresponding specific VRU that calculates.

After the generation system matrix, further comprise: from the sytem matrix that is generated, select L _AIndividual row vector generates new matrix, wherein, and L _AMore than or equal to the columns of described sytem matrix and be less than or equal to the line number of described sytem matrix, with newly-generated matrix as sytem matrix.

Before the first and second data block signals according to sytem matrix that is generated and reception carry out the detection of data field multi-plot joint, further comprise: utilize the result of multi-plot joint channel estimating that the first and second data block signals that receive are handled, elimination in transmission course the midamble block signal to the interference of the described first and second data block signals, obtain the first and second clean data block signals, and utilize the first and second clean data block signals to carry out the data field multi-plot joint and detect.

Elimination midamble block signal in transmission course comprises the interference of the described first and second data block signals: estimate the interference of this sub-district midamble sign indicating number to two data blocks, deduct the interference of this sub-district midamble sign indicating number to two data blocks respectively from the baseband signal of first and second data blocks that receive; Estimate of the interference of the midamble sign indicating number of neighbor cell, from the baseband signal of having eliminated first and second data blocks that this sub-district midamble sign indicating number disturbs, deduct of the interference of the midamble sign indicating number of neighbor cell respectively two data blocks to two data blocks.

Multi-plot joint checkout gear of the present invention comprises: pilot tone midamble territory multi-plot joint channel estimating unit and data field multi-plot joint detecting unit; Wherein,

Described midamble territory multi-plot joint channel estimating unit is used to estimate the channel impulse response CIR of each neighbor cell and this sub-district, CIR according to each neighbor cell, in neighbor cell, select N strong signal footpath in all multipath signals of each user, CIR according to this sub-district, from all multipath signals of each user of this sub-district, select M strong signal footpath, utilize selected N+M strong signal directly to carry out midamble territory multi-plot joint channel estimating again, wherein, N and M are natural number;

Described data field multi-plot joint detecting unit is used to utilize the midamble territory multi-plot joint channel estimation results generation system matrix of midamble territory multi-plot joint channel estimating unit output, and carry out the data field multi-plot joint according to the first and second data block signals of sytem matrix that is generated and reception and detect, obtain the first and second data block data carried by data.

Wherein, midamble territory multi-plot joint channel estimating unit comprises: the strong signal of neighbor cell directly selects module, the strong signal in this sub-district directly to select module, multi-plot joint channel estimation module and at least one channel estimation module; Wherein,

Each channel estimation module is respectively applied for the channel impulse response CIR that estimates designated cell in neighbor cell and this sub-district, obtains the CIR of each neighbor cell and this sub-district;

The strong signal of described neighbor cell directly selects module to be used for selecting N strong signals footpaths according to the CIR of the channel estimation module of corresponding neighbor cell output all multipath signals of all users in the neighbor cell;

The CIR that the strong signal in described this sub-district directly selects module to be used for exporting according to the channel estimation module of corresponding this sub-district selects M strong signal footpaths from all multipath signals of all users of this sub-district;

N+M the strong signal that described multi-plot joint channel estimation module is used for directly selecting module and the strong signal in this sub-district directly to select module to select according to the strong signal of neighbor cell directly carries out the multi-plot joint channel estimating.

Data field multi-plot joint detecting unit comprises: multi-plot joint sytem matrix generation module and least mean-square error MMSE detection module; Wherein,

Described multi-plot joint sytem matrix generation module is used for generating the required sytem matrix of data field multi-plot joint detection according to the multi-plot joint channel estimation results of multi-plot joint channel estimation module output;

Described MMSE detection module is used for the sytem matrix that generates according to multi-plot joint sytem matrix generation module and receives the first and second data block signals carrying out the multi-plot joint detection, obtains the first and second data block data.

Data field multi-plot joint detecting unit further comprises: the midamble interference cancellation module, be used to eliminate midamble sign indicating number that this sub-district and neighbor cell use this sub-district is received the interference of two data blocks in the same burst packets, generate the receiving baseband signal of the first and second clean data blocks and output to described MMSE detection module.

This shows, the described multi-plot joint channel estimation process of the embodiment of the invention is at first selected the stronger strong signal footpath of signal in all multipath signals of each user in neighbor cell, also in this sub-district, select the stronger strong signal footpath of signal simultaneously in all multipath signals of each user, and then utilize the strong signal of user in selected neighbor cell and this sub-district directly to carry out midamble territory multi-plot joint channel estimating, can obtain good CIR estimated performance thus.

In addition, detect, thereby also can obtain big joint-detection gain because the described data field multi-plot joint of embodiment of the invention detecting unit utilizes the output of midamble territory multi-plot joint channel estimating unit to carry out the data field multi-plot joint.

Description of drawings

To make clearer above-mentioned and other feature and advantage of the present invention of those of ordinary skill in the art by describing the preferred embodiments of the present invention in detail below with reference to accompanying drawing, in the accompanying drawing:

Fig. 1 has shown the described multi-plot joint checkout gear of embodiment of the invention internal structure;

Fig. 2 selects method 1 flow chart in strong signal footpath for the strong signal selection module of the described neighbor cell of the embodiment of the invention;

Fig. 3 selects method 2 flow charts in strong signal footpath for the strong signal selection module of the described neighbor cell of the embodiment of the invention;

Fig. 4 selects method 3 flow charts in strong signal footpath for the strong signal selection module of the described neighbor cell of the embodiment of the invention;

Fig. 5 selects the method flow diagram in strong signal footpath for the strong signal selection module in described this sub-district of the embodiment of the invention;

Fig. 6 is the method flow diagram of the described multi-plot joint channel estimating of the embodiment of the invention;

Fig. 7 is the method flow diagram of the described generation system matrix of the embodiment of the invention;

Fig. 8 is described multi-plot joint channel estimating of the embodiment of the invention and multi-cell joint detection method flow chart.

Embodiment

In order to solve the problems of the technologies described above, to the invention provides and realize the device that multi-plot joint channel estimating and multi-plot joint detect and realize multi-plot joint channel estimating and multi-cell joint detection method.

The device and method that multi-plot joint channel estimating of the present invention and multi-plot joint detect is at first selected the stronger strong signal footpath of signal in all multipath signals of each user in neighbor cell, also in this sub-district, select the stronger strong signal footpath of signal simultaneously in all multipath signals of each user, and then utilize the strong signal of user in selected neighbor cell and this sub-district directly to carry out midamble territory multi-plot joint channel estimating.Because the multi-plot joint channel estimation process utilization of above-mentioned midamble territory is that the strong signal in all multipath signals of user is directly in user and this sub-district in the neighbor cell, therefore, compare with existing multi-plot joint detection technique, midamble provided by the invention territory multi-plot joint channel estimation process can obtain better CIR estimated performance.And, because it is to carry out on the basis of midamble territory multi-plot joint channel estimating that data field multi-plot joint of the present invention detects, the output of midamble territory multi-plot joint channel estimating will be as the input of data field multi-plot joint detection, thereby, along with the raising of CIR estimated performance, the data field multi-plot joint detects also can obtain better joint-detection gain.

Below in conjunction with accompanying drawing and embodiments of the invention technical scheme of the present invention is described in further detail.

Fig. 1 is the described multi-plot joint checkout gear of an embodiment of the invention internal structure schematic diagram.As shown in Figure 1, the described multi-plot joint checkout gear of present embodiment mainly comprises: midamble territory multi-plot joint channel estimating unit and two parts of data field multi-plot joint detecting unit.

Wherein, described midamble territory multi-plot joint channel estimating unit is used for the strong signal of all multipath signals selections footpath of each user in neighbor cell, from all multipath signals of each user of this sub-district, select strong signal footpath simultaneously, and then utilize selected strong signal directly to carry out midamble territory multi-plot joint channel estimating;

Described data field multi-plot joint detecting unit is used for carrying out the detection of data field multi-plot joint according to the midamble territory multi-plot joint channel estimation results of midamble territory multi-plot joint channel estimating unit output.

As shown in Figure 1, described midamble territory multi-plot joint channel estimating unit comprises: at least one channel estimation module (CHE), the strong signal of neighbor cell directly select module (SPC-AC), the strong signal in this sub-district directly to select module (SPC-LC), multi-plot joint channel estimation module (MC-JCHE).Particularly, described channel estimation module is used to estimate the CIR of designated cell; The strong signal of described neighbor cell directly selects module to be used for selecting to determine according to the channel estimation results of corresponding neighbor cell channel estimation module all multipath signals of all users in the neighbor cell the stronger strong signal footpath of signal of number; The strong signal in described this sub-district directly selects module to be used for selecting to determine from all multipath signals of all users of this sub-district according to the channel estimation results of the channel estimation module of this sub-district the stronger strong signal footpath of signal of number; Described multi-plot joint channel estimation module is used for directly carrying out the multi-plot joint channel estimating according to the strong signal that the strong signal of neighbor cell directly selects module and the strong signal in this sub-district directly to select module to select.

Above-mentioned channel estimation module, the strong footpath of neighbor cell select module, strong footpath, this sub-district to select module and multi-plot joint channel estimation module to realize midamble territory multi-plot joint channel estimation function jointly.Those skilled in the art will appreciate that above-mentioned midamble territory multi-plot joint channel estimating unit can also use separately as discrete multi-plot joint channel estimating apparatus.

As shown in Figure 1, described data field multi-plot joint detecting unit comprises: multi-plot joint sytem matrix generation module, midamble interference cancellation module (MIC) and least mean-square error (MMSE) detection module.Particularly, described multi-plot joint sytem matrix generation module is used for generating the required sytem matrix of data field multi-plot joint detection according to the multi-plot joint channel estimation results of multi-plot joint channel estimation module output; Described midamble interference cancellation module is used to eliminate midamble sign indicating number that this sub-district and neighbor cell use this sub-district is received the interference of two data blocks in the same burst packets, generates the receiving baseband signal of clean (clean) first and second data blocks; The MMSE detection module is used for carrying out multi-plot joint according to first and second data blocks of the sytem matrix of multi-plot joint sytem matrix generation module generation and the output of midamble interference cancellation module and detects, and obtains the first and second data block data.

Above-mentioned multi-plot joint sytem matrix generation module, midamble interference cancellation module and MMSE detection module have been realized data field multi-plot joint measuring ability jointly.

Further describe the processing procedure of above-mentioned each module below in conjunction with accompanying drawing.

1. channel estimation module

As previously mentioned, described channel estimation module is mainly used in designated cell is carried out channel estimating, determines the CIR of this sub-district correspondence.The number of employed channel estimation module will be carried out the number of the sub-district of channel estimating, the i.e. sum of this sub-district and neighbor cell corresponding to needs in the described multi-plot joint checkout gear of present embodiment.As shown in Figure 1, total C+1 of the described channel estimation module of present embodiment, wherein, C channel estimation module corresponds respectively to C neighbor cell, be respectively applied for the channel estimating of this C neighbor cell, the another one channel estimation module will be used for the channel estimating of this sub-district corresponding to this sub-district.In order to express easily represent c neighbor cell with sub-district c, wherein, c=1,2 ..., C; And represent this sub-district with sub-district C+1.

To be example with the channel estimation module of sub-district c correspondence below, describe the operating process of each channel estimation module in detail.

In order to express easily, suppose antenna k by receiving terminal in the c of sub-district _aThe chip-level base band sequence table that receives is shown

, wherein, k _a=0 ..., K _a-1, K _aThe total number of expression receiving terminal antenna; I=0,1 ..., L _Bu-1, L _BuThe length of a burst packets of expression, for example, L in time division duplex-S-CDMA (TD-SCDMA) system _Bu=848; R represents " reception ", represents that this signal is a received signal.Generally, each burst packets comprises two data blocks and the midamble piece between two data blocks.Described channel estimation module utilizes the midamble piece in the received signal to carry out channel estimating.The length of supposing each data block is L _Data, L in the TD-SCDMA system _Data=352; In addition, the length of supposing each midamble piece is L _Mid, L in the TD-SCDMA system _Mid=144.

The channel estimation module chip-level base band sequence of known input respective cell c can be expressed as ${\underset{\‾}{e}}_{m,c,i}^{\left(k_a\right)}={\underset{\‾}{e}}_{r,i+L_{\mathrm{data}}+W_{\max ,c}}^{\left(k_a\right)},$ i＝0，...，L-1。Wherein, definition W _{Max, c}∈ N is the maximal possible length of channel impulse response window, and makes L=L _Mid-W _{Max, c}+ 1.Further can with

Write as column vector

Form.

In addition, associating (combined) CIR that supposes sub-district c is

Can obtain ${\underset{\‾}{h^{\′\′}}}_c^{\left(k_a\right)}={[{\underset{\‾}{h^{\′\′}}}_c^{(k_a,0)T}{\underset{\‾}{h^{\′\′}}}_c^{(k_a,1)T}...{\underset{\‾}{h^{\′\′}}}_c^{(k_a,k_m)T}...{\underset{\‾}{h^{\′\′}}}_c^{(k_a,K_{m,c}-1)T}]}^T,$ k _a=0 ..., K _a-1, wherein,

For midamble skew among the c of sub-district is k _mThe time antenna k _aCIR estimate,

In T representative

Transposition, and K _{M, c}Midamble skew among the expression sub-district c.

Can further be expressed as ${\underset{\‾}{h^{\′\′}}}_c^{(k_a,k_m)}={[{\underset{\‾}{h^{\′\′}}}_{c,1}^{(k_a,k_m)}{,\underset{\‾}{h^{\′\′}}}_{c,2}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{c,i}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{c,W_c}^{(k_a,k_m)}]}^T,$ Wherein, Representative

The fading coefficients in i footpath, W _cFor the CIR window of sub-district c long.

For the channel estimation module of respective cell c, the chip-level base band sequence of input

Can also be expressed as ${\underset{\‾}{e}}_{m,c}^{\left(k_a\right)}={\underset{\‾}{G}}_c{\underset{\‾}{h^{\′\′}}}_c^{\left(k_a\right)}+{n_c}^{\left(k_a\right)},$ c＝1，2，...，C+1。Wherein, G _cBe the channel estimate matrix of sub-district c,

Be antenna k among the c of sub-district _aThe noise column vector.

From foregoing description as can be seen, the major function of the channel estimation module of respective cell c is the chip-level base band sequence according to input

Calculate

Preferably, described channel estimation module can use ZF (ZF, zero-forcing) detector or MMSE detector calculating

If channel estimation module uses the ZF detector, and antenna k among the hypothesis sub-district c _aThe noise column vector

Covariance be σ _c ²I, wherein, I is unit matrix (unitary matrix), σ _c ²For

Noise power.Described channel estimation module can calculate the associating CIR of respective cell by following formula (1).

${\underset{\‾}{h^{\′\′}}}_c^{\left(k_a\right)}={\left({{\underset{\‾}{G}}_c}^H{\underset{\‾}{G}}_c\right)}^{-1}{{\underset{\‾}{G}}_c}^H{\underset{\‾}{e}}_{m,c}^{\left(k_a\right)},k_a=0,...,K_a-1---\left(1\right)$

If channel estimation module uses the MMSE detector, and antenna k among the hypothesis sub-district c _aThe noise column vector

Covariance be σ _c ²I, described channel estimation module can calculate the associating CIR of respective cell by following formula (2).

${\underset{\‾}{h^{\′\′}}}_c^{\left(k_a\right)}={({{\underset{\‾}{G}}_c}^H{\underset{\‾}{G}}_c+{\mathrm{\σ}}_c^{2}I)}^{-1}{{\underset{\‾}{G}}_c}^H{\underset{\‾}{e}}_{m,c}^{\left(k_a\right)},k_a=0,...,K_a-1---\left(2\right)$

If G _cFor Theory of Circular Matrix (is that the element that comprises of each row in the matrix is identical, and each classifies the vector that the next column element moves up and obtains behind the element as), at this moment, the channel estimation module of respective cell c can calculate the associating CIR of sub-district c by following formula (3).

${\underset{\‾}{h^{\′\′}}}_c^{\left(k_a\right)}=\mathrm{IDFT}({\left({\underset{\‾}{G}}_{\mathrm{DFT}.c}\right)}^{-1}\·\mathrm{DFT}\left({\underset{\‾}{e}}_{m,c}^{\left(k_a\right)}\right)),k_a=0,...,K_a-1---\left(3\right)$

Wherein, operator DFT () represents discrete Fourier transform (DFT), the contrary discrete Fourier transform (DFT) of IDFT () expression, ( G _DFT.c) ^-1Expression G _DFT.cContrary. G _DFT.cFor the diagonal angle vector is $\underset{\‾}{g}=\mathrm{DFT}\left({\underset{\‾}{m}}_p^c\right).$ Diagonal matrix, ${\underset{\‾}{m}}_P^c=({\underset{\‾}{m}}_1^c,{\underset{\‾}{m}}_2^c,...,{\underset{\‾}{m}}_P^c)$ The basic midamble sign indicating number of expression sub-district c.

By above-mentioned formula as can be seen, at the channel estimate matrix of known cell c G _cThe time can according to the input chip-level base band sequence

Calculate the associating CIR of sub-district c.

To provide the channel estimate matrix that obtains sub-district c below G _cMethod.

Suppose to use ${\underset{\‾}{m}}_P^c=({\underset{\‾}{m}}_1^c,{\underset{\‾}{m}}_2^c,...,{\underset{\‾}{m}}_P^c)$ The basic midamble sign indicating number of expression sub-district c (each component of this sign indicating number is a plural number), wherein, P is the length of described basic midamble sign indicating number.At i=(P+1) ..., L _Max, L _Max=L _Mid+ (K _{M, c}-1) W _cThe time, order ${\underset{\‾}{m}}_i^c={\underset{\‾}{m}}_{i-P}^c,$ Can basis m _p ^cGenerate ${\underset{\‾}{m}}^c=({\underset{\‾}{m}}_1^c,{\underset{\‾}{m}}_2^c,...,{\underset{\‾}{m}}_P^c,...,{\underset{\‾}{m}}_{L_{\max }}^c).$

Further suppose

And

Consider at i=(P+1) ..., L _MaxThe time, order ${\underset{\‾}{m}}_i^c={\underset{\‾}{m}}_{i-P}^c,$

Can be again by m _i ^c, i=1,2 ..., P is expressed as:

Can obtain G thus _{Temp, c}, with G _{Temp, c}Preceding W _{Max, c}After-1 row deletion, promptly can generate G _c.This shows, G _cBe (L _Mid-W _{Max, c}+ 1) * W _cK _{M, c}The dimension matrix, wherein, W _{Max, c}W need satisfy condition _cK _{M, c}≤ L _Mid-W _{Max, c}+ 1, and if L _Mid-W _cK _{M, c}〉=W _c-1, W then _{Max, c}=W _c, otherwise, W _{Max, c}=L _Mid-W _cK _{M, c}+ 1.

In addition, G _DFT.cFor the diagonal angle vector is $\underset{\‾}{g}=\mathrm{DFT}\left({\underset{\‾}{m}}_p^c\right)$ Diagonal matrix.

By the aforementioned calculation process as can be seen, channel estimate matrix G _cBe associated with the basic midamble sign indicating number that sub-district c uses, therefore, under the situation of the basic midamble sign indicating number of known cell c, channel estimate matrix G _cAlso be known.Therefore, the described channel estimation module of present embodiment can be according to known channel estimate matrix G _cAnd input signal, by above-mentioned formula (1), (2) or (3) calculate the associating CIR of sub-district c

2. the strong signal of neighbor cell is directly selected module

As previously mentioned, the strong signal of described neighbor cell is directly selected module to be mainly used in associating CIR according to each neighbor cell of the channel estimation module of respective cell output to select to determine (for example N) of number in all multipath signals of neighbor cell strong signal directly.

In actual applications, the strong signal of neighbor cell directly selects module can adopt several different methods to select strong signal footpath from all multipath signals.Present embodiment has specifically provided following three kinds of method for optimizing of selecting N strong signal footpath from all multipath signals of neighbor cell.

Method 1: the associating CIR according to each neighbor cell selects N multipath signal the strongest from all multipath signals of neighbor cell.Its detailed process mainly comprises as shown in Figure 2:

Step 201: the power that calculates each multipath signal in all neighbor cells.

In this step, the power of described each multipath signal of neighbor cell can calculate by following formula (4):

$v_{c,i}^{k_m}=\frac{1}{K_a}\underset{k_a=0}{\overset{K_a-1}{\mathrm{\Σ}}}{\left|\right|{\underset{\‾}{h^{\′\′}}}_{c,i}^{(k_a,k_m)}\left|\right|}^2,$ i＝1，2，...，W _c，k _m＝0，...，K _m，c-1，c＝1，2，...，C (4)

Step 202:, therefrom select N multipath signal the strongest according to the power of each multipath signal in all neighbor cells of step 201 calculating.

In order to realize above-mentioned steps,, then claim this multipath signal to be strong signal footpath if the power of a multipath signal of definition is higher than another multipath signal.According to above definition, just can from all multipath signals, select N the strongest multipath signal of power according to the power of each multipath signal in all neighbor cells, promptly the performance number of selected multipath signal should be all multipath signal performance numbers Middle maximum N.

Method 2: at first select the stronger CIR of power, and in the multipath signal of selected CIR correspondence, select N multipath signal the strongest.This method can directly be selected block search N the shared resource in peak signal footpath by reducing the hunting zone strong signal of saving neighbor cell.Its concrete grammar mainly comprises as shown in Figure 3:

Step 301: calculate the power of each CIR, promptly calculate each CIR correspondence all multipath signals power and.

The power of described each CIR of this step can utilize following formula (5) to calculate:

$v_c^{k_m}=\underset{i=1}{\overset{W_c}{\mathrm{\Σ}}}\underset{k_a=0}{\overset{K_a-1}{\mathrm{\Σ}}}{\left|\right|{\underset{\‾}{h^{\′\′}}}_{c,i}^{(k_a,k_m)}\left|\right|}^2,k_m=0,...,K_{m,c}-1,c=\mathrm{1,2},...,C---\left(5\right)$

Step 302: power average value and the variance of calculating all CIR.

Wherein, the computing formula of the power average value of CIR and variance is respectively shown in formula (6) and (7).

${\mathrm{\μ}}_v=\frac{1}{K_{m,\mathrm{Total}}}\underset{c=1}{\overset{C}{\mathrm{\Σ}}}\underset{k_m=0}{\overset{K_{m,c}-1}{\mathrm{\Σ}}}{v}_c^{k_m},$ $K_{m,\mathrm{Total}}=\underset{c=1}{\overset{C}{\mathrm{\Σ}}}{K}_{m,c}---\left(6\right)$

${\mathrm{\σ}}_v^2=\frac{1}{K_{m,\mathrm{Total}}}\underset{c=1}{\overset{C}{\mathrm{\Σ}}}\underset{k_m=0}{\overset{K_{m,c}-1}{\mathrm{\Σ}}}{(v_c^{k_m}-{\mathrm{\μ}}_v)}^2---\left(7\right)$

Step 303: according to selecting the stronger CIR of power among the power average value of the determined CIR of step 302 and variance all CIR in described neighbor cell.

In this step, the method for selection specifically may further comprise the steps:

At first, the mean value and the variance of all CIR power that calculate according to step 302 are determined a threshold value α=μ _v+ β σ _v, wherein, β 〉=0 is predetermined thresholding coefficient, for example, and β=1.

Then, select from all CIR of described neighbor cell according to the threshold value of determining $v_c^{k_m}>\mathrm{\α}$ CIR.Be the stronger CIR of the described power of this step at this selected CIR.

Step 304: from the stronger CIR of selected power, select N the strongest multipath signal of power.

In this step, if the sum of all multipath signals that the selected CIR of step 303 is comprised is less than or equal to N, then N is updated to the number of the multipath signal that the selected CIR of step 303 comprised.In this case, all multipath signals that CIR comprised that each selected power is stronger are the selected N of this step multipath signal the strongest.

This shows, method 2 is at first selected the stronger CIR of power by step 301-303 from all CIR, and then from the stronger multipath signal that CIR comprised of selected power, select N the strongest multipath signal of power, the hunting zone in N peak signal footpath is reduced to the stronger multipath signal that CIR comprised of power from all multipath signals, thereby reduced the hunting zone in N peak signal footpath greatly, reduced amount of calculation.

Method 3: at first select the stronger CIR of power, and in selected CIR, select N multipath signal the strongest.Method 3 also can be by reducing to search for N the required resource of the strongest multipath signal of scope saving search in N peak signal footpath.Its concrete grammar mainly comprises as shown in Figure 4:

Step 401: order $N_{\mathrm{shift}}=\left|\frac{N}{N_{\mathrm{path}}}\right|,$ Wherein, N _PathThe number of representing the strong signal footpath among the CIR, and N _PathShould satisfy greater than 1 and $\frac{N}{K_{m,\mathrm{Total}}}\≤N_{\mathrm{path}}\≤\underset{c=\mathrm{1,2},...C}{\min \left\{W_c\right\}}$ Condition.For example, make N _Path=4.

Step 402: from all CIR of all neighbor cells, select N _ShiftIndividual CIR, wherein, the power of each selected CIR should be all CIR power

In the strongest N _ShiftIndividual.

Step 403: at the selected N of step 402 _ShiftSelect N the strongest multipath signal of power among the individual CIR.

Similar with method 2, method 3 is at first selected N by step 401 and 402 from all CIR _ShiftThe strongest CIR of individual power, and then from the multipath signal that selected CIR comprised, select N the strongest multipath signal of power, N peak signal hunting zone directly is reduced to N from all multipath signals _ShiftThe multipath signal that individual CIR comprised, thus N peak signal hunting zone directly reduced greatly, reduced amount of calculation.

After the strong signal of described neighbor cell directly selects to select module all multipath signals in neighbor cell N peak signal footpath, this N selected peak signal footpath will be sent to multi-plot joint channel estimation function module.In order to express easily, selected peak signal directly will use by multipath index i midamble offset index k _mAnd tlv triple (i, the k of cell index c composition _m, c) sign.

${\underset{\‾}{h^{\′\′}}}_c^{(k_a,k_m)}={[{\underset{\‾}{h^{\′\′}}}_{c,1}^{(k_a,k_m)},{\underset{\‾}{h^{\′\′}}}_{c,2}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{c,i}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{c,W_c}^{(k_a,k_m)}]}^T$ Middle corresponding each selected multipath signal (i, k _m, fading coefficients c)

To in follow-up data field multi-plot joint testing process, obtain upgrading.

3, the strong signal in this sub-district is directly selected module

As previously mentioned, the strong signal in described this sub-district directly selects module to be mainly used in to select in all multipath signals in this sub-district to determine (for example M) multipath signal that power is the strongest of number.Directly select process as shown in Figure 5 by the strong signal that this module is finished, mainly comprise:

Step 501: the power that calculates each multipath signal in this sub-district.

In this step, the power of each multipath signal can calculate by following formula (8) in this sub-district:

$v_{C+1,i}^{k_m}=\frac{1}{K_a}\underset{k_a=0}{\overset{K_a-1}{\mathrm{\Σ}}}{\left|\right|{\underset{\‾}{h^{\′\′}}}_{C+1,i}^{(k_a,k_m)}\left|\right|}^2,$ i＝1，2，...，W _C+1，k _m＝0，...，K _m，C+1-1 (8)

Step 502:, therefrom select M the strongest multipath signal of power according to the power of each multipath signal in this sub-district of step 501 calculating.

In this step, select the method for M the multipath signal that power is the strongest specifically to comprise the steps:

At first, order

If N _{Local, path}＞W _C+1, then make N _{Lcal, path}=W _C+1If N _{Local, shift}=M-K _{M, C+1}N _{Local, path}＞0 and N _{Local, path}＜W _C+1, then at k _m=0 ..., N _{Local, shift}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}}+1,$ And at k _m=N _{Local, shift}..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}}.$ Otherwise, at k _m=0 ..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}}.$

Then, the power of all multipath signals that calculate according to step 501

Each midamble is offset pairing CIR from this sub-district In all multipath signals in select

The strongest multipath signal of individual power.

At last, utilize formula $M=\underset{k_m=0}{\overset{K_{m,C+1}-1}{\mathrm{\Σ}}}{N}_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}$ Upgrade M.

Directly select module selects M peak signal footpath all multipath signals in this sub-district after at the strong signal in described this sub-district, this M selected peak signal footpath will be sent to multi-plot joint channel estimation function module.In order to express easily, selected peak signal directly will use by multipath index i midamble offset index k _mAnd tlv triple (i, the k of cell index C+1 composition _m, C+1) sign.

${\underset{\‾}{h^{\′\′}}}_{C+1}^{(k_a,k_m)}={[{\underset{\‾}{h^{\′\′}}}_{C+1,1}^{(k_a,k_m)},{\underset{\‾}{h^{\′\′}}}_{C+1,2}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{C+1,}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{C+1,W_{C+1}}^{(k_a,k_m)}]}^T$ Multipath signal (i, the k of middle corresponding each selection _m, fading coefficients C+1)

To in the multi-plot joint testing process of follow-up data territory, obtain upgrading.

4, multi-plot joint channel estimation module

As previously mentioned, this multi-plot joint channel estimation function module is mainly used in according to directly select module from the strong signal of neighbor cell N strong signal footpath with from the strong signal in this sub-district and directly selects M strong signal of module directly to carry out the multi-plot joint channel estimating.

The described multi-plot joint channel estimating of present embodiment, the number N in the strong signal footpath of selecting in neighbor cell and the strong signal footpath number of selecting in this sub-district need satisfy following condition:

(1)(N+M)≤L

(2)0＜M≤K _m，C+1W _C+1

(3)0＜N≤L-M

The detailed process of the multi-plot joint channel estimating that is realized by the described multi-plot joint channel estimation module of present embodiment mainly comprises as shown in Figure 6:

Step 601: calculate the required parameters such as multi-plot joint channel estimate matrix of multi-plot joint channel estimating.

In this step, known, being input as of multi-plot joint channel estimation module Output should be multi-plot joint CIR column vector In order to carry out midamble territory multi-plot joint channel estimating, the multi-plot joint channel estimation module need at first be known respective antenna k _a=0 ..., K _a-1 multi-plot joint channel estimate matrix JG

Known JGRole and G in the multi-plot joint channel estimation process of midamble territory _cRole is basic identical in single cell channel estimation procedure.And,

In the multi-plot joint channel estimation process of midamble territory role with

Role is also basic identical in single cell channel estimation procedure.

Since the multipath signal index i of each selected multipath signal, midamble offset index k _mAnd cell index c or C+1 will be sent to this multi-plot joint channel estimation module.It is 1 to N+M multipath signal that the multi-plot joint channel estimation module renumbers selected multipath signal.

Below, use i, k _mAnd c represents the multipath signal index through the individual multipath signal of l ' that rearranges, midamble offset index and cell index respectively.Can be with matrix G _{Temp, c}L row copy to matrix JG _TempL ' row, wherein, l=W _ck _m+ i and l '=1 ..., N+M.So, obtain matrix JG _TempThen, deletion again JG _TempIn preceding W _{Max, C+1}-1 row generates JG

In addition, can also be with column vector

Capable the copying to of l

L ' OK, wherein, k _a=0 ..., K _a-1, obtain column vector

Being used for follow-up data field multi-plot joint detects.

And the power of the individual multipath signal of l ' can calculate by following formula (9).

$V_{l^{\′}}=v_{c,i}^{k_m}=\frac{1}{K_a}\underset{k_a=0}{\overset{K_a-1}{\mathrm{\Σ}}}{\left|\right|{\underset{\‾}{h^{\′\′}}}_{c,i}^{(k_a,k_m)}\left|\right|}^2,$ l′＝1，...，N+M (9)

Step 602: utilize the multi-plot joint channel estimate matrix that generates to carry out midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector

The described combined channel of this step is estimated main basis ${\underset{\‾}{e}}_m^{\left(k_a\right)}=\underset{\‾}{\mathrm{JG}}{\underset{\‾}{{\mathrm{Jh}}^{\′\′}}}^{\left(k_a\right)}+{\underset{\‾}{\mathrm{Jn}}}^{\left(k_a\right)}$ Calculate Wherein,

K for antenna in the multi-plot joint channel estimation process of midamble territory _aNoisy vector.

Preferably, the described midamble of this step territory multi-plot joint channel estimating can use ZF detector or MMSE detector to estimate

When using the MMSE detector, the multi-plot joint channel estimation module can calculate by following formula (10)

${\underset{\‾}{{\mathrm{Jh}}^{\′\′}}}^{\left(k_a\right)}={({\underset{\‾}{\mathrm{JG}}}^H{\left(R_{\mathrm{Jn}}^{\left(k_a\right)}\right)}^{-1}\underset{\‾}{\mathrm{JG}}+{\left(R_{\mathrm{Jh}}^{\left(k_a\right)}\right)}^{-1})}^{-1}{\underset{\‾}{\mathrm{JG}}}^H{\left(R_{\mathrm{Jn}}^{\left(k_a\right)}\right)}^{-1}{\underset{\‾}{e}}_{m,C+1}^{\left(k_a\right)},$ k _a＝0，...，K _a-1 (10)

Wherein,

For

Covariance matrix and,

For Covariance matrix.

When $R_{\mathrm{Jn}}^{\left(k_a\right)}={\mathrm{\σ}}_{\mathrm{Jn}}^{2}I,$ And σ _Jn ²Expression

During the noise power of each element,

Can further be expressed as ${\underset{\‾}{{\mathrm{Jh}}^{\′\′}}}^{\left(k_a\right)}={({\underset{\‾}{\mathrm{JG}}}^H\underset{\‾}{\mathrm{JG}}+{\mathrm{\σ}}^2{\left(R_{\mathrm{Jn}}^{\left(k_a\right)}\right)}^{-1})}^{-1}{\underset{\‾}{\mathrm{JG}}}^H{\underset{\‾}{e}}_{m,C+1}^{\left(k_a\right)},$ k _a=0 ..., K _a-1, wherein

$R_{\mathrm{Jh}}^{\left(k_a\right)}=E\left({\underset{\‾}{{\mathrm{Jh}}^{\′\′}}}^{\left(k_a\right)}{\underset{\‾}{{\mathrm{Jh}}^{\′\′}}}^{\left(k_a\right)H}\right)=E\left\{\frac{1}{K_a}\underset{k_a=0}{\overset{K_a-1}{\mathrm{\Σ}}}\left({\underset{\‾}{{\mathrm{Jh}}^{\′\′}}}^{\left(k_a\right)}{\underset{\‾}{{\mathrm{Jh}}^{\′\′}}}^{\left(k_a\right)H}\right)\right\}=\left[\begin{array}{cccc}{V}_1& 0& 0...& 0\\ 0& V_2& 0...& 0\\ .& .& .& .\\ .& .& .& .\\ .& .& .& .\\ 0& 0& 0& V_{N+M}\end{array}\right],$ And, V _{L '}, l '=1 ..., N+M, expression

The power of the individual multipath signal of l '.

Can estimate multi-plot joint CIR column vector by above-mentioned steps 601 and 602

In order to simplify the aforementioned calculation process, reduce the computation complexity of multi-plot joint channel estimating, another embodiment of the present invention gives a kind of multiple cell combined channel estimation method of simplification, and its main process comprises:

In the multi-plot joint channel estimation process of midamble territory, if (N+M)＜L, can suitably reduce the multi-plot joint channel estimate matrix JGWith

Line number, realize to reduce the purpose of computation complexity.Specifically, can be at first from the multi-plot joint channel estimate matrix JGWith

After the conduct of the middle L ' of selection row is upgraded JGWith

Wherein, N+M≤L '≤L.Matrix after the renewal JGWith With the matrix before the renewal JGWith

Compare, selected each row remains unchanged, and nonoptional row is deleted.After this, the matrix after the utilization of multi-plot joint channel estimation module is upgraded _JGWith

Carry out midamble territory multi-plot joint channel estimating by foregoing method.

Because the multi-plot joint channel estimate matrix after upgrading JGWith

Compare line number with the matrix before the renewal and significantly reduced, therefore correspondingly, computation complexity also greatly reduces.

5, multi-plot joint sytem matrix generation module

The described data field multi-plot joint of present embodiment detects and can utilize the MMSE detection algorithm to realize.The system model that uses in the joint detection algorithm based on MMSE is e=Ad+n, and wherein, e is the chip-level sequence that is received, and d is a data symbol sequence, and n is the chip-level noise sequence, and A is a sytem matrix.

According to above system model, in order to realize the data field joint-detection, obtain data symbol sequence, must at first know sytem matrix A.The major function of this multi-plot joint sytem matrix generation module just is the generation system matrix A.

The method of the generation system matrix that present embodiment provides mainly comprises as shown in Figure 7:

Step 701: upgrade corresponding C IR according to the multi-plot joint channel estimation results.

Specifically, in this step, the method for the corresponding CIR of described renewal is that (i, km c), use for the individual multipath signal of selected l '

In the individual element of l ' upgrade ${\underset{\‾}{h^{\′\′}}}_c^{(k_a,k_m)}={[{\underset{\‾}{h^{\′\′}}}_{c,1}^{(k_a,k_m)},{\underset{\‾}{h^{\′\′}}}_{c,2}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{c,i}^{(k_a,k_m)},...,{\underset{\‾}{h^{\′\′}}}_{c,W}^{(k_a,k_m)}]}^T$ In Wherein, l '=1,2 ..., N+M.

Step 702: specific (proprietary) CIR that calculates each UE.

In this step, adopt following formula (11) to calculate the specific CIR of described each UE.

${\underset{\‾}{h^{\′\′}}}_{\mathrm{av},c}^{(k_a,k_m)}=\frac{\sqrt{K_{\mathrm{VRU}}^{\left(k_m\right)}}}{\underset{\underset{\mathrm{same\; UE}}{\mathrm{all}{k}_m\mathrm{of}}}{\mathrm{\Σ}}{K}_{\mathrm{VRU}}^{\left(k_m\right)}}\underset{\underset{\mathrm{same\; UE}}{\mathrm{all}{k}_m\mathrm{of}}}{\mathrm{\Σ}}\sqrt{K_{\mathrm{VRU}}^{\left(k_m\right)}}{\underset{\‾}{h^{\′\′}}}_c^{(k_a,k_m)}---\left(11\right)$

Wherein,

For using midamble skew k _mThe number of virtual resource unit (VRU, Virtual Resource Unit), for example, when SF=8, K _VRU=2.Generally, VRU resource that can provide and a SF are that 16 RU is identical.

Step 703: carry out the channel estimating reprocessing.

In this step, utilize following formula (12) to carry out described channel estimating reprocessing:

${\underset{\‾}{h^{\′}}}_{c,i}^{(k_a,k_m)}=\left\{\begin{array}{c}{\underset{\‾}{h^{\′\′}}}_{\mathrm{av},c,i}^{(k_a,k_m)};\frac{1}{K_a}\underset{\mathrm{all}{k}_a}{\mathrm{\Σ}}{\left|\right|{\underset{\‾}{h^{\′\′}}}_{\mathrm{av},c,i}^{(k_a,k_m)}\left|\right|}^2\≥\mathrm{\Γ}\\ 0;\mathrm{else}\end{array}\right.,$ i＝1，...，W _c，k _a＝0，...，K _a-1，k _m＝0，...，K _m，c-1 (12)

After carrying out above-mentioned channel estimating reprocessing, CIR

To be updated to

Then,

To be output to the midamble interference cancellation module.Wherein, Γ represents to be used for deleting the noise gate of CIR window noise multipath signal.It is that noise power adds 3dB that Γ preferably can be set.

Step 704: the CIR that calculates specific VRU.

In this step, utilize following formula (13) to calculate the CIR of specific VRU.

${\underset{\‾}{h}}_c^{(k_a,k_u)}=\sqrt{\frac{\left(\frac{16}{S{F}^{\left(k_{\mathrm{ru}}\right)}}\right)}{K_{\mathrm{VRU}}^{\left(f\left(k_{\mathrm{ru}}\right)\right)}}}{\underset{\‾}{h^{\′}}}_c^{(k_a,f\left(k_{\mathrm{ru}}\right))},$ k _ru＝0，...，K _ru，c-1，i＝1，...，W _c (13)

Wherein, f (k _Ru) the expression number is k _RuThe employed midamble of RU skew,

The spread spectrum coefficient of representing this RU, and K _{Ru, c}The quantity of all relevant RU in the expression sub-district c.

Step 705: the quantity of determining the VRU that the participation multi-plot joint detects.

The line number of supposing the system matrix A is K _a(L _Data+ W _C+1-1), columns is K _RuN ₁₆, N wherein ₁₆=22, K _RuFor participating in the number of the VRU that multi-plot joint detects.Because K _RuN ₁₆≤ K _a(L _Data+ W _C+1-1), therefore has

Step 706: from described neighbor cell, select K _{Ru, adjacent}=K _Ru-K _{Ru, C+1}Individual VRU, wherein, K _{Ru, C+1}Number for the VRU that activates in this sub-district.

Definition $u(c,k_{\mathrm{ru}})=\underset{k_a=0}{\overset{K_a-1}{\mathrm{\Σ}}}{\left|\right|{\underset{\‾}{h}}_c^{(k_a,k_{\mathrm{ru}})}\left|\right|}^2$ Be VRUk among the c of sub-district _RuPower, k wherein _Ru=0,1 ..., K _{Ru, c}-1 and c=1,2 ..., C.Thereby, can be according to power u (c, the k of all VRU in the neighbor cell _Ru) selection K _{Ru, adjacent}The VRU of individual power maximum correspondence is as selected VRU.

Step 707: the associating CIR that calculates selected VRU in each local VRU and the special neighbourhood sub-district.

Consider scrambler and channel code, the associating CIR of the specific VRU of described correspondence is ${\underset{\‾}{b}}_c^{(k_a,k_{\mathrm{ru}})}={\underset{\‾}{h}}_c^{(k_a,k_{\mathrm{ru}})}*{\underset{\‾}{c}}_c^{\left(k_{\mathrm{ru}}\right)},k_a=0,...,K_a-1,$ Wherein,

Be VRUk among the c of sub-district _RuScrambler and the product of channel code.

Step 708: calculate sytem matrix A according to step 707.

At first, utilize formula

Generate the matrix of corresponding each antenna

According to above-mentioned formula,

Comprise N ₁₆=22 skews Wherein, Move down Q element at every turn, Q=16 for example, K simultaneously moves right _RuIndividual element.

Comprise the interior VRU's of all local VRU and selected neighbor cell

All K _RuIndividual VRU from 1 to K _RuArrange.Preceding K _{Ru, C+1}Individual VRU is local VRU, the K of back _{Ru, adjacent}Individual VRU is from neighbor cell.If the associating CIR of l corresponding VRU is

Then

For

L row.

Next, pass through computing formula $A=\left[\begin{array}{c}{A}^{\left(0\right)}\\ .\\ .\\ .\\ A^{(K_a-1)}\end{array}\right]$ Obtain sytem matrix A.This sytem matrix A will be as the output of multi-plot joint sytem matrix generation module.

6, midamble interference cancellation module

The major function of described midamble interference cancellation module is to eliminate the interference of the midamble sign indicating number of this sub-district and neighbor cell to two data blocks.Described midamble interference cancellation module can be estimated the interference of this sub-district midamble sign indicating number to two data blocks, from the baseband signal of first and second data blocks that receive this interference is cut respectively then.In addition, the midamble sign indicating number that described midamble interference cancellation module can also be estimated neighbor cell cuts this interference to the interference of two data blocks and from the purer baseband signal of having eliminated first and second data blocks that this sub-district midamble sign indicating number disturbs, obtains the baseband signal of the first and second further pure data blocks.

Suppose that the baseband signal of first data block that is received is

${\underset{\‾}{e}}_{r1,i}^{\left(k_a\right)}={\underset{\‾}{e}}_{r,i}^{\left(k_a\right)},$ I=0,1 .., L _Data+ W _C+1-2, the baseband signal of second data block that is received is

${\underset{\‾}{e}}_{r2,i}^{\left(k_a\right)}={\underset{\‾}{e}}_{r,i+L_{\mathrm{data}}+L_{\mathrm{mid}}}^{\left(k_a\right)},$ i＝0，1，..，L _data+W _C+1-2。

Use the midamble sign indicating number of known sub-district c

And the specific CIR of UE

The midamble sign indicating number of reconstruct (rebuild) sub-district c is to the interference of two data blocks, from the baseband signal of first and second data blocks, cut the estimation structure of interference of the midamble sign indicating number of this sub-district and neighbor cell then respectively, obtain the clean baseband signal of first and second data blocks With

Make up following two column vectors $e_1=\left[\begin{array}{c}{\underset{\‾}{e}}_1^{\left(0\right)}\\ .\\ .\\ .\\ \underset{\‾}{{\underset{\‾}{e}}_1^{(K_a-1)}}\end{array}\right]$ With $e_2=\left[\begin{array}{c}{\underset{\‾}{e}}_2^{\left(0\right)}\\ .\\ .\\ .\\ \underset{\‾}{{\underset{\‾}{e}}_2^{(K_a-1)}}\end{array}\right].$ And the row amount e that will make up ₁And e ₂Output as the midamble interference cancellation module.

7, MMSE detection module

Be used to from the sytem matrix A of multi-plot joint sytem matrix generation module and from the clean data block signal e of midamble interference cancellation module ₁And e ₂, carry out the MMSE multi-plot joint and detect.

Particularly, described MMSE detection module is by carrying out following matrix operation formula (14) from described data block signal e ₁And e ₂In calculate the first data block d ₁With the second data block d ₂

$d_l={(A^H{R}_{n,l}^{-1}A+R_{d,l}^{-1})}^{-1}{A}^H{R}_{n,l}^{-1}{e}_l---\left(14\right)$

Wherein, l=1 represents first data block, and l=2 represents second data block, e _l=Ad _l+ n _l, R _{N, l}Be noise sequence n _lCovariance matrix, R _{D, l}Be d _lCovariance matrix.At this, $R_{d,l}=I_{K_{\mathrm{ru}}{N}_{16}}$ And $R_{n,l}={\mathrm{\σ}}^2{I}_{K_a(L_{\mathrm{data}}+W_{C+1}-1)},$ Wherein, I _DRepresent the unit matrix (unitary matrix) and the σ of D * D dimension _{N, l} ²Be n _lThe variance of noise, suppose σ in the present embodiment _{N, l} ²Known.

d _lIn jK _Ru+ i element Wherein, j=0 ..., N ₁₆-1 and i=0,1 ..., K _{Ru, C+1}-1, represent j+1 the symbol of i+1 VRU of this l data block in sub-district.L=1 represents first data block, and l=2 represents second data block.d _lElement arrange since 0.

This shows, at known system matrix A and data block signal e ₁And e ₂Situation under, by above-mentioned matrix operation, can obtain the first data block d in the burst packets that transmitting terminal sends ₁With the second data block d ₂Thereby, realize that multi-plot joint detects.

From foregoing description as can be seen, the described mdamble of present embodiment territory multi-plot joint channel estimating unit is at first selected the stronger strong signal footpath of signal in all multipath signals of each user in neighbor cell, also in this sub-district, select the stronger strong signal footpath of signal simultaneously in all multipath signals of each user, and then utilize the strong signal of user in selected neighbor cell and this sub-district directly to carry out midamble territory multi-plot joint channel estimating, can obtain good CIR estimated performance.The described data field multi-plot joint of present embodiment detecting unit utilizes the output of midamble territory multi-plot joint channel estimating unit to carry out the data field multi-plot joint and detects, thereby can obtain better joint-detection gain.

The described multi-plot joint sytem matrix of the foregoing description generation module uses the specific associating CIR of VRU to generate sytem matrix A.Wherein, the dimension of sytem matrix A is K _a(L _Data+ W _C+1-1) * K _RuN ₁₆Like this, work as K _{Ru, adjacent}When bigger, for the data block that obtains this sub-district will need very big amount of calculation.

For this reason, an alternative embodiment of the invention provides a kind of improved sytem matrix generation module, be used to produce a small-scale sytem matrix A, just, improved sytem matrix generation module reduces amount of calculation by line number and/or the columns that reduces sytem matrix A.

In order to reduce the columns of sytem matrix A, should at first in neighbor cell, select suitable VRU.If the dimension of sytem matrix A is defined as K _a(L _Data+ W _C+1-1) * K _{Ru, fix}N ₁₆, and K _{Ru, C+1}≤ K _{Ru, fix}≤ K _Ru, can from all VRU of neighbor cell, select K _{Ru, fix, adjacent}=K _{Ru, fix}-K _{Ru, C+1}Individual VRU.Wherein, selected VRU should be the highest K of power among all VRU of all neighbor cells _{Ru, fix, adjacent}Individual VRU.

Next, use the specific associating CIR of the VRU that in neighbor cell, selects

(the specific CIR of c ≠ C+1) and all VRU of this sub-district $({\underset{\‾}{b}}_c^{(k_a,k_{\mathrm{ru}})},c=C+1)$ Generate sytem matrix A.The sytem matrix A that present embodiment generated is called the improved sytem matrix A that is used to reduce amount of calculation.

In the present embodiment, with all selected K _{Ru, fix}Individual VRU again from 1 to K _{Ru, fix}Arrange wherein preceding K _{Ru, C+1}Individual VRU is local VRU, the K of back _{Ru, fix, adjacent}Individual VRU is from neighbor cell.If the specific associating CIR of selected l VRU is

Then

L row will for Wherein, l=1,2 .., K _{Ru, fix}

At this moment, can utilize formula $A=\left[\begin{array}{c}{A}^{\left(0\right)}\\ .\\ .\\ .\\ A^{(K_a-1)}\end{array}\right]$ Calculate improved sytem matrix A, wherein,

Compare with last embodiment,

Move down Q element, K simultaneously moves right at every turn _{Ru, fix}Individual element.

Utilize described improved sytem matrix A, the MMSE detection module still can pass through formula (14), $d_l={(A^H{R}_{n,l}^{-1}A+R_{d,l}^{-1})}^{-1}{A}^H{R}_{n,l}^{-1}{e}_l$ Calculate d _lWherein, d _lIn jK _{Ru, fix}+ i element

J=0 ..., N ₁₆-1, i=0 ..., K _{Ru, C+1}-1, represent j+1 symbol of i+1 VRU correspondence in l the data block in this sub-district.L=1 represents first data block, and l=2 represents second data block.

At K _{Ru, fix}=K _{Ru, C+1}The time, improved multi-plot joint sytem matrix generation module will generate single cell system matrix, and this matrix comprises the associating CIR of all corresponding VRU in this sub-district.Thus, improved multi-plot joint sytem matrix generation module will deteriorate to single cell system matrix generation module.Corresponding, the MMSE detection module is also with the community associated detection of fill order.

In order further to reduce computation complexity, can also further reduce the line number of sytem matrix A, for example, the MMSE detection module can be selected L from sytem matrix A _ARow generates the sytem matrix A that simplifies, thereby the line number of sytem matrix A is defined as L _A, L wherein _AK satisfies condition _{Ru, fix}N ₁₆≤ L _A≤ K _a(L _Data+ W _C+1-1).

Utilize the sytem matrix A of described further simplification, the MMSE detection module still can pass through formula (14) $d_l={(A^H{R}_{n,l}^{-1}A+R_{d,l}^{-1})}^{-1}{A}^H{R}_{n,l}^{-1}{e}_l,$ Calculate d _lWherein, d _lIn jK _{Ru, fix}+ i element J=0 ..., N ₁₆-1, i=0 ..., K _{Ru, C+1}-1, represent j+1 symbol of i+1 VRU correspondence in l the data block in this sub-district.L=1 represents first data block, and l=2 represents second data block.

Except that above-mentioned multi-plot joint checkout gear, embodiments of the invention give the method for multi-plot joint channel estimating and based on the multi-cell joint detection method of this multiple cell combined channel estimation method.

Describe the method and the multi-cell joint detection method of the multi-plot joint channel estimating that present embodiments provide in detail below in conjunction with accompanying drawing 8.As shown in Figure 8, the described multiple cell combined channel estimation method of present embodiment mainly comprises:

Step 801: respectively all neighbor cells are carried out channel estimating, obtain the associating CIR of each neighbor cell, and from the multipath signal of all neighbor cells, select to determine number according to the associating CIR of each neighbor cell that obtains, for example N, strong signal footpath.

To carry out the method for channel estimating identical for channel estimation module in this step is described carries out channel estimating to neighbor cell method and the foregoing description.In addition, this step is described select to determine the strong signal footpath of number from the multipath signal of all neighbor cells method also can use in the foregoing description the strong signal of neighbor cell directly to select module to select to determine the strong signal method directly of number, i.e. Fig. 2, Fig. 3 or method shown in Figure 4 from the multipath signal of all neighbor cells.

Step 802: channel estimating is carried out in this sub-district, obtain the CIR of this sub-district,, from the multipath signal of this sub-district, select to determine number according to the CIR of this sub-district, for example M, strong signal footpath.

The described method that channel estimating is carried out in this sub-district of this step can adopt above-mentioned channel estimation module to carry out the method for channel estimating, and the described strong signal method of selecting to determine number from the multipath signal of this sub-district directly also can adopt the strong signal in above-mentioned this sub-district directly to select module to select the method in strong signal footpath, method promptly shown in Figure 5 from all multipath signals of this sub-district.

Need to prove that present embodiment is to the not restriction of execution sequence of step 801 and 802, promptly step 801 and 802 can executed in parallel also can be carried out with the random order sequencing.

Step 803: utilize step 801 and the 802 strong signals of selecting directly to carry out the multi-plot joint channel estimating, obtain the associating CIR and the multi-plot joint CIR column vector of corresponding each sub-district.

The described method of carrying out the multi-plot joint channel estimating of this step can adopt multi-plot joint channel estimation module in the foregoing description to carry out the method for multi-plot joint channel estimating, method promptly shown in Figure 6.

This shows,, can finish the multi-plot joint channel estimating, obtain the associating CIR and the multi-plot joint CIR column vector of corresponding each sub-district by carrying out above-mentioned steps 801-803.And the multipath signal that is used for the multi-plot joint channel estimating in the present embodiment is the strong signal of all multipath signals footpath in neighbor cell and this sub-district, and therefore, above-mentioned multi-plot joint channel estimating can obtain good CIR estimated performance.

After executing above-mentioned steps 803, can also further carry out following steps 804.

Step 804: the matrix of generation system as a result that utilizes the multi-plot joint channel estimating, and then carry out the data field multi-plot joint according to the first and second data block signals of sytem matrix that is generated and reception and detect, obtain the first and second data block data carried by data.

In this step, the method of generation system matrix both can be with reference to the method for multi-plot joint sytem matrix generation module generation system matrix in the foregoing description, be method shown in Figure 7, can also generate the method for improving sytem matrix with reference to improved multi-plot joint sytem matrix generation module in the foregoing description.The method that described data field multi-plot joint detects then can be carried out the method that the data field multi-plot joint detects with reference to MMSE module in the foregoing description.

In addition, in above-mentioned steps 804, can also before carrying out the detection of data field multi-plot joint, utilize the result of multi-plot joint channel estimating that the first and second data block signals that receive are handled earlier in order further to improve multi-plot joint detection gain, elimination midamble block signal in transmission course obtains the first and second clean data block signals to the interference of the described first and second data block signals.And in subsequent operation, the first and second clean data block signals are carried out multi-plot joint detect.

Because in step 804, sytem matrix is to obtain according to the multi-plot joint channel estimation results that step 801-803 obtains, therefore, and at the CIR estimated performance preferably under the situation, the sytem matrix that obtains is also just truer, detects gain thereby can obtain bigger multi-plot joint.

The above only is preferred embodiment of the present invention, and is in order to restriction the present invention, within the spirit and principles in the present invention not all, any modification of being done, is equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (25)

1. a multiple cell combined channel estimation method is characterized in that, comprising:

Estimate the channel impulse response CIR of each neighbor cell and this sub-district;

CIR according to each neighbor cell selects N strong signal footpath from the multipath signal of all neighbor cell user correspondences, according to the CIR of this sub-district, select M strong signal footpath from the multipath signal of this community user correspondence, and wherein, N and M are natural number;

Utilize selected N+M strong signal directly to carry out the multi-plot joint channel estimating, obtain multi-plot joint CIR column vector.

2. multiple cell combined channel estimation method according to claim 1 is characterized in that, the described step in N strong signal footpath of selecting from the multipath signal of all neighbor cell user correspondences comprises:

Calculate the power of each multipath signal in all neighbor cells, therefrom select N the multipath signal that power is the strongest as described N strong signal footpath.

3. multiple cell combined channel estimation method according to claim 1 is characterized in that, the described step in N strong signal footpath of selecting from the multipath signal of all neighbor cell user correspondences comprises:

Calculate the power of each CIR, and calculate power average value and the variance of all CIR;

Mean value and variance according to all CIR power that calculate are determined a threshold value, select the CIR of power greater than described thresholding from all CIR of described neighbor cell;

From selected CIR, select N the multipath signal that power is the strongest as described N strong signal footpath.

4. multiple cell combined channel estimation method according to claim 1 is characterized in that, the described step in N strong signal footpath of selecting from the multipath signal of all neighbor cell user correspondences comprises:

Determine the number in strong signal footpath among the CIR, and determine the number N of strong jamming CIR to be selected according to the number in strong signal footpath among each CIR that determines _Shift, wherein, N _ShiftBe natural number;

According to the power of each CIR, from all CIR of neighbor cell, select the strongest N of power _ShiftIndividual CIR;

From selected N _ShiftSelect N the multipath signal that power is the strongest as described N strong signal footpath in pairing all multipath signals of individual CIR.

5. multiple cell combined channel estimation method according to claim 1 is characterized in that, the described step in M strong signal footpath of selecting from the multipath signal of this community user correspondence comprises:

Calculate the power of each multipath signal in this sub-district, therefrom select M the multipath signal that power is the strongest as described M strong signal footpath.

6. multiple cell combined channel estimation method according to claim 5 is characterized in that, M multipath signal that power is the strongest of described selection comprises as the step in described M strong signal footpath:

Order

Wherein, K _{M, C+1}Midamble skew for this sub-district;

If N _{Local, path}＞W _C+1, then make N _{Local, path}=W _C+1, W wherein _C+1For the CIR window of this sub-district long;

If N _{Local, shift}=M-K _{M, C+1}N _{Local, path}＞0 and N _{Local, path}＜W _C+1, then at k _m=0 ..., N _{Local, shift}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}}+1,$ And at k _m=N _{Local, shift}..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$ Otherwise, at k _m=0 ..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$

According to the power of all multipath signals that calculate, from being offset all multipath signals of pairing CIR, each midamble of this sub-district selects

The strongest multipath signal of individual power;

Utilize formula $M=\underset{k_m=0}{\overset{K_{m,C+1}-1}{\mathrm{\Σ}}}{N}_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}$ Upgrade M.

7. multiple cell combined channel estimation method according to claim 1 is characterized in that, the described step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises:

Generate the required multi-plot joint channel estimate matrix of multi-plot joint channel estimating, and utilize the multi-plot joint channel estimate matrix that generates to carry out midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector.

8. multiple cell combined channel estimation method according to claim 1 is characterized in that, the described step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises:

Generate the required multi-plot joint channel estimate matrix of multi-plot joint channel estimating, therefrom select L ' row vector to regenerate the multi-plot joint channel estimate matrix of simplification, utilize the multi-plot joint channel estimate matrix of described simplification to carry out midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector, wherein, N+M≤L '≤L _Mid-W _{Max, c}+ 1, L _MidThe length of representing each midamble piece, W _{Max, c}The maximal possible length of expression channel impulse response window.

9. a multi-plot joint channel estimating apparatus is characterized in that, comprising: the strong signal of neighbor cell directly selects module, the strong signal in this sub-district directly to select module, multi-plot joint channel estimation module and at least one channel estimation module; Wherein,

Each channel estimation module is respectively applied for the channel impulse response CIR that estimates designated cell in neighbor cell and this sub-district;

The strong signal of described neighbor cell directly selects module to be used for selecting N strong signals footpaths according to the CIR of the channel estimation module of corresponding neighbor cell output all multipath signals of all users in the neighbor cell;

The CIR that the strong signal in described this sub-district directly selects module to be used for exporting according to the channel estimation module of corresponding this sub-district selects M strong signal footpaths from all multipath signals of all users of this sub-district;

N+M the strong signal that described multi-plot joint channel estimation module is used for directly selecting module and the strong signal in this sub-district directly to select module to select according to the strong signal of neighbor cell directly carries out the multi-plot joint channel estimating, obtain multi-plot joint CIR column vector, wherein, N and M are natural number.

10. a multi-cell joint detection method is characterized in that, comprising:

Estimate the channel impulse response CIR of each neighbor cell and this sub-district;

CIR according to each neighbor cell selects N strong signal footpath from the multipath signal of all neighbor cell user correspondences, according to the CIR of this sub-district, select M strong signal footpath from the multipath signal of this community user correspondence, and wherein, N and M are natural number;

Utilize selected N+M strong signal directly to carry out the multi-plot joint channel estimating, obtain multi-plot joint CIR column vector;

Utilize the matrix of generation system as a result of multi-plot joint channel estimating;

The first and second data block signals according to sytem matrix that is generated and reception carry out the detection of data field multi-plot joint, obtain the first and second data block data carried by data.

11. associated detecting method according to claim 10 is characterized in that, the described step in N strong signal footpath of selecting from the multipath signal of all neighbor cell user correspondences comprises:

Calculate the power of each multipath signal in all neighbor cells, therefrom select N the multipath signal that power is the strongest as described N strong signal footpath.

12. associated detecting method according to claim 10 is characterized in that, the described step in N strong signal footpath of selecting from the multipath signal of all neighbor cell user correspondences comprises:

Calculate the power of each CIR, and calculate power average value and the variance of all CIR;

Mean value and variance according to all CIR power that calculate are determined a threshold value, select the CIR of power greater than described thresholding from all CIR of described neighbor cell;

From selected CIR, select N the multipath signal that power is the strongest as described N strong signal footpath.

13. associated detecting method according to claim 10 is characterized in that, the described step in N strong signal footpath of selecting from the multipath signal of all neighbor cell user correspondences comprises:

Determine the number in strong signal footpath among the CIR, and determine the number N of strong jamming CIR to be selected according to the number in strong signal footpath among each CIR that determines _Shift, wherein, N _ShiftBe natural number;

According to the power of each CIR, from all CIR of neighbor cell, select the strongest N of power _ShiftIndividual CIR;

From selected N _ShiftSelect N the multipath signal that power is the strongest as described N strong signal footpath in pairing all multipath signals of individual CIR.

14. associated detecting method according to claim 10 is characterized in that, the described step in M strong signal footpath of selecting from the multipath signal of this community user correspondence comprises:

Calculate the power of each multipath signal in this sub-district, therefrom select M the multipath signal that power is the strongest as described M strong signal footpath.

15. associated detecting method according to claim 14 is characterized in that, M multipath signal that power is the strongest of described selection comprises as the step in described M strong signal footpath:

Order

Wherein, K _{M, C+1}Midamble skew for this sub-district;

If N _{Local, path}＞W _C+1, then make N _{Local, path}=W _C+1, W wherein _C+1For the CIR window of this sub-district long;

If N _{Local, shift}=M-K _{M, C+1}N _{Local, path}＞0 and N _{Local, path}＜W _C+1, then at k _m=0 ..., N _{Local, shift}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}}+1,$ And at k _m=N _{Local, shift}..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$ Otherwise, at k _m=0 ..., K _{M, C+1}-1 o'clock, order $N_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}=N_{\mathrm{local},\mathrm{path}};$

According to the power of all multipath signals that calculate, from being offset all multipath signals of pairing CIR, each midamble of this sub-district selects The strongest multipath signal of individual power;

Utilize formula $M=\underset{k_m=0}{\overset{K_{m,C+1}-1}{\mathrm{\Σ}}}{N}_{\mathrm{local},\mathrm{path}}^{\left(k_m\right)}$ Upgrade M.

16. associated detecting method according to claim 10 is characterized in that, the described step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises:

Generate the required multi-plot joint channel estimate matrix of multi-plot joint channel estimating, and utilize the multi-plot joint channel estimate matrix that generates to carry out pilot tone midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector.

17. associated detecting method according to claim 10 is characterized in that, the described step of utilizing selected N+M strong signal directly to carry out the multi-plot joint channel estimating comprises:

Generate the required multi-plot joint channel estimate matrix of multi-plot joint channel estimating, therefrom select L ' row vector to regenerate the multi-plot joint channel estimate matrix of simplification, utilize the multi-plot joint channel estimate matrix of described simplification to carry out midamble territory multi-plot joint channel estimating, obtain multi-plot joint CIR column vector, wherein, N+M≤L '≤L _Mid-W _{Max, c}+ 1, L _MidThe length of representing each midamble piece, W _{Max, c}The maximal possible length of expression channel impulse response window.

18. associated detecting method according to claim 10 is characterized in that, the described step of the matrix of generation system as a result of multi-plot joint channel estimating of utilizing comprises:

Upgrade corresponding community associated CIR according to the multi-plot joint channel estimation results;

Calculate the specific CIR of user equipment (UE) according to the community associated CIR after upgrading, and carry out the channel estimating reprocessing;

Calculate the CIR of particular virtual Resource Unit VRU according to the CIR of the corresponding particular UE after the channel estimating reprocessing;

Determine the quantity of the VRU that the participation multi-plot joint detects, and select the VRU of respective numbers;

Calculate the associating CIR of selected VRU in corresponding each local VRU and the special neighbourhood sub-district, and according to the associating CIR generation system matrix of the corresponding specific VRU that calculates.

19. associated detecting method according to claim 18 is characterized in that, further comprises after the generation system matrix:

From the sytem matrix that is generated, select L _AIndividual row vector generates new matrix, wherein, and L _AMore than or equal to the columns of described sytem matrix and be less than or equal to the line number of described sytem matrix, with newly-generated matrix as sytem matrix.

20. associated detecting method according to claim 10 is characterized in that, before the first and second data block signals according to sytem matrix that is generated and reception carry out the detection of data field multi-plot joint, further comprises:

Utilize the result of multi-plot joint channel estimating that the first and second data block signals that receive are handled, elimination in transmission course the midamble block signal to the interference of the described first and second data block signals, obtain the first and second clean data block signals, and utilize the first and second clean data block signals to carry out the data field multi-plot joint and detect.

21. associated detecting method according to claim 20 is characterized in that, described elimination midamble block signal in transmission course comprises the interference of the described first and second data block signals:

Estimate of the interference of this sub-district midamble sign indicating number, from the baseband signal of first and second data blocks that receive, deduct of the interference of this sub-district midamble sign indicating number respectively two data blocks to two data blocks;

Estimate of the interference of the midamble sign indicating number of neighbor cell, from the baseband signal of having eliminated first and second data blocks that this sub-district midamble sign indicating number disturbs, deduct of the interference of the midamble sign indicating number of neighbor cell respectively two data blocks to two data blocks.

22. a multi-plot joint checkout gear is characterized in that, comprising: pilot tone midamble territory multi-plot joint channel estimating unit and data field multi-plot joint detecting unit; Wherein,

Described midamble territory multi-plot joint channel estimating unit is used to estimate the channel impulse response CIR of each neighbor cell and this sub-district, CIR according to each neighbor cell, in neighbor cell, select N strong signal footpath in all multipath signals of each user, CIR according to this sub-district, from all multipath signals of each user of this sub-district, select M strong signal footpath, utilize selected N+M strong signal directly to carry out midamble territory multi-plot joint channel estimating again, wherein, N and M are natural number;

Described data field multi-plot joint detecting unit is used to utilize the midamble territory multi-plot joint channel estimation results generation system matrix of midamble territory multi-plot joint channel estimating unit output, and carry out the data field multi-plot joint according to the first and second data block signals of sytem matrix that is generated and reception and detect, obtain the first and second data block data carried by data.

23. multi-plot joint checkout gear according to claim 22, it is characterized in that described midamble territory multi-plot joint channel estimating unit comprises: the strong signal of neighbor cell directly selects module, the strong signal in this sub-district directly to select module, multi-plot joint channel estimation module and at least one channel estimation module; Wherein,

Each channel estimation module is respectively applied for the channel impulse response CIR that estimates designated cell in neighbor cell and this sub-district;

The strong signal of described neighbor cell directly selects module to be used for selecting N strong signals footpaths according to the CIR of the channel estimation module of corresponding neighbor cell output all multipath signals of all users in the neighbor cell;

The CIR that the strong signal in described this sub-district directly selects module to be used for exporting according to the channel estimation module of corresponding this sub-district selects M strong signal footpaths from all multipath signals of all users of this sub-district;

N+M the strong signal that described multi-plot joint channel estimation module is used for directly selecting module and the strong signal in this sub-district directly to select module to select according to the strong signal of neighbor cell directly carries out the multi-plot joint channel estimating.

24. multi-plot joint checkout gear according to claim 22 is characterized in that, described data field multi-plot joint detecting unit comprises: multi-plot joint sytem matrix generation module and least mean-square error MMSE detection module; Wherein,

Described multi-plot joint sytem matrix generation module is used for generating the required sytem matrix of data field multi-plot joint detection according to the multi-plot joint channel estimation results of multi-plot joint channel estimation module output;

Described MMSE detection module is used for the sytem matrix that generates according to multi-plot joint sytem matrix generation module and receives the first and second data block signals carrying out the multi-plot joint detection, obtains the first and second data block data.

25. multi-plot joint checkout gear according to claim 24 is characterized in that, described data field multi-plot joint detecting unit further comprises:

The midamble interference cancellation module, be used to eliminate midamble sign indicating number that this sub-district and neighbor cell use this sub-district is received the interference of two data blocks in the same burst packets, generate the receiving baseband signal of the first and second clean data blocks and output to described MMSE detection module.

Images