CN101127548A - Non cooperative virtual MIMO technology - Google Patents

Abstract

The utility model discloses a non-cooperative virtual MIMO technology, which belongs to wireless communication field. Reasonable users are chosen to construct co-channel user combination; space-division multiplex of the co-channel users can be realized to increase total system capacity by times without requesting that direction communication ability is arranged between mobile terminals. For an uplink, an uplink co-channel signal can be separated through a minimum quadratic mean deviation algorithm, a zero forcing algorithm, a serial interference elimination algorithm, a maximum likelihood algorithm and other methods. For a downlink, the co-channel interference is obviously restrained for the receiving signal of each user through the downlink signal pretreatment technology; each user can individually receive signal treatment without other user cooperation to greatly simplify the receiver of the users. The non-cooperative virtual MIMO technology can improve total capacity of broadband wireless communication system by times through cooperating with orthogonal frequency division multiplexing technology. The non-cooperative virtual MIMO technology is more particularly suitable for the wireless communication system of division dual.

Description

Non cooperative virtual MIMO technology

Technical field

Technical field under the technical program is a wireless communication field.

Background technology

Multiple-input and multiple-output (MIMO) technology is subjected to broad research because of it can significantly improve message capacity.Yet generally portable mobile termianl (MT) can't be installed a plurality of antennas, makes the many antennas in single antenna MT and base station can't constitute real mimo system.The virtual MIMO technology of Ti Chuing utilizes the cooperation between a plurality of MT that the space faces mutually to constitute mimo system in recent years.Yet cooperative virtual MIMO technology requires must have direct communication capacity between cooperation MT, thereby has information security, system complex, is difficult to a series of problems such as realization.

The present invention proposes non cooperative virtual MIMO technology, its characteristics are: because the MT end does not carry out collaborative process, thereby do not require between common road MT to have direct communication capacity, overcome the shortcoming of cooperative virtual MIMO technology, but can realize common road user's space division multiplexing equally, overall system capacity is increased exponentially.For up link, select rational user to constitute common road user combination, at the multi-antenna base station receiving terminal,, a plurality of road users' altogether upward signal can be separated by Minimum Mean Square Error (MMSE) algorithm, ZF (ZF) algorithm, serial interference elimination (SIC) algorithm, maximum likelihood algorithm etc.For down link, select rational user to constitute common road user combination, by the downstream signal preprocessing technology that proposes, common road in each user's received signal disturbs (CCI) obviously to be suppressed, each user need not just can carry out the received signal processing separately by other user collaboration, makes receiver user obtain simplifying.Use with OFDM (OFDM) technical tie-up, the present invention can make the total capacity of system of broadband wireless communication be increased exponentially.The present invention is particularly useful for the wireless communication system of time division duplex (TDD).

Summary of the invention

The search technique of user's combination rationally in the virtual MIMO system has been proposed, the technical problem that solve:, make the non cooperative virtual MIMO system can obtain bigger capacity and the better error rate (BER) performance by selecting rational user's combination.Technical scheme: at first estimate the mimo channel matrix, the Generalized Inverse Matrix of calculating channel matrix.Can make the user of the Frobenius norm minimum of Generalized Inverse Matrix be combined as the optimal user combination, the less user of Frobenius norm of Generalized Inverse Matrix is combined as reasonable user's combination.

Proposed the preprocessing technology of virtual MIMO system downstream signal, the technical problem that solve: do not have between MT under the situation of direct communication ability, each user can't utilize other user's received signal to carry out collaborative process.The downstream signal preprocessing technology that proposes can make the CCI in each user's received signal obviously be suppressed, and each user need not cooperate just can carry out the received signal processing separately.Technical scheme: at first estimate the down channel matrix, constitute column vector with the downstream signal that is total to the road user, Generalized Inverse Matrix with the down channel matrix multiply by this column vector, the downstream signal column vector after obtaining anticipating, emission respectively on a plurality of antennas in base station accordingly.

Embodiment

Represent common road number of users in the non cooperative virtual MIMO system with M, represent antenna for base station quantity with N.In order to make receiving terminal can separate M road subscriber signal altogether effectively, require N 〉=M.

The search technique of up reasonable user's combination of narrow band channel.At first the up channel matrix H of N * M is estimated that the Generalized Inverse Matrix V of calculating channel matrix H can make the user of the Frobenius norm minimum of V be combined as the optimal user combination, the less user of Frobenius norm of V is combined as reasonable user's combination.According to the ZF criterion, the Generalized Inverse Matrix of H is provided by following algorithm:

V ^zf＝(H ^HH) ^-1H ^H (1)

According to the MMSE criterion, the Generalized Inverse Matrix of H also can be provided by following algorithm:

V ^mmse＝(H ^HH+σ ²I _M) ^-1H ^H (2)

σ wherein ²Expression base station normalization received noise power, I _MBe M * M unit matrix, subscript ^HThe conjugate transpose computing of representing matrix.The search algorithm of optimal user combination is as follows:

$H^{\mathrm{opt}}=\arg \underset{H}{\min }\sqrt{\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{\left|v_{\mathrm{ij}}\right|}^2}---\left(3\right)$

V wherein _IjThe value of the capable j row of i of expression Generalized Inverse Matrix V.Can make

Less user is combined as reasonable up user's combination.

The isolation technics of the common road upward signal of narrow band channel.M the estimated value that the user transmits at any one time

${\hat{x}}_m=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{v}_{\mathrm{mi}}{y}_i,m=\mathrm{1,2},\·\·\·M---\left(4\right)$

Y wherein _iThe normalization received signal value of expression base station i antenna.Formula (1) result of calculation substitution formula (4) with reasonable up user's combination is the ZF algorithm, and formula (2) the result of calculation substitution formula (4) that makes up with reasonable up user is the MMSE algorithm.

The search technique of descending reasonable user's combination of narrow band channel.At first the down channel matrix G to M * N estimates, the Generalized Inverse Matrix W of calculating channel matrix G can make the user of the Frobenius norm minimum of W be combined as the optimal user combination, and the less user of Frobenius norm of W is combined as reasonable user's combination.According to the ZF criterion, the Generalized Inverse Matrix of G is provided by following algorithm:

W ^zf＝G ^H(GG ^H) ^-1(5)

According to the MMSE criterion, the Generalized Inverse Matrix of G also can be provided by following algorithm:

W ^mmse＝G ^H(GG ^H+σ ²I _M) ^-1(6)

σ wherein ²Expression MT normalization received noise power.The search algorithm of optimal user combination is as follows:

$G^{\mathrm{opt}}=\arg \underset{G}{\min }\sqrt{\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}{\left|w_{\mathrm{ij}}\right|}^2}---\left(7\right)$

W wherein _IjThe value of the capable j row of i of expression Generalized Inverse Matrix W.Can make Less user is combined as reasonable downlink user combination.

For the wireless communication system of TDD, if up user combination is identical with the downlink user combination, then up channel matrix and down channel matrix have following relation:

G＝H ^T (8)

Subscript wherein ^TThe transposition computing of representing matrix.If down channel matrix G need reappraise, can obtain by formula (8).

The preprocessing technology of the common road downstream signal of narrow band channel.N antenna for base station transmitting at any one time

$x_n=\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{w}_{\mathrm{ni}}{s}_i,n=\mathrm{1,2},\·\·\·,N---\left(9\right)$

S wherein _iRepresent that i is total to the downstream signal value that the road user will launch.Formula (5) result of calculation substitution formula (9) with reasonable downlink user combination is the ZF algorithm, and formula (6) the result of calculation substitution formula (9) that makes up with reasonable downlink user is the MMSE algorithm.

Below provide the theory analysis of downstream signal preprocessing technology.Know by formula (9), anticipate the emission column vector of back antenna for base station

x＝[x ₁，x ₂，…，x _N] ^T＝Ws＝W[s ₁，s ₂，…，s _M] ^T (10)

Road user's normalization receives column vector altogether

y＝[y ₁，y ₂，…，y _M] ^T＝Gx+n＝GWs+n (11)

N=[n wherein ₁, n ₂..., n _M] ^TExpression normalization receives the noise column vector.With formula (5) substitution formula (11),

y＝GG ^H(GG ^H) ^-1s+n＝s+n (12)

Formula (12) illustrates that the preprocessing technology of downstream signal makes each common road user's received signal y _m, m=1,2 ..., only comprise self downstream signal s among the M _mWith interchannel noise n _m, and other common road user's downstream signal has been suppressed.Formula (6) substitution formula (11) is also obtained same conclusion.

The search technique of up reasonable user's combination of broad-band channel.Combine with the OFDM technology, non cooperative virtual MIMO technology can work in broad-band channel.The MIMO-OFDM system all constitutes an arrowband mimo system on each subcarrier.At first to the up channel matrix H (k) of a plurality of N * M, k=1,2 ..., K estimates that K represents the number of subcarriers of OFDM.The Generalized Inverse Matrix V (k) of calculating channel matrix H (k).According to the ZF criterion, the Generalized Inverse Matrix of H (k) is provided by following algorithm:

V ^zf(k)＝(H ^H(k)H(k)) ^-1H ^H(k)，k＝1，2，…，K (13)

According to the MMSE criterion, the Generalized Inverse Matrix of H (k) also can be provided by following algorithm:

V ^mmse(k)＝(H ^H(k)H(k)+σ ²I _M) ^-1H ^H(k)，k＝1，2，…，K (14)

The search algorithm that optimal user carefully closes is as follows:

$H^{\mathrm{opt}}\left(k\right)=\arg \underset{H\left(k\right)}{\min }\sqrt{\underset{k=1}{\overset{K}{\mathrm{\Σ}}}\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\underset{j=1}{\overset{N}{\mathrm{\Σ}}}{\left|v_{\mathrm{ij}}\left(k\right)\right|}^2}---\left(15\right)$

V wherein _Ij(k) value of the capable j row of the i of expression Generalized Inverse Matrix V (k).Can make

Less user is combined as reasonable up user's combination.

The isolation technics of the common road upward signal of broad-band channel.The estimated value that m user k subcarrier transmits

${\hat{x}}_m\left(k\right)=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{v}_{\mathrm{mi}}\left(k\right)y_i\left(k\right),m=\mathrm{1,2},\·\·\·M---\left(16\right)$

Y wherein _i(k) the normalization received signal value of expression base station i antenna on the k subcarrier.Formula (13) result of calculation substitution formula (16) with reasonable downlink user combination is the ZF algorithm, and formula (14) the result of calculation substitution formula (16) that makes up with reasonable downlink user is the MMSE algorithm.

The search technique of descending reasonable user's combination of broad-band channel.At first to the down channel matrix G (k) of a plurality of M * N, k=1,2 ..., K estimates, the Generalized Inverse Matrix W (k) of calculating channel matrix G (k).According to the ZF criterion, the Generalized Inverse Matrix of G (k) is provided by following algorithm:

W ^zf(k)＝G ^H(k)(G(k)G ^H(k)) ^-1，k＝1，2，…，K (17)

According to the MMSE criterion, the Generalized Inverse Matrix of G (k) also can be provided by following algorithm:

W ^mmse(k)＝G ^H(k)(G(k)G ^H(k)+σ ²I _M)-1，k＝1，2，…K (18)

σ wherein ²Expression MT normalization received noise power.The search algorithm of optimal user combination is as follows:

$G^{\mathrm{opt}}\left(k\right)=\arg \underset{G\left(k\right)}{\min }\sqrt{\underset{k=1}{\overset{K}{\mathrm{\Σ}}}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}{\left|w_{\mathrm{ij}}\left(k\right)\right|}^2}---\left(19\right)$

W wherein ^Ij(k) value of the capable j row of the i of expression Generalized Inverse Matrix W (k).Can make

Less user is combined as reasonable downlink user combination.

For the wireless communication system of TDD, if up user combination is identical with the downlink user combination, then up channel matrix and down channel matrix have following relation:

G(k)＝H ^T(k)，k＝1，2，…K (20)

If down channel matrix G (k) need reappraise, can obtain by formula (20).

The preprocessing technology of the common road downstream signal of broad-band channel.N antenna for base station transmitting on the k subcarrier

$x_n\left(k\right)=\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{w}_{\mathrm{ni}}\left(k\right)s_i\left(k\right),n=\mathrm{1,2},\·\·\·,N---\left(21\right)$

S wherein _i(k) expression i is total to the downstream signal value that the road user will launch on the k subcarrier.Formula (17) result of calculation substitution formula (21) with reasonable downlink user combination is the ZF algorithm, and formula (18) the result of calculation substitution formula (21) that makes up with reasonable downlink user is the MMSE algorithm.

Claims (6)

1. non cooperative virtual MIMO technology.With the immediate prior art of this theme: have direct communication capacity between the cooperation portable terminal (MT) that cooperative virtual multiple-input and multiple-output (MIMO) technology, this specification requirement space are faced mutually.This theme is characterised in that: because the MT end does not carry out collaborative process, thereby do not require between common road MT to have direct communication capacity, overcome the shortcoming of cooperative virtual MIMO technology, but can realize common road user's space division multiplexing equally, the total capacity of system is improved.For up link, select rational user to constitute common road user combination, at the multi-antenna base station receiving terminal,, a plurality of road users' altogether upward signal can be separated by ZF (ZF) algorithm, Minimum Mean Square Error (MMSE) algorithm, serial interference elimination (SIC) algorithm, maximum likelihood algorithm etc.For down link, select rational user to constitute common road user combination, by the downstream signal preprocessing technology that proposes, the common road in each common road user's received signal disturbs (CCI) obviously to be suppressed, and the cooperation that each user need not other user just can be carried out received signal separately and handle.Use with OFDM (OFDM) technical tie-up, non cooperative virtual MIMO technology can make the total capacity of system of broadband wireless communication be improved.Non cooperative virtual MIMO technology is specially adapted to the wireless communication system of time division duplex (TDD).

2. quote the non cooperative virtual MIMO technology described in 1.Its additional technical characterictic is: the search technique that has provided user's combination rationally in the virtual MIMO system.At first estimate the mimo channel matrix, the Generalized Inverse Matrix of calculating channel matrix.Can make the user of the Frobenius norm minimum of Generalized Inverse Matrix be combined as the optimal user combination, can make the less user of Frobenius norm of Generalized Inverse Matrix be combined as reasonable user's combination.

3. quote the non cooperative virtual MIMO technology described in 2.Its additional technical characterictic is: the search technique that has provided up reasonable user's combination in the virtual MIMO system.Represent up altogether road number of users in the non cooperative virtual MIMO system with M, represent antenna for base station quantity with N.Under the arrowband situation, can make

Less user is combined as reasonable up user's combination.V wherein _IjThe value of the capable j row of i of expression Generalized Inverse Matrix V, Generalized Inverse Matrix V=(H ^HH) ^-1H ^HPerhaps V=(H ^HH+ σ ²I _M) ^-1H ^H, H is the estimation of up channel matrix, σ ²Expression base station normalization received noise power, I _MBe M * M unit matrix, subscript ^HThe conjugate transpose computing of representing matrix.Under wideband scenarios, can make

Less user is combined as reasonable up user's combination, and wherein K represents the number of subcarriers of OFDM, v _Ij(k) value of the capable j row of the i of expression Generalized Inverse Matrix V (k), Generalized Inverse Matrix V (k)=(H ^H(k) H (k)) ^-1H ^H(k), k=1,2 ..., K or V (k)=(H ^H(k) H (k)+σ ²I _M) ^-1H ^H(k), k=1,2 ..., K, H (k) is the estimation of up channel matrix.

4. quote the non cooperative virtual MIMO technology described in 2.Its additional technical characterictic is: the search technique that has provided descending reasonable user's combination in the virtual MIMO system.Represent descending altogether road number of users in the non cooperative virtual MIMO system with M, represent antenna for base station quantity with N.Under the arrowband situation, can make

Less user is combined as reasonable downlink user combination.W wherein _IjThe value of the capable j row of i of expression Generalized Inverse Matrix W, Generalized Inverse Matrix W=G ^H(GG ^H) ^-1Perhaps W=G ^H(GG ^H+ σ ²I _M) ^-1, G is the estimation of down channel matrix, σ ²Expression MT normalization received noise power.For the wireless communication system of TDD, if up user combination is identical with the downlink user combination, then up channel matrix and down channel matrix have following relation: G=H ^T, if down channel matrix G need reappraise, can be by G=H ^TRelation obtains.Under wideband scenarios, can make

Less user is combined as reasonable downlink user combination, w _Ij(k) value of the capable j row of the i of expression Generalized Inverse Matrix W (k), Generalized Inverse Matrix W (k)=G ^H(k) (G (k) G ^H(k)) ^-1, k=1,2 ..., K or W (k)=G ^H(k) (G (k) G ^H(k)+σ ²I _M) ^-1, k=1,2 ... K, G (k) is the estimation of down channel matrix.For the wireless communication system of TDD, if up user combination is identical with the downlink user combination, then up channel matrix and down channel matrix have following relation: G (k)=H ^T(k), k=1,2 ... K, if down channel matrix G (k) need reappraise, can be by G (k)=H ^T(k), k=1,2 ... K concerns acquisition.

5. quote the non cooperative virtual MIMO technology described in 1.Its additional technical characterictic is: the preprocessing technology that has provided common road downstream signal.At first estimate the down channel matrix, constitute column vector with road user's downstream signal altogether, multiply by this column vector with the Generalized Inverse Matrix of down channel matrix, the downstream signal column vector after obtaining anticipating is launched respectively on a plurality of antennas in base station accordingly.

6. quote the non cooperative virtual MIMO technology described in 5.Its additional technical characterictic is: under the arrowband situation, and n antenna for base station transmitting at any one time $x_n=\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{w}_{\mathrm{ni}}{s}_i,$ n＝1，2，...，N。S wherein _iRepresent that i is total to the downstream signal value that the road user will launch, w _IjThe value of the capable j row of i of expression Generalized Inverse Matrix W, Generalized Inverse Matrix W=G ^H(GG ^H) ^-1Perhaps W=G ^H(GG ^H+ σ ²I _M) ^-1, G is the estimation of down channel matrix.Under wideband scenarios, n antenna for base station transmitting on the k subcarrier $x_n\left(k\right)=\underset{i=1}{\overset{M}{\mathrm{\Σ}}}{w}_{\mathrm{ni}}\left(k\right)s_i\left(k\right),n=\mathrm{1,2},\·\·\·,N.$ S wherein _i(k) expression i is total to the downstream signal value that the road user will launch, w on the k subcarrier _Ij(k) value of the capable j row of the i of expression Generalized Inverse Matrix W (k), Generalized Inverse Matrix W (k)=G ^H(k) (G (k) G ^H(k)) ^-1, k=1,2 ..., K or W (k)=G ^H(k) (G (k) G ^H(k)+σ ²I _M) ^-1, k=1,2 ... K, G (k) is the estimation of down channel matrix.