SCHEDULING METHOD FOR MULTI-USER MIMO

• US 20090154419A1
• Filed: 12/11/2008
• Published: 06/18/2009
• Est. Priority Date: 12/17/2007
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: YOSHIDA, SHOUSEI, KIMATA, Masayuki
• Granted Patent: US 8,254,328 B2
• Time in Patent Office: Days
• US Class Current: 370/330
• CPC Class Codes: H04L 25/022 of frequency response
• CPC Class Codes: H04L 5/0023 Time-frequency-space
• CPC Class Codes: H04L 5/0037 Inter-user or inter-terminal allocation
• CPC Class Codes: H04L 5/006 Quality of the received signal, e.g. BER, SNR, water filling
• CPC Class Codes: H04L 5/0089 due to addition or removal of users or terminals

Abstract

A priority computation process computes a priority of each user for each RB (Resource Brock) using a reception SINR. A maximum priority user selection/RB allocation process selects a user with the maximum priority for an unallocated RB and allocates the RB to the user. A frequency axis/space axis unallocated RB presence determination process proceeds to scheduling for a next user if there is an unallocated RB on the frequency or space axis. A projected channel vector update process updates a projected channel vector of an unselected user by GS orthogonalization. An orthogonal coefficient computation process computes an orthogonal coefficient. A corrected SINR computation process computes a corrected SINR. A next MIMO layer priority computation process computes priorities of unselected users for a corresponding RB in the next multiple MIMO layer. The priorities of the unselected users are used in the next user scheduling processing.

Claims

1. An MU-MIMO scheduling method for allocating frequency- and space-divided resource blocks to users, comprising:

allocating one or more resource blocks to a user with the maximum priority by determining the maximum priority used in allocation of resource blocks to users such that a priority based on reception Signal to Interference and Noise power Ratios for each resource block is used for the first multiple MIMO layer, and a priority based on corrected Signal to Interference and Noise power Ratios is used for the second and onward multiple MIMO layers;

computing a corrected Signal to Interference and Noise power Ratio of an unselected user for resource blocks in the multiple MIMO layer next to that of the allocated resource blocks;

computing a priority represented by a function of the corrected Signal to Interference and Noise power Ratio; and

allocating resource blocks to a next user by using priorities for unallocated resource blocks including the resource blocks for which the priority has been computed.

2. The MU-MIMO scheduling method as claimed in claim 1, wherein a Channel Quality Indicator is used as the reception Signal to Interference and Noise power Ratio.

3. The MU-MIMO scheduling method as claimed in claim 1, wherein resource blocks are allocated to users without posing any restriction on the order of allocation of resource blocks on the frequency axis and the space axis.

4. The MU-MIMO scheduling method as claimed in claim 1, wherein all the resource blocks on the frequency axis are allocated to users for each space axis.

5. The MU-MIMO scheduling method as claimed in claim 1, wherein the computation of the corrected Signal to Interference and Noise power Ratio is performed by updating a projected channel vector of an unselected user for an n-th user, computing an orthogonal coefficient of resource blocks of the next multiple MIMO layer based on the projected channel vector, and multiplying the reception Signal to Interference and Noise power Ratio by the orthogonal coefficient.

6. The MU-MIMO scheduling method as claimed in claim 5, wherein the update of the projected channel vector is performed by updating a projected channel vector on a complementary space of an orthonormal system corresponding to a user already selected by GS orthogonalization, for each predetermined frequency resolution.

7. The MU-MIMO scheduling method as claimed in claim 6, wherein the predetermined frequency resolution is of a subcarrier or a subcarrier group corresponding to a coherent bandwidth.

8. The MU-MIMO scheduling method as claimed in claim 5, wherein the update of the projected channel vector is performed by a method in which, when a projected channel vector of a subcarrier k of a user j in the m-th multiple MIMO layer is represented by h_j^(m)(k), where 1≦

k≦

K, 1≦

j≦

N_u, K is an average number of subcarriers, and N_u is a number of users, and a selected user in the m-th multiple MIMO layer is represented by J_m, a projected channel vector of the subcarrier k in the (m+1)-th multiple MIMO layer is updated so as to satisfy the following expression (11).

$\begin{array}{cc}{h}_j^{\left(m+1\right)}<br><br>\left(k\right)=h_j^{\left(m\right)}<br><br>\left(k\right)-\frac{\left(h_{J_m}^{{\left(m\right)}^H}<br><br>\left(k\right)<br><br>h_j^{\left(m\right)}<br><br>\left(k\right)\right)<br><br>h_{J_m}^{\left(m\right)}<br><br>\left(k\right)}{{<br><br>h_{J_m}^{\left(m\right)}<br><br>\left(k\right)<br><br>}^2}& \left(11\right)\end{array}$

9. The MU-MIMO scheduling method as claimed in claim 5, wherein the computation of the orthogonal coefficient is performed by a method in which power of the projected channel vector of unselected users is averaged or totalized over a predetermined frequency bandwidth, and is normalized with an averaged or totalized channel power over the same bandwidth.

10. The MU-MIMO scheduling method as claimed in claim 9, wherein the predetermined frequency bandwidth is a bandwidth of a resource block or of a single carrier.

11. The MU-MIMO scheduling method as claimed in claim 5, wherein the computation of orthogonal coefficient is performed by a method in which, when a projected channel vector of a subcarrier k of a user j in the (m+1)-th multiple MIMO layer is represented by h_j^(m+1)(k), and a channel vector of the subcarrier k of the user j is represented by h_j(k), an orthogonal coefficient ρ

_j^(m+1)(0≦

ρ

_j^(m+1)≦

1) of the user j in the (m+1)-th multiple MIMO layer is computed so as to satisfy the following expression (12).

$\begin{array}{cc}{\mathrm{\rho <br><br>}}_j^{\left(m+1\right)}=\frac{\frac{1}{K}<br><br>\underset{k=1}{\overset{K}{\sum <br><br>}}<br><br>{<br><br>h_j^{\left(m+1\right)}<br><br>\left(k\right)<br><br>}^2}{\frac{1}{K}<br><br>\underset{k=1}{\overset{K}{\sum <br><br>}}<br><br>{<br><br>h_j<br><br>\left(k\right)<br><br>}^2}& \left(12\right)\end{array}$

12. The MU-MIMO scheduling method as claimed in claim 1, wherein the computation of priority is performed by using a corrected Signal to Interference and Noise power Ratio as a priority.

13. The MU-MIMO scheduling method as claimed in claim 1, wherein the computation of priority is performed by computing a support data rate based on a corrected Signal to Interference and Noise power Ratio, and using the support data rate weighted with an inverse of an average support data rate as the priority.

14. The MU-MIMO scheduling method as claimed in claim 1, wherein the computation of priority is performed by computing a comprehensive priority by adding a priority of a request for delay or a request for retransmission to a priority represented by a function of the corrected Signal to Interference and Noise power Ratio.

15. The MU-MIMO scheduling method as claimed in claim 1, wherein users are narrowed down to those with high priority when selecting users for the first multiple MIMO layer, and users are selected from only those users for the second and onward multiple MIMO layers.

19. A base station using the scheduling method as claimed in claim 1 for allocating resource blocks or carriers to mobile stations or transmission devices.

20. A scheduler that carries out scheduling for allocating frequency- and space-divided resource blocks or carriers to mobile stations or transmission devices by using the MU-MIMO scheduling method as claimed in claim 1.

21. A program product that causes a base station to implement an operation for allocating frequency- and space-divided resource blocks or carriers to mobile stations or transmission devices by using the MU-MIMO scheduling method as claimed in claim 1.

16. A scheduling method for use in an MU-MIMO communication system allocating frequency- and space-divided carriers to users, the method being for performing scheduling for allocating carriers to users and comprising:

computing a priority of each user for each carrier based on a reception Signal to Interference and Noise power Ratio of each of resource blocks forming a carrier;

selecting a user with the maximum priority for an unallocated carrier;

allocating the unallocated carrier to the user with the maximum priority;

updating a projected channel vector of a carrier of an unselected user by GS orthogonalization to obtain an orthogonal coefficient if there is an unallocated carrier in the space axis direction, and computing a corrected Signal to Interference and Noise power Ratio by multiplying the reception Signal to Interference and Noise power Ratio of the unselected user by the orthogonal coefficient;

computing a priority of the unselected user of the carrier based on the computed corrected Signal to Interference and Noise power Ratio; and

allocating the unselected user with the maximum priority to the unallocated carrier in the space axis direction.

18. The scheduling method as claimed in claim 16, wherein each carrier allocated to the users is composed of a single resource block or an allowable number of resource blocks in the frequency axis direction.

17. A scheduling method for use in an MU-MIMO communication system allocating frequency- and space-divided carriers to users, the method being for performing scheduling for allocating carriers to users and comprising:

computing a priority of each user for each carrier based on a reception Signal to Interference and Noise power Ratio of each of resource blocks forming a carrier;

selecting a user with the maximum priority for an unallocated carrier;

allocating the unallocated carrier to the user with the maximum priority;

updating a projected channel vector of a carrier of an unselected user by GS orthogonalization to obtain an orthogonal coefficient if there is an allowable next multiple MIMO layer in the space axis direction and there is no unallocated carrier on the frequency axis, and computing a corrected Signal to Interference and Noise power Ratio by multiplying the reception Signal to Interference and Noise power Ratio of the unselected user by the orthogonal coefficient;

computing a priority of the unselected user for the carrier based on the computed corrected Signal to Interference and Noise power Ratio; and

allocating the unselected user with the maximum priority to the unallocated carrier in the next multiple MIMO layer.