RELATED APPLICATION INFORMATION

[0001] This applicatio , claims priority to provisional application serial number 61/883,548 filed on September 27, 2013, incorporated herein by reference.

BACKGROUND

Technical Field

[0002] The present inverrtion relates to signal processing, and more particularly to two- stage preceding and user grouping for large scale multiple-input luuldpfe-output (M1MO) systems.

Description of the R a ed Art

[0003] Most of the existing work directed to large scale 1MO (multi-iupttt-multi - output) systems considers a Time Division Duplex (TDD) mode. However, there are as many FDD (Frequency Division Duplex) systems as TDD system in real-world deployments. To equip a large amount of antennas m the base station (BS _. ) of a FDD system, a two-stage preceding scheme has been proposed.. In particular, all the users in the system are divided into several groups. The first stage preceding is designed to suppress the interferences among different groups. The second stage preceding is designed to. suppress the interferences among different users within each group. Given the preceding methods, bow to group users has a direct impact on the throughput performance. Fixing the number of groups, the existing grouping method may group too marr users in some particular groups and group none or too few users to the other groups. Also, with groups being formed, scheduling which users of each, group are to transmit and which users are to r mai silent also has big impacts on (he throughput. As is evident,, sueh an approach is not a sufficient way lo utilize the precious spectrum resources,

[0004] These and other dra wbacks and disadvantages of the prior art are addressed, by th present principles, which are directed to two-stage preceding and user grouping for large scale multiple-input multiple-output (MIM ^' O) systems.

[0005] According to an aspect of the present princip es, a method, is pro vided. The method, includes determining respective weighted likelihoods corresponding to a plurality of users In. a nndtiple-input ma!iiple-output communication system. The method further includes forming a plurality of user groups from the plurality of users using an iterative K-means clustering technique applied to the plurality of users. The iterative -means clustering technique is responsive to the respective weighted likelihoods.

[0006] According to another aspect of the resent principles, a method is provided. The method includes calculating respective rate gains for each of a. plurality of users responsive to respectively adding each of the plurality of users to a given time slot in a multi le- input multiple-output communication system. The method further includes finding a maximum rate gain from among the respective rate gains. The method additionally includes scheduling a given one of the plurality of users with the maximum, rate gain for (he g en, time slot. [0007] These and other features and ^' advantages will, become appa ent from the following detailed description of illustrative embodiments thereof, which is to be connection with the accompanying drawings.

BRIEF DESCRIPTION OF DRA INGS

[0008] The disclosure will provide details in the .following descriptio of preferred embodiments with reference to the .following figures wherein:

[0009] FIG. 1. shows an exemplary processing system 1.00 to which the present principles may be applied, in accordance with an embodiment of the present principles;

[0010] FIG. 2 shows an exemplary system 200 for preceding, user grouping, and user scheduling in a large scale multiple-input multiple-output IMO) communication system, in accordance with an embodiment of he present principles;

[0011 J FIG. 3 shows an exemplar method 300 for assigning users to difference groups in a multiple-input mulu le-ouiput (M1MO) communication system using K-nreans clustering, in. accordance with an embodiment of the present principles;

[0012] PIG. 4 shows an exemplary method 400 for user scheduling in a multiple-input multiple-output ( ΪΜΟ) communication system using -means clustering, in accordance with an embodiment of the present principles; and

[0013] HO. 5 sho ws an exemplary method 500 for user grouping with joint group load balancing and preceding design, in a multiple-inpat multiple-output (M!MOj

communicat o system using -means: clustering, in accordance with an embodiment of the present principles.

[00 ί 4] The. present principles, are directed to two-stage preceding and user grouping for large scale ip le-in t multiple-ontpnt (MMO) systems. That is. various

embodiments of the present principles are directed to- user grouping, user scheduling, and load balancing problems relating to large scale MEMO systems.

] 0015] in an embodiment, we propose several nser grouping methods to achieve higher throughput or better load halancing. For example, in an embodiment, we propose an improved Knneaos clustering algorithm for general user grouping. In an embodiment, we also propose a user grouping algorithm with joint load balancing and preceding (prebeaniform ng) design.

[0016] ¾. an em odimen , we also propose a dynamic user scheduling algorithm to enhance the system throughput. In an embodiment, the proposed dynamic user scheduling algorithm Is a greedy algorithm, which achieves a local maximum at each step. Thus, a high throughput can be expected.

[001?] The proposed improved K-nreans clustering algorithm and dynamic user schedulin algorithm advantageously achieve higher system throughput than existing approaches. In an embodiment, the proposed improved K-means clustering algorithm considers the weights of difFereut Eigen odes, which brings about better classification of different users. The proposed loading balancing algorithm advantageousl solves the problem of unbalanced group loading for large scale 1MO system.

[0018] FIX). I shows an exemplary processing system 1.00 to which the present, principles may be applied, according to an embodiment of the present principles. Is shown. The processing system 100 includes at least one processor (CPU) 104 operative I y coupled to other components via a. system bus 102, A cache 106, a Read Only Memory (ROM) 108, a Random Access Memory (RAM) .1 10, an input/output (I/O) adapter 120, a sound adapter 130, a network adapter 140, a user interface adapter 150, and. a display adapter 160, are operatively coupled to the s stem bus 102,

10019] A first storage device .122 and a second storage device 124 are operatively coupled, to system bus 102 by the /O adapter 120. The storag devices 122 and 124 can he any of a disk storage device (e.g.,, a magnetic or optical disk storage device), a solid state magnetic device, and so forth. The storage devices 122 aud 124 ca be the same type of storage device or diffe en types of storage devices.

[0020] A speaker 132 is operative coupled to system bus 102 by the sound adapte 130. A transceiver 142 is operatively coupled to system bus 1.02 by network adapter 140, A display device 1.62 is operatively coupled to system bus 102 by display adapter 160.

[002 I] A ^' first user input device 152, a second user input device J 54, and a third user input device 156 are operatively coupled to system bus 102 by user interface adapter 150. The user input devices 1 52, .154, and 1 6 can be any of a keyboard., a mouse, a keypad, an image capture device, a motion sensing device, a microphone, a device incorporating the functionality of at least two of the preceding devices, and so forth. Of course, other types of input devices can also be used, while maintaining the spirit of the present principles. The. user input devices 152, 154, and 1 56 can be the same type of user input device or different types of user input devices. The user input devices 152, 1 54» and 156 are used to input and output, information to and from system 100.

|O022J Of course, the processing system 100 may also include other elements . ^' not shown);, s readily contemplated by one of skill in the art, as wed as omit certain elements, For example, various other input devices and/or out ut devices cm be included in processing system 100, depending upon the particular implementation of the same, as readily understood by one of ordinary skill in the art. For example, various types of wireless and/or wired input and/or output devices can be used.. Moreover _* additional processors, controllers, memories, and so forth, its various co figurations can also be utilized, as readily appreciated by one of ordinary skill in the art. These and other variations of the processing system 1 0 are readily contemplated by one of ordinary skill in the art given the teachings of the present principles provided herein.

[0023] Moreover, it is to be appreciated that system 200 described below wit respect to PIG, 2 is a system for implementing respective embodiments of the present principles. Pari or all of processing system 100 may be implemented in one or more of the elements of system 200.

[0024] Further, it is to be appreciated that processing system 100 may perform at least part of the method described, herein including, for example, at least part of any of methods 300-700 of F!Gs.. 3-7. Simihrly, part or all of system 200 ma be used to perform at least pari of any of methods 300-700 of FIGs. 3-7.

[0025] FIG. .2 shows an exemplary system 200 for preceding, user grouping, and: user scheduling in a large scale multiple-input multiple-output (MMO) communication system, in. accordance with an embodiment of the present principles,

[0026] The system 200 includes a -means user elnsterer 21 , a dynamic user scheduler 220, and a joint load balancer and preeoder 230, at! interconnected by a bus 266. The -means user elnsterer 210 determines user groups. The dynamic user scheduler 220 determines use schedules. The joint load balancer and preeoder 230 determines user groups. In an embodiment, at least the K-means elusterer 21.0 performs method 300 of F 1(5. 3, at ast the ynamic user scheduler 220 performs method 400 of FIG. 4, arsd. at least the joint load balancer and precoder 230 performs method 500 of FIG. 5.

[0027] FIG, 3 shows an. exemplary method. 300 for assigning users to difference groups in a .multiple -input multiple-output (Ml ^' O) coin muni cation syst m using K-means clustering, in accordance with an. embodiment of the present principles, Method 300 can. be considered to correspond to algorithm 1 described herein,

[00.28] Method. 300 uses wei hted likelihood as a criterion to assign users. The weighted likelihood takes the weight of differ n Eigenmodes into consideration. The update of the group center and total likelihood also considers the weights of different Eigenmodes,

[0029] At step 301, find the eovariance matrix tor each, user, and compute ¾ using Ei en-decomposition ¾ ^::; PA-AUiA.

l ^' 0030 ^' j At step 302, for each gr p, randomly pick U* as the group center V _g .

[00311 At ste 303 , find the weighted likelihood L(R _k , V ,} defined in liquation i l ).

[0032] At step 304, assign the user to the group with the maximum weighted likelihood using Equation (2),

[0033] At step 305, update the group center V _e using Equation (3),

[0034 At step 306, compute the likelihood of eac user to its assigned group. Add the

Likelihood as the total iikelihood,

[0035] At step 3 7, determine whether or not the likelihood converges. If so, then the method terminates. Otherwise, the method returns to step 303. The following equations apply to method 300:

L( _fe /¾) - !! (i ,A ^¾ ¾. (1 ) g _k * ^ ar m ;¾L(R _fci ¾) (2)

where etg denotes the unitary matrix after Eigeu-decomposltion, ¾ -denotes the member set of r u g, and \S \ denotes the num er of group members in group g.

[0037] FIG, 4 shows an exemplary method 400 for user scheduling in. a n ltip!e-inpnt multiple-output (MIMO) comn nieaiion system using -means clustering, in accordance with, an embodiment of the present principles. Method 400 can he considered to correspond to algorithm 2 described herein,

[0038] Method 400 computes the throughput gains of dding each user given the set of users already scheduled (initially empty), and then finds the maximum one amon them. Each time, the user with, the maximum throughput gain is added, to the set of users already scheduled. By doing this repeatedly, we can Una the set of users to schedule. | ^' 0039j At ste 401 , initially, define the user set as lf~{ L K}, where the group member set S _g is empty for ail g.

[00401 At step 402, determine whether or not the total number of scheduled users in a! I groups is greater than or equal to the total number of beams of all groups, If so, then the method is terminated. Otherwise, the method proceeds to step 403,

[0041] At step 403, determine -whether or not U is empty . If so, then the method is terminated. Otherwise, the method continues to step 404,

[0042] At step 404. begin a. Loop A, for each unscheduled user. [0043] A step 405, pick one user from U„ and assume it is scheduled in its assigned group.

[0044] At step 06, perform zero _, forcing esmJ nning or regularized zero forcing beam&rmirrg given ifce group members already scheduled and the prebeamfbrming matrix rbr the user grouping.

[0045] At step 407. compute the rate gain K(k g) as the difference of throughput with and without having user k scheduled.

[0046] At step 408, cud Loop A.

[0047] At step 409, f nd the maximum of K(k, g) denoted as (k\ g

[0048] At step 10, determine whether or not K(k\ g ^* ) >0.

[0049] At step 41 1 , schedule the user with the maximum ra e gain K( add k* into S _8* , and eliminate k* from U.

[0050] It is to be appreciated that steps 405, 406, and 407 are repeated for ail unscheduled users.

[0051 ] FIG. 5 shows an exemplary method 500 far user grouping with joint group load balancing and preceding design in a multiple-input multiple-output (MIMO)

communication system usifig K-meaas clustering, in accordance with m embodiment of the present principles. Method 500 can be considered to correspond to algorithm 3 described herein.

[0052] At. step 501 , Hud the ovariance matrix for each user, and perform Eigen- decomposition to compute L as ¾ ^~ TJ^A ^-dk ·

[0053] At step 502, use K-means or improved K--means grouping algorithm to find the user association with each group. [0054] At step 503, update the ^' group center V _g using Equation (3) or (5).

[0055] At step 504, find, pre-beanrforramg matrices.

[0056] At step 505, approximate the average SI Ri ^' k, g) using Equation (4) or the large scale approximation method described herein. In an embodiment the large scale approximation .method is described with respect to algorithm 4. Of course, the present principles are not limited to solely the large scale approximation method described herein and, thus, other large scale approximation methods can also be used while maintaining the spirit of the present principles.

[005?] At step 506, optimize the utility using algorithm in Table 2,

[00581 At step 507, determine, whether or not the utility has converg d, if so, then, the method is terminated. Otherwise, the method returns to step 503.

[0059] The following equations apply to method 500:

where SINR. is defined as the Signal, to Interference pins Noise Ratio,

[0060] V _p - eOo{™∑ _½¾ iU4 ^i; ) (5) (For K-means Glittering Algorithm), where eig denotes the unitary matrix, after eigen-deeompostlon S _¾ denotes the member set of grou g, and. I I -denotes the number of group members in group g.

[0061] TABLE 1 shows various notions applicable to the present principles. ID accordance with m embodiment of the present principles. TABLE !

I LHm Hsiou Meas sg

K x 1 Sie;nals received bv all users yg .¾ X I gnals recei ed, by group g d 5 x i Signals transmitted for ail users

4? ¾ x 1 Signals transmitted for group g

M 1 x 1 Number of antennas at the BS iY M x K Channel, between BS and users e b x K Effective chasms! between BS

I 1 Number of da a stream

V M x 5 Preceding matrix

M X Covariance matrix for user k x l Instantaneous channel t or user k

B x s Premeamibrmkm matrix

P 1? x S Precoding matrix

x Unitary matrix composed of x Diagonal matrix composed of

M X Af Group center of group g

X dominant, eigenvalues for group

% Premea rbrming matrix for

% ¾ x ¾ Precoding matrix within each.

.. K X 1 No se vector

¾ ^' ¾ x I Noise vector within each group i? 1 X 1 Azimuth angie of user

,s 1 X 1. Distance between BS and user

Λ I x 1 Angle spread

r 1 X 1 Radius of scattering ring

G 1 X 1 Number of groups

a i x 1 Parameter for RZFBF precoding

P 1 x 1 Total transmit power of BS

.JiL i 1 X J: [0062] Algorithm 1 ; Improved K-means Clustering Algorithm

and

[00631 Algorithm 2; Greedy algorithm, tor d namic user selection, and beamormng with determined user grouping

! User grouping g is given:

2 Initially set f£ ~ {!, K}, the weighted um rate r; _w 0 and ¾ ' - 0 for g-L , Q:

¾ White Ts ion conditions (∑ _g |¾| 0 _S or U - 0) ?¾ wi ■satisfied do ≠ 5 ;

j and {.¾) :

({S' _e ) _t {¾}) - t ((¾X (¾})} i

V. ml

is End

[0064} Algorithm 3 : User grouping with joint group load balancing and precedi ng design algorithm

1 Perform Algorithm 1 to obtain user group I ^' D d likelihood (22)

t s end

[0065 J Algorithm 4; Large Scale SiMR Approximation

n end

ii end [0066] Embodiments described herein may he entirely hardware, ntirel software or including both hardware and software elements, In. a preferred embodiment, the present invention is implemented in. software, which includes but is not limited to firmware, residen software, microcode, etc,

j0067] Embodiments may include a computer program product accessible from a. computer-usable or computer-readable medium providing program code for use by or in connection with a computer or any instruction execution system, A compute -usable or computer readable medium may include auy apparatus that stores, communicates, propagates, or transports the program lor use by or in connection with the instruction, execution system, apparatus, or device. The medium can be magnetic,, optical, electronic, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium. The medium may include a computer-readable medium such as a semiconductor or solid state memory, magnetic tape, a removable. computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disk and an optical, disk, etc.

[0068] it is to be appreciated that the use of any of the following "/",. "and/or", and "at least one of, for example, in the eases of Bf ^;i A and/or B" and " least one of A. and " is intended to encompass the selection, of the first fisted option (A) only, or the seiectiou of the second listed option (B) only, or the selection of both options (A and B). As a further example, in the cases of "A, B, and/or C" and "at least one of A, B, and C ^* \ such phrasing Is Intended to encompass the selection of the first listed option (A.) only, or the selection of the second listed option (B) only, or the selection of die third listed option (C) only, or the selection of the first and the second listed options (A. and 13} only, or the selection of the first mi third listed options (A and C) only, ox the selection of the second and third listed options (B and C) only, or the selection, of all three options (A and B and C This may be extended, as readily apparent by one of ordinary skill, m tltis and related arts, for as many items listed.

[0069] The foregoing is to understood as being in every respect iilustraiive a exemplary, bin not restrictive, and the scope of the invention disclosed herein is not to be determined from the Detailed. Description, but rather from the claims as interpreted according to the fell breadth permitted by the patent laws. Additional, information is provided in. an appendix, to the application entitled, "Additional Information". It is to be understood that the embodiments shown, and described herein are only illustrative of the principles of the present invention and that those skilled in the art may Implement various modifications without departing from the scope and spirit of the invention, Those skilled in the art could impkmeM various other feature combinations without departing from the scope and spirit of the invention.