WO2015067200A1 - 多输入输出系统的导频调度方法及协同设备 - Google Patents

多输入输出系统的导频调度方法及协同设备 Download PDF

Info

Publication number
WO2015067200A1
WO2015067200A1 PCT/CN2014/090531 CN2014090531W WO2015067200A1 WO 2015067200 A1 WO2015067200 A1 WO 2015067200A1 CN 2014090531 W CN2014090531 W CN 2014090531W WO 2015067200 A1 WO2015067200 A1 WO 2015067200A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
cells
sequence
rate
search
Prior art date
Application number
PCT/CN2014/090531
Other languages
English (en)
French (fr)
Inventor
金石
李明梅
杜颖钢
高西奇
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2015067200A1 publication Critical patent/WO2015067200A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04J—MULTIPLEX COMMUNICATION
    • H04J13/00—Code division multiplex systems
    • H04J13/16—Code allocation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L5/00—Arrangements affording multiple use of the transmission path
    • H04L5/003—Arrangements for allocating sub-channels of the transmission path
    • H04L5/0048—Allocation of pilot signals, i.e. of signals known to the receiver
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W72/00—Local resource management
    • H04W72/12—Wireless traffic scheduling

Definitions

  • Embodiments of the present invention relate to the field of communications, and in particular, to a pilot scheduling method and a cooperative device of a multiple input/output system.
  • Massive MIMO Very Large MIMO or Massive MIMO
  • the feedback amount increases linearly with the number of antennas.
  • the number of antennas at the base station increases, when the number of antennas is large, the time required for feedback will be large.
  • channel state information is obtained mainly by utilizing channel reciprocity in a massive MIMO system.
  • pilot pollution since the dimension of the pilot signal space is always limited, it is inevitable that users of different cells always transmit simultaneously using the same pilot, thereby causing the base station to be indistinguishable, forming so-called "pilot pollution".
  • pilot pollution When the base station does not cooperate, as the number of base station antennas increases indefinitely, the uncorrelated noise and fast fading effects can be averaged out.
  • the system performance is mainly caused by inter-cell interference caused by pilot pollution, and whether it is uplink or not In the downlink, the equivalent signal to interference and noise ratio is only related to the large-scale fading factor; and when there is pilot pollution, increasing the transmit power of the uplink pilot has no meaning for improving the channel estimation performance.
  • a Massive MIMO multi-cell multi-user system considering no cooperation between base stations, including L cells, each cell containing K single-antenna users (multi-antenna users can be regarded as multiple single-antenna users), performing network-wide frequency reuse, All of these users are located on the same time-frequency resource block, and one user is selected from each cell in turn to form a combination of L users. It is the user index selected in the cell l corresponding to the combination ⁇ k .
  • the users in the combination use the same pilot sequence for channel estimation, and the users in the combination use mutually orthogonal pilot sequences, so that L cells can share the total number of orthogonal pilot sequences with K, and the users in the cell do not mutually interference.
  • Optimal pairing of user combinations ⁇ k using the same pilot sequence can reduce interference caused by pilot pollution. Pilot scheduling is a global optimization problem, and different user combinations will affect the difference between the rate and the rate.
  • search volume is very large, and the number of users and the number of cells are slightly increased, which will lead to a sharp search volume. Increase, not practical in practical applications.
  • the embodiment of the invention provides a pilot scheduling method and a cooperative device for a multi-input and output system, which can achieve better pilot scheduling effects under the condition of low algorithm complexity, and comprehensively consider the computational overhead and integration of the multi-input and output system. Pilot effect.
  • a pilot scheduling method for a multiple input/output system wherein the system includes L cells, and each of the L cells has a maximum of K user equipment UEs, and the method includes: Determining an initial user sequence, where the initial user sequence is a user sequence of the L cells, each user sequence includes K user combinations of the L cells, and each user combination includes a maximum of L UEs, and the K user combinations
  • the UEs in each user combination belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations respectively; according to the initial user sequence Performing a tabu search to obtain an optimized user sequence, wherein the optimized user sequence is a user sequence of the L cells; performing pilot scheduling on the UEs in the L cells according to the optimized user sequence, where the L cells
  • the same user combination belonging to the optimized user sequence The UEs share the same pilot sequence.
  • performing a tabu search according to the initial user sequence to obtain an optimized user sequence is specifically implemented according to the initial user sequence, performing a tabu search according to a pilot scheduling optimization criterion to obtain an optimization.
  • the pilot scheduling optimization criterion comprising: a criterion for maximizing the system and rate of the L cells; or a criterion for maximizing a minimum rate of UEs in the L cells; a system for causing the L cells And the criterion of the highest rate criterion and the rate requirement of the quality of service QoS of the UE of the L cells.
  • the method further includes: acquiring a large-scale fading of a base station of each of the L cells to other cells of the L cell a factor, wherein a large-scale fading factor of a base station of each of the L cells to other cells of the L cell is used to determine a rate of the UE in the L cells, and a sum rate of the system is in the L cells The sum of the rates of all access to the UE.
  • the specific implementation is: when the pilot scheduling optimization criterion is a criterion for maximizing the system and rate of the L cells, according to the The initial user sequence is subjected to the tabu search optimization criterion according to the pilot scheduling optimization criterion to obtain an optimized user sequence.
  • the specific user sequence is implemented by assigning the initial user sequence to the historical user sequence and the current user sequence; and performing the history after the steps a and b are performed N times.
  • User sequence as the optimized user sequence, N is a positive integer not greater than L, where
  • B1 performing location switching of the UEs to be exchanged by any X user combinations in the current user sequence to obtain multiple neighborhood sequences of the current user sequence, and obtaining a system corresponding to multiple neighborhood sequences of the current user sequence Rate, and take out a first neighborhood sequence that maximizes the system and rate, wherein the cell to be exchanged performs the selected cell for the 1st cycle in the process of performing N times cyclically, each time the N cycles are executed
  • the selected cells are different, the system and rate are determined by the large-scale fading factor, and X is a positive integer greater than 1 and not greater than K;
  • the first neighborhood sequence satisfies the special criterion of the tabu search, assign the first neighborhood sequence to the historical user sequence and the current user sequence, and add the first neighborhood sequence to the search contrain table. Or if the first neighborhood sequence does not satisfy the special criterion of the tabu search, assigning the second neighborhood sequence in the plurality of neighborhood sequences of the current user sequence that is not in the search tab list and having the largest system and rate to the current a sequence of users, and adding the second neighborhood sequence to the search forbidden table, wherein the feature criterion is that the system and rate of the first neighborhood sequence is greater than a sum rate of the historical user sequence, or the feature criterion is the first The system and rate of a neighborhood sequence is greater than or equal to the sum rate of the historical user sequence.
  • the specific implementation is: the value of X is 2.
  • the specific implementation is: the predetermined number of times is K times.
  • the specific implementation is:
  • rate( ⁇ opt ) represents the system and rate corresponding to the user sequence of the L cells
  • ⁇ k represents the kth user combination in the user sequence of the L cells
  • rate( ⁇ k ) represents the kth user combination
  • ⁇ jk1 represents a large-scale fading factor of the UE belonging to the cell 1 in the L cells and belonging to the k-th user combination to the base station to which the cell j belongs.
  • determining the initial user The sequence is specifically implemented to randomly determine a user sequence of the L cells as the initial user sequence.
  • determining that the initial user sequence is specifically implemented as : performing greedy search on the plurality of user combinations of the L cells according to the principle that the system and the rate corresponding to the initial user sequence are maximized, and adding the first user combination in each search in the greedy search process to the initial a sequence of users, wherein the first user combination is a combination of users that can join the initial user sequence and the highest rate in each search, and any one of the initial user sequences exists only in one user combination of the initial user sequence.
  • the greedy search is performed on multiple user combinations of the L cells according to the principle that the system and the rate corresponding to the initial user sequence are the largest. And adding the local optimal user combination obtained by each search in the greedy search process to the initial user sequence, specifically, obtaining the respective sum rates of the multiple user combinations according to the large-scale fading factor to form the multiple And a set of rates of the user combination, wherein the sum rate in the sum rate set is in one-to-one correspondence with the plurality of user combinations of the L cells; and the preset number of times of step c and step d is repeated, wherein the preset number of times is not greater than K:
  • determining the initial user sequence specifically includes: after the greedy search is completed, If the number C of user combinations joining the initial user sequence is less than K, KC user combinations are selected from the L cells to join the initial user sequence to form the initial user sequence.
  • the specific implementation is: the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and the Rate is Each of the dimensions of the table corresponds to the one of the L cells, and the subscripts of the first dimension in the Rate table respectively correspond to the UEs of the first cell in the L cells, where the first dimension corresponds to the first Community.
  • a pilot scheduling method for a multi-input and output system wherein the system includes L cells, and each of the L cells has a maximum of K user equipment UEs, and the method includes: Obtaining a plurality of user combinations of the L cells, wherein any one of the multiple user combinations is used to form a user sequence of the L cells, and each user sequence includes K user combinations of the L cells, each The UEs in each of the K user combinations belong to different cells in the L cells, and the UEs in each of the L cells belong to the K, respectively.
  • Different user combinations in the user combinations performing a greedy search on the plurality of user combinations of the L cells, and adding the first user combination in each search in the greedy search process to the optimized user sequence, wherein the first The user combination is a user combination that can join the initial user sequence and satisfy the search condition in each search, and any one of the optimized user sequences exists only in the optimized user sequence.
  • a combination of a user; pilot for scheduling the UE in a cell of L optimize the user based on the sequence, wherein the L cells belonging to the same section of the UE share the same guide the user to optimize the user a sequence of combined pilot sequence.
  • a greedy search is performed on multiple user combinations of the L cells, and a user combination that best matches the search condition in each search in the greedy search process is added to
  • the optimized user sequence is implemented by performing greedy search on multiple user combinations of the L cells according to the pilot scheduling optimization criterion, and Adding the first user combination in the greedy search process to the optimized user sequence, the search condition is used to make the optimized user sequence conform to the pilot scheduling optimization criterion, and the pilot scheduling optimization criterion includes the following criterion: a criterion for the system and rate of the cells; or a criterion for maximizing the minimum rate of the UEs in the L cells; or a criterion for maximizing the system and rate of the L cells and satisfying the quality of service QoS of the UEs of the L cells The criteria for rate requirements.
  • the specific implementation is: when the pilot scheduling optimization criterion is a criterion for maximizing the system and rate of the L cells,
  • the frequency scheduling optimization criterion performs greedy search on the plurality of user combinations of the L cells, and the first user combination in the greedy search process is added to the optimized user sequence, which is implemented according to: the system and the rate of the L cells are maximized.
  • the user sequence is combined with the user with the highest rate.
  • the greedy search is performed on multiple user combinations of the L cells according to a criterion that maximizes the system and rate of the L cells.
  • Adding the first user combination in the greedy search process to the optimized user sequence is specifically implemented as: acquiring a large-scale fading factor of each base station of the L cells to other cells of the L cell, and according to the large scale The fading factor acquires a sum rate of the plurality of user combinations of the L cells to form a sum rate set of the plurality of user combinations of the L cells, where the sum rate in the sum rate set is combined with multiple users of the L cells One-to-one correspondence; repeating the preset number of times of step c and step d, wherein the preset number of times is not greater than K:
  • the method further includes: adding the optimization after the greedy search is completed If the number C of user combinations of the user sequence is less than K, KC user combinations are selected from the L cells to be added to the optimized user sequence to form the optimized user sequence.
  • the specific implementation is: the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and the Rate table is Each of the dimensions corresponds to one of the L cells, and the subscripts of the first dimension in the Rate table respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. .
  • a collaborative device for a multi-input and output system wherein the system includes L cells, and each of the L cells has a maximum of K user equipment UEs, and the coordinated device includes: a unit, configured to determine an initial user sequence, where the initial user sequence is a user sequence of the L cells, each user sequence includes K user combinations of the L cells, and each user combination includes at most L UEs, where The UEs in each of the K user combinations belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations; a unit, performing a tabu search according to the initial user sequence to obtain an optimized user sequence, wherein the optimized user sequence is a user sequence of the L cells; and a scheduling unit, configured to: in the L cells according to the optimized user sequence The UE performs pilot scheduling, where UEs of the same user combination belonging to the optimized user sequence among the L cells share the same pilot sequence.
  • the determining unit is specifically configured to perform, according to the initial user sequence, a tabu search according to a pilot scheduling optimization criterion to obtain an optimized user sequence
  • the pilot scheduling optimization criterion includes the following: a criterion: a criterion for maximizing the system and rate of the L cells; or a criterion of maximizing a minimum rate of UEs in the L cells; or a criterion for maximizing the system and rate of the L cells and satisfying the L A criterion for the rate requirement of the quality of service QoS of the UE of the cell.
  • the collaboration device further includes: an acquiring unit, configured to acquire each UE of the L cells to another cell of the L cell a large-scale fading factor of a base station, wherein a large-scale fading factor of a base station of each of the L cells to other cells of the L cell is used to determine a rate of the UE in the L cells, and a rate of the system The sum of the rates of all access UEs in the L cells.
  • the specific implementation is: when the pilot scheduling optimization criterion is a criterion for maximizing the system and rate of the L cells, the search The unit is specifically configured to assign the initial user sequence to the historical user sequence and the current user sequence; after performing N times in steps a and b, the historical user sequence is used as the optimized user sequence, and N is a positive integer not greater than L, wherein
  • B1 performing location switching of the UEs to be exchanged by any X user combinations in the current user sequence to obtain multiple neighborhood sequences of the current user sequence, and obtaining a system corresponding to multiple neighborhood sequences of the current user sequence Rate, and take out a first neighborhood sequence that maximizes the system and rate, wherein the cell to be exchanged performs the selected cell for the 1st cycle in the process of performing N times cyclically, each time the N cycles are executed
  • the selected cells are different, the system and rate are determined by the large-scale fading factor, and X is a positive integer greater than 1 and not greater than K;
  • the first neighborhood sequence satisfies the special criterion of the tabu search, assign the first neighborhood sequence to the historical user sequence and the current user sequence, and add the first neighborhood sequence to the search contrain table. Or if the first neighborhood sequence does not satisfy the special criterion of the tabu search, assigning the second neighborhood sequence in the plurality of neighborhood sequences of the current user sequence that is not in the search tab list and having the largest system and rate to the current a sequence of users, and adding the second neighborhood sequence to the search forbidden table, wherein the feature criterion is that the system and rate of the first neighborhood sequence is greater than a sum rate of the historical user sequence, or the feature criterion is the first The system and rate of a neighborhood sequence is greater than or equal to the sum rate of the historical user sequence.
  • the specific implementation is: the value of X is 2.
  • the specific implementation is: the predetermined number of times is K times.
  • the possible implementation manner of the fifth possible implementation manner of the third aspect in the sixth possible implementation manner, the specific implementation is: the L
  • the system and rate corresponding to the user sequence of the cell are represented by the following formula:
  • rate( ⁇ opt ) represents the system and rate corresponding to the user sequence of the L cells
  • ⁇ k represents the kth user combination in the user sequence of the L cells
  • rate( ⁇ k ) represents the kth user combination
  • ⁇ jk1 represents a large-scale fading factor of the UE belonging to the cell 1 in the L cells and belonging to the k-th user combination to the base station to which the cell j belongs.
  • the determining unit Specifically, it is used to randomly determine a user sequence of the L cells as the initial user sequence.
  • the search unit It is also used for greedy for multiple user combinations of L cells according to the principle that the system and rate corresponding to the greedy search user sequence are maximized. Searching, and adding a first user combination in each search in the greedy search process to the greedy search user sequence, wherein the first user combination is capable of joining the greedy search user sequence and the maximum rate in each search User combination, any one of the greedy search user sequences exists only in a user combination of the greedy search user sequence; the determining unit is specifically configured to determine the greedy search user sequence as the initial user sequence.
  • the acquiring unit is further configured to obtain, according to the large-scale fading factor, a respective sum rate of the multiple user combinations.
  • the search unit is specifically configured to repeatedly perform step c and Step d preset times, wherein the preset number of times is not greater than K:
  • the determining unit is further configured to perform the greedy search Thereafter, if the number C of user combinations joining the initial user sequence is less than K, KC user combinations are selected from the L cells to join the initial user sequence to form the initial user sequence.
  • the specific implementation is: the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and the Rate is Each of the dimensions of the table corresponds to the one of the L cells, and the subscripts of the first dimension in the Rate table respectively correspond to the UEs of the first cell in the L cells, where the first dimension corresponds to the first Community.
  • a cooperative device of a multi-input and output system wherein the system includes L cells, and each of the L cells has a maximum of K user equipment UEs, and the coordinated device includes: acquiring a unit, configured to acquire multiple user combinations of the L cells, where any one of the multiple user combinations is used to form a user sequence of the L cells, and each user sequence includes K cells of the L cells User combination, each user combination includes a maximum of L UEs, and the UEs in each of the K user combinations belong to different cells in the L cells, and UEs in each of the L cells Each of the K user combinations is a different user combination; the search unit is configured to perform greedy search on the plurality of user combinations of the L cells, and join the first user combination in each search in the greedy search process.
  • the first user combination is a user combination that can join the initial user sequence and satisfy the search condition in each search, and the optimized user sequence
  • the scheduling unit is configured to perform pilot scheduling on the UEs in the L cells according to the optimized user sequence, where the L users belong to the optimized user sequence
  • the UEs of the same user combination share the same pilot sequence.
  • the searching unit is specifically configured to perform greedy search on multiple user combinations of the L cells according to pilot scheduling optimization criteria, and perform the greedy search
  • the first user combination in the process is added to an optimized user sequence
  • the search condition is used to conform the optimized user sequence to the pilot scheduling optimization criterion
  • the pilot scheduling optimization criterion includes the following criterion: a system that makes the L cells And a criterion of the highest rate; or a criterion that maximizes the minimum rate of the UEs in the L cells; or a criterion that maximizes the system and rate of the L cells and satisfies the rate requirement of the quality of service QoS of the UEs of the L cells Guidelines.
  • the specific implementation is: when the pilot scheduling optimization criterion is a criterion for maximizing the system and rate of the L cells, Performing a greedy search on the plurality of user combinations of the L cells according to the pilot scheduling optimization criterion, and adding the first user combination in the greedy search process to the optimized user sequence, where the searching unit is specifically configured to make the L
  • the system and rate of the community is the largest
  • the criterion performs a greedy search on the plurality of user combinations of the L cells, and adds the first user combination in the greedy search process to the optimized user sequence, wherein the first user combination is capable of joining the optimized user in each search.
  • the sequence is combined with the user with the highest rate.
  • the acquiring unit is further configured to obtain, by using the UE, each UE of the L cells to other cells of the L cell. a large-scale fading factor of the base station, and acquiring a sum rate of the plurality of user combinations of the L cells according to the large-scale fading factor to form a sum rate set of the plurality of user combinations of the L cells, where the sum rate set And a rate one-to-one correspondence with a plurality of user combinations of the L cells; performing greedy search on the plurality of user combinations of the L cells according to a criterion that the systems and rates of the L cells are maximized, and the greedy The first user combination in the search process is added to the optimized user sequence, and the search unit is specifically configured to repeatedly perform the preset number of steps c and d, wherein the preset number of times is not greater than K:
  • the searching unit is further configured to join the greedy search after the completion of the greedy search If the number C of user combinations for optimizing the user sequence is less than K, KC user combinations are selected from the L cells to be added to the optimized user sequence to form the optimized user sequence.
  • the specific implementation is: the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and the Rate table is Each of the dimensions corresponds to one of the L cells, and the subscripts of the first dimension in the Rate table respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. .
  • the pilot scheduling method and the cooperative device of the multiple input/output system search for a plurality of user combinations of the L cells to obtain an optimized user sequence of L cells, and optimize the user according to the user.
  • the sequence performs pilot scheduling on the UEs of the L cells, and can obtain a better pilot scheduling effect when the algorithm complexity is low, and balance the computational overhead and the integrated pilot effect of the multi-input and output system.
  • FIG. 1 is a schematic diagram of an application scenario according to an embodiment of the present invention.
  • FIG. 2 is a flow chart of a method for greedy search scheduling according to an embodiment of the present invention.
  • FIG. 3 is a specific flowchart of greedy search according to an embodiment of the present invention.
  • FIG. 4 is a flow chart of a method for tabu search scheduling according to an embodiment of the present invention.
  • FIG. 5 is a specific flowchart of the tabu search in the embodiment of the present invention.
  • 6 is a comparison diagram of effects of greedy search scheduling and random scheduling in an embodiment of the present invention.
  • FIG. 7 is another comparison diagram of greedy search scheduling and random scheduling according to an embodiment of the present invention.
  • FIG. 8 is a comparison diagram of effects between tabu search scheduling and random scheduling according to an embodiment of the present invention.
  • FIG. 9 is another effect comparison diagram of tabu search scheduling and random scheduling according to an embodiment of the present invention.
  • FIG. 10 is a schematic structural diagram of a collaborative device according to an embodiment of the present invention.
  • FIG. 11 is another schematic structural diagram of a collaborative device according to an embodiment of the present invention.
  • FIG. 12 is a schematic diagram of still another structure of a collaborative device according to an embodiment of the present invention.
  • FIG. 13 is a schematic diagram of still another structure of a collaborative device according to an embodiment of the present invention.
  • GSM Global System of Mobile Communication
  • CDMA Code Division Multiple Access
  • WCDMA Wideband Code Division Multiple Access
  • GPRS General Packet Radio Service
  • LTE Long Term Evolution
  • a User Equipment which may also be called a Mobile Terminal, a mobile user equipment, or the like, may communicate with one or more core networks via a radio access network (eg, RAN, Radio Access Network).
  • the user equipment may be a mobile terminal, such as a mobile phone (or "cellular" phone) and a computer with a mobile terminal, for example, a portable, pocket, handheld, computer built-in or in-vehicle mobile device,
  • the wireless access network exchanges languages and/or data.
  • the base station may be a base station (BTS, Base Transceiver Station) in GSM or CDMA, or a base station (NodeB) in WCDMA, or an evolved base station (eNB or e-NodeB, evolutional Node B) in LTE.
  • BTS Base Transceiver Station
  • NodeB base station
  • eNB evolved base station
  • e-NodeB evolutional Node B
  • FIG. 1 is a schematic diagram of an application scenario according to an embodiment of the present invention.
  • each cell in a multi-input and output system composed of L cells, each cell includes one base station equipped with M antennas and a maximum of K single antenna UEs.
  • M antennas For multi-antenna UEs, it can be considered as multiple single-antenna UEs.
  • FIG. 1 is only an application scenario of the embodiment of the present invention.
  • the method and device of the embodiment of the present invention are not limited to the application in the application scenario shown in FIG. 1 .
  • the UEs mentioned herein all refer to users currently accessing one of the L cells.
  • r jk1 represents the distance from the kth UE in the cell 1 of the L cells to the base station in the cell j
  • ⁇ is an attenuation index
  • z jk1 represents the pair of the kth UE in the cell 1 to the base station in the cell j
  • the number of normal random variables satisfies 10log 10 (z jkl ) to CN(0, ⁇ shadow ), and ⁇ jk1 represents the large-scale fading factor of the kth UE in the cell 1 to the base station to which the cell j belongs
  • CN(0, ⁇ shadow ) indicates a Gaussian distribution with a mean of 0, a variance of ⁇ shadow , and a ⁇ shadow representing the variance of the shadow fading of a log Gaussian distribution.
  • the number M of antennas at the base station is a fixed value, and the noise floor must exist.
  • the SINR is not difficult to obtain, and the SINR can be used to calculate the capacity of the system.
  • M ⁇ the number of antennas at the base station is infinite, that is, M ⁇
  • the influence of noise is negligible.
  • the equivalent signal-to-interference ratio (SIR) of the kth UE in cell j is as shown in formula (1.2). Show
  • the embodiment of the present invention calculates the capacity of the system by taking SIR as an example.
  • the UEs in the L cells are grouped to obtain a user sequence of the L cells.
  • Each user sequence includes K user combinations of the L cells, and each user combination includes a maximum of L UEs, and the UEs in each user combination of the K user combinations belong to different L cells.
  • a cell in which each of the L cells belongs to a different user combination of the K user combinations, and UEs of the L user group that belong to the same user combination of the optimized user sequence share the same pilot segment. sequence.
  • the L cells share a total of K orthogonal pilot sequences, and the users in the cell do not interfere with each other.
  • a quasi-orthogonal sequence with good orthogonality but not completely orthogonal can be used in many cases, and no distinction is made here.
  • pilot scheduling optimization criteria When the pilot scheduling is optimized, different user sequences may be acquired based on different optimization purposes, and pilot scheduling of the UEs of the L cells according to the user sequence.
  • the criteria for achieving optimization purposes may be referred to as pilot scheduling optimization criteria.
  • the collaborative device may search for a sequence of users of the L cells according to a criterion that maximizes the system and rate of the L cells.
  • the sum rate of the system can be regarded as the sum of the sum rates of the K user combinations.
  • the UEs of the same user combination share the same pilot sequence, which can ensure that L cells share a total of K orthogonal pilot sequences, and the users in the cell do not interfere with each other, which can reduce pilot pollution.
  • the cooperative device can search for the user sequence of the L cells according to the criterion that the system and rate of the L cells are the largest.
  • the cooperative device may search for a user sequence of L cells according to a criterion that maximizes the minimum user rate in the L cells.
  • a criterion that maximizes the minimum user rate in the L cells.
  • the rate of each user can be obtained.
  • the minimum user rate in this mode can be obtained; the sequence combination of all users is traversed, and the minimum user rate is found to be the largest. That kind of sequence combination can be confirmed as the desired sequence combination.
  • the minimum rate of the UE in the mth user combination ⁇ m in the user sequence can be expressed by the formula (1.4):
  • the cooperative device can search for the user sequence of the L cells according to the criterion of maximizing the minimum user rate in the L cells.
  • the collaborative device may search for the user sequence of the L cells according to criteria that satisfy the quality of service (QoS) requirements of the L cells and the maximum rate requirement of the user.
  • QoS quality of service
  • Equation (1.4) when considering QoS, considering the maximum rate required by the user, Equation (1.4) can be modified to Equation (1.8):
  • equation (1.8) introduces user QoS considerations.
  • the maximum rate required by each user is different.
  • R is infinite; for voice users, R is a small fixed value that does not require a very high signal-to-interference ratio to meet demand.
  • the collaborative device can search for the user sequence of the L cells according to the QoS requirements of the user and the criteria for maximizing the users and rates in the L cells.
  • the collaborative device may also search for the user sequence of the L cells according to other criteria, which is not limited herein.
  • Table 1.1 is a sequence of users of L cells in the embodiment of the present invention.
  • ⁇ represents the user sequence
  • it contains user combinations ⁇ 1 , ⁇ 2 , ..., ⁇ k , ..., ⁇ K for a total of K user combinations.
  • P_1, p_2, ..., p_l, ..., p_L represent the L cells, respectively.
  • the UE included in the cell is a set of elements on the right side of the symbol indicating the cell in the table; the UE included in the combination is a set of elements below the symbol indicating the cell in the table.
  • FIG. 2 is a flow chart of a method for greedy search scheduling according to an embodiment of the present invention, and the method of FIG. 2 is performed by a cooperative device of a multiple input/output system.
  • the system includes L cells, and each of the L cells has a maximum of K UEs.
  • the method includes:
  • the user combination of the multiple user combinations is used to form a user sequence of the L cells, and each user sequence includes K user combinations of the L cells, and each user combination includes at most L UEs, and the K
  • the UEs in each of the user combinations belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations.
  • the user combinations of L cells are a total of K L user combinations.
  • the multiple user combinations may be all or part of user combinations of L cells.
  • the UE mentioned in the embodiment of the present invention is a single antenna UE, and the multi-antenna UE can be regarded as multiple single antenna UEs.
  • a four-antenna UE can be considered as four single-antenna UEs in the embodiment of the present invention.
  • the first user combination is a user combination that can join the initial user sequence and meet the search condition in each search, and any one of the optimized user sequences exists only in one user combination of the optimized user sequence.
  • step 202 is specifically implemented as: performing greedy search on the multiple user combinations of the L cells according to the pilot scheduling optimization criterion, and adding the first user combination in the greedy search process to the optimized user sequence, the search
  • the condition is used to conform the optimized user sequence to the pilot scheduling optimization criterion, the pilot scheduling optimization criterion including a criterion of: a criterion for maximizing a system and a rate of the L cells; or making a UE of the L cells The criterion of the minimum rate maximum; or the criterion that maximizes the system and rate of the L cells and satisfies the rate requirement of the quality of service QoS of the UEs of the L cells.
  • the pilot scheduling optimization criterion may also be other pilot scheduling optimization criteria, which is not limited by the embodiment of the present invention.
  • the search condition is a condition capable of satisfying the purpose of pilot scheduling optimization. Pilot scheduling optimization criteria
  • the search condition is used to select a user combination with the highest median rate of multiple user combinations.
  • the sum rate of the user combination is the sum of the rates of all UEs in the user combination.
  • the pilot scheduling optimization criterion is a criterion for maximizing the minimum rate of UEs in the L cells
  • the search condition is used to select a user combination with the smallest minimum rate of UEs in multiple user combinations.
  • the UEs of the same user combination belonging to the optimized user sequence among the L cells share the same pilot sequence.
  • the optimized user sequence may be sent to the base station to which the L cells belong, so that the base station performs pilot scheduling on the UE of the cell under the jurisdiction of the base station according to the optimized user sequence.
  • the partial sequence corresponding to the first cell in the optimized user sequence may be sent to the base station to which the first cell of the L cells belongs, so that the base station to which the first cell belongs is based on the partial sequence corresponding to the first cell in the optimized user sequence. Pilot scheduling is performed on the UE of the first cell.
  • the first cell may be any one of L cells. Of course, there may be other scheduling modes, which are not limited herein.
  • the algorithm complexity can be obtained. In the lower case, better pilot scheduling effect is obtained, and the computational overhead and integrated pilot effect of the multi-input and output system are balanced.
  • the greedy search is performed on multiple user combinations of the L cells according to pilot scheduling optimization criteria.
  • adding the first user combination in the greedy search process to the optimized user sequence is implemented as: performing greedy search on multiple user combinations of the L cells according to a criterion that maximizes the system and rate of the L cells, and The first user combination in the greedy search process is added to the optimized user sequence, wherein the first user combination is a combination of users who can join the optimized user sequence and the highest rate in each search.
  • performing greedy search on the plurality of user combinations of the L cells according to a criterion that the system and the rate of the L cells are the largest, and adding the first user combination in the greedy search process to the optimized user sequence is implemented as Obtaining a large-scale fading factor of a base station of each of the L cells to other cells of the L cell, and acquiring a sum rate of the multiple user combinations of the L cells according to the large-scale fading factor to form the L a sum rate set of a plurality of users of the cell, wherein a sum rate in the sum rate set is in one-to-one correspondence with a plurality of user combinations of the L cells;
  • Step 2.1 and Step 2.2 for the preset number of times, where the preset number of times is not greater than K:
  • the method further includes: after the greedy search is completed, if the number C of user combinations that join the optimized user sequence is less than K, the KC user combinations are selected from the L cells to be added to the optimized user sequence, To form the optimized user sequence.
  • the user combinations in the optimized user sequence are K, the collection of UEs that optimize the user sequence is equal to the collection of UEs of the L cells.
  • the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and each dimension of the Rate table corresponds to one of the L cells, and the first dimension of the Rate table is The subscripts respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. It should be noted that the first cell may be any one of L cells.
  • FIG. 3 is a specific flowchart of greedy search according to an embodiment of the present invention.
  • the method of Figure 3 is performed by a collaborative device of a multiple input output system.
  • the cooperative device may be set on a certain base station of the multiple input/output system, or may be a certain network element device independent of the base station.
  • An application scenario of the method shown in FIG. 3 can be referred to the description of FIG. 1 , and details are not described herein again.
  • taking the maximum sum rate of the systems in the L cells as an example, for the L cells
  • User combinations perform greedy searches to get an optimized user sequence.
  • the cooperative device may acquire a large-scale fading factor of K UEs of each of the L cells by using suitable means.
  • suitable means for example, the embodiment of the present invention is not limited herein by indicating that the base station measurement is obtained, or is calculated by a calculation model in combination with a map navigation or the like.
  • one UE is selected according to each cell to form a user combination criterion, and there are K L different user combinations in the L cells.
  • the cell may be regarded as a UE with KS signals being 0, and then selected according to each cell.
  • a UE forms a user combination criterion, selects K L user combinations, and then eliminates the UE with a signal of 0 in the user combination, and removes the repeated user combination, and obtains all user combinations of L cells. Therefore, there are at most K L different user combinations among the L cells.
  • greedy search can be performed on all user combinations of L cells, and greedy search can be performed on some user combinations of L cells.
  • a greedy search is performed on a partial user combination of L cells, only some user combinations of the optimized user sequence may be obtained in the end, and other user combinations for optimizing the user sequence need to be determined by other means, for example, from L.
  • the UE is randomly selected to form other user combinations that optimize the user sequence.
  • the pilot scheduling optimization criterion is a criterion for maximizing the sum rate of the systems in the L cells
  • the greedy search is performed on the plurality of user combinations of the L cells
  • the sum rate values of the plurality of user combinations need to be calculated.
  • the sum rate values of the plurality of user combinations may constitute a sum rate set, and the elements (and rates) in the sum rate set are in one-to-one correspondence with the user combinations in the plurality of user combinations.
  • a K-dimensional Rate table may be used to store the sum rate of the sum rate set.
  • the Rate table may be a K-dimensional array, and each dimension of the Rate table corresponds to one of the L cells, for example, the first dimension corresponds to the first cell, and the first dimension corresponds to the first one.
  • the cell, the Lth dimension corresponds to the Lth cell, and so on.
  • One subscript of each dimension corresponds to one UE of the cell corresponding to the dimension, for example, the first cell has K UEs, and the first dimension has K subscripts corresponding to the K UEs, the first If there are K-2 UEs in the cell, the first dimension has K-2 subscripts corresponding to the K-2 UEs, and so on.
  • the sum value of the user combination may be placed in the corresponding position in the Rate table.
  • N K
  • K K
  • the final search result has only one user sequence. If N is less than K, there may be multiple final search results, and one of them can be selected as the result of greedy search.
  • the loop count variable is incremented by 1.
  • a counter can be used for loop counting.
  • step 308 is performed.
  • step 306 is performed.
  • the first user combination corresponding to the current rate rate neutralization rate maximum value is taken out and added to the optimized user sequence.
  • the element ⁇ m with the highest rate can be obtained, namely:
  • the first user combination corresponding to ⁇ m is taken out, that is, the user combination taken out for this iteration is added to the optimized user sequence.
  • the first user combination is If the second user combination corresponding to the element in the Rate table contains One or more of the UEs, the second element is deleted or set to zero.
  • the element in the Rate table that includes the dimension subscript corresponding to each UE in the ⁇ m may be deleted or set to 0.
  • the optimized user sequence at this time is a complete user sequence, which can be directly output.
  • the optimized user sequence at this time is an incomplete user sequence, and the remaining user combinations in the optimized user sequence need to be obtained by other means.
  • the first user combination may be randomly selected from the L cells to join the optimized user sequence until the user combinations in the optimized user sequence are K, wherein the UE in the first user combination has not appeared in the UE of the optimized user sequence.
  • the collection of UEs that optimize the user sequence is equal to the collection of UEs of the L cells.
  • the L cells can be scheduled according to the optimized user sequence.
  • the UEs belonging to the same user combination in the L cells share the same pilot sequence.
  • FIG. 3 is only a specific implementation manner of the embodiment of the present invention.
  • the step of FIG. 3 can also be adjusted to obtain the search result.
  • step 303 can be placed before steps 301 and 302, and the loop counter can start from 0.
  • the embodiment of the present invention is not limited herein.
  • FIG. 6 is a comparison diagram of effects of greedy search scheduling and random scheduling in an embodiment of the present invention.
  • the number of cells is 2
  • the comparison effect of a rate difference between a greedy search schedule and a random user under random scheduling is as shown in FIG. 6.
  • FIG. 7 is another comparison diagram of greedy search scheduling and random scheduling according to an embodiment of the present invention.
  • the number of cells is 3, the comparison effect of a rate difference between a greedy search schedule and a random user under random scheduling is as shown in FIG. 7.
  • each of the L cells includes K UEs, and the greedy search is all user combinations for L cells.
  • the number of loop executions for greedy search is K times.
  • the method in the embodiment of the present invention has a significant improvement in user rate compared to the random scheduling, and the number of users in the cell is increased. The more obvious the effect.
  • Table 3-3 is a comparison of the complexity of the exhaustive search scheduling and the greedy search scheduling of the present invention.
  • the cooperative device may also search for the user sequence of the L cells by using other pilot scheduling optimization criteria, for example, a criterion for maximizing the minimum rate of users in the L cells, and the like. There are no restrictions here.
  • FIG. 4 is a flow chart of a method for tabu search scheduling according to an embodiment of the present invention, and the method of FIG. 4 is performed by a cooperative device of a multi-input and output system.
  • the system includes L cells, and each of the L cells has a maximum of K UEs.
  • the method includes:
  • the initial user sequence is a user sequence of the L cells, and each user sequence includes K user combinations of the L cells, and each user combination includes at most L UEs, and each of the K user combinations
  • the UEs in the user combination belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations.
  • the UE mentioned in the embodiment of the present invention is a single antenna UE, and the multi-antenna UE can be regarded as multiple single antenna UEs.
  • a four-antenna UE can be considered as four single-antenna UEs in the embodiment of the present invention.
  • the optimized user sequence is a user sequence of the L cells.
  • the tabu search is performed based on the initial user sequence.
  • the switching cells in the L cells it is first necessary to specify the switching cells in the L cells, and then exchange the UEs belonging to the switching cells in the initial user sequence to obtain a neighborhood sequence, and obtain a historical optimal user sequence by comparison.
  • the neighborhood switching can be performed multiple times; there can also be multiple cells performing the switching.
  • one of the L cells may be designated as a switching cell one by one.
  • the step 402 is specifically implemented as: performing, according to the initial user sequence, a tabu search according to a pilot scheduling optimization criterion to obtain an optimized user sequence, where the pilot scheduling optimization criterion includes the following criterion: a system that makes the L cells And a criterion of the highest rate; or a criterion that maximizes the minimum rate of the UEs in the L cells; or a criterion that maximizes the system and rate of the L cells and satisfies the rate requirement of the quality of service QoS of the UEs of the L cells Guidelines.
  • the pilot scheduling optimization criterion may also be other pilot scheduling optimization criteria, which is not limited by the embodiment of the present invention.
  • the optimized user sequence may be sent to the base station to which the L cells belong, so that the base station performs pilot scheduling on the UE of the cell under the jurisdiction of the base station according to the optimized user sequence.
  • the partial sequence corresponding to the first cell in the optimized user sequence may be sent to the base station to which the first cell of the L cells belongs, so that the base station to which the first cell belongs is based on the partial sequence corresponding to the first cell in the optimized user sequence. Pilot scheduling is performed on the UE of the first cell.
  • the algorithm may be better in the case of low algorithm complexity.
  • the pilot scheduling effect balances the computational overhead of the multiple input and output systems and the integrated pilot effect.
  • the method further includes: acquiring a large-scale fading factor of a base station of each of the L cells to other cells of the L cell, where each UE of the L cells reaches the The large-scale fading factor of the base station of the other cell of the L cell is used to determine the rate of the UE in the L cells, and the sum rate of the system is the sum of the rates of all the access UEs in the L cells.
  • pilot scheduling optimization criterion is a criterion for maximizing the system and rate of the L cells, according to the initial user sequence, performing a tabu search according to a pilot scheduling optimization criterion to obtain an optimized user.
  • the sequence can be implemented as:
  • the historical user sequence is taken as the optimized user sequence, and N is a positive integer not greater than L, wherein
  • the first neighborhood sequence satisfies the special criterion of the tabu search, assign the first neighborhood sequence to the historical user sequence and the current user sequence, and add the first neighborhood sequence to the search contraindication.
  • assigning the second neighborhood sequence in the plurality of neighborhood sequences of the current user sequence that is not in the search forbidden table and having the largest system and rate Giving a current user sequence, and adding the second neighborhood sequence to the search forbidden table wherein the feature criterion is that the system and rate of the first neighborhood sequence is greater than a sum rate of the historical user sequence, or the special criterion is The system and rate of the first neighborhood sequence is greater than or equal to the sum rate of the historical user sequence.
  • the value of X is 2, that is, only two UEs in the same cell are exchanged when acquiring the neighborhood of the user sequence.
  • the predetermined number of times is K times.
  • the maximum number of iterations of a round of tabu search is set to K times, which can achieve a certain balance in the pilot scheduling effect and algorithm performance.
  • rate( ⁇ opt ) represents the system and rate corresponding to the user sequence of the L cells
  • ⁇ k represents the kth user combination in the user sequence of the L cells
  • rate( ⁇ k ) represents the kth user combination
  • ⁇ jk1 represents a large-scale fading factor of the UE belonging to the cell 1 in the L cells and belonging to the k-th user combination to the base station to which the cell j belongs.
  • the large-scale fading factor ⁇ jkl of the L cells can be expressed by the formula (4.2):
  • r jk1 represents the distance from the kth UE in the cell 1 to the base station in the cell j
  • ⁇ is the attenuation index
  • z jk1 represents the base station logarithm of the kth UE to the cell j in the cell 1
  • Normal random variable satisfying 10log 10 (z jkl ) ⁇ CN(0, ⁇ shadow )
  • ⁇ jkl represents the large-scale fading factor of the kth UE in the cell 1 to the base station in the cell j
  • CN(0, ⁇ shadow ) represents zero mean and variance ⁇ shadow Gaussian distribution
  • ⁇ shadow indicates the number of shadow fading variance Gaussian distribution.
  • determining the initial user sequence is specifically implemented by: randomly determining a user sequence of the L cells as the initial user sequence.
  • determining the initial user sequence is specifically implemented as: performing greedy search on the multiple user combinations of the L cells according to a principle that the system and the rate corresponding to the initial user sequence are maximized, and the greedy The first user combination in each search in the search process is added to the initial user sequence, wherein the first user combination is a combination of users who can join the initial user sequence and the highest rate in each search, the initial user sequence Any one of the UEs exists only in one user combination of the initial user sequence.
  • the greedy search is performed on the plurality of user combinations of the L cells according to the principle that the system and the rate corresponding to the initial user sequence are maximized, and the local optimal user combination obtained by each search in the greedy search process is added.
  • the initial user sequence can be implemented as:
  • the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and each dimension of the Rate table corresponds to one of the L cells, and the first dimension of the Rate table is The subscripts respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. It should be noted that the first cell may be any one of L cells.
  • determining the initial user sequence may also be implemented: after the greedy search is completed, if the number C of user combinations joining the initial user sequence is less than K, then KC user combinations are selected from the L cells to join. The initial user sequence is formed to form the initial user sequence.
  • multiple users of the L cells are combined into all or part of the user combinations of all user combinations of the L cells.
  • FIG. 5 is a specific flowchart of the tabu search in the embodiment of the present invention.
  • the method of Figure 5 is performed by a collaborative device of a multiple input output system.
  • the cooperative device may be set on a certain base station of the multiple input/output system, or may be a certain network element device independent of the base station.
  • An application scenario of the method shown in FIG. 5 can be referred to the description of FIG. 1 , and details are not described herein again.
  • the user sequence of the L cells is searched by taking the maximum sum rate of the systems in the L cells as an example.
  • Collaborative equipment can be based on formula (1.1) A large-scale fading factor for each UE in the L cells is obtained.
  • the cooperative device can also obtain the large-scale fading factor by other means, which is not limited herein.
  • the collaborative device may initialize an initial user sequence of the tabu search.
  • the collaborative device may randomly design a user sequence of L cells as the initial user sequence index, or the cooperative device may also use the user sequence derived according to the greedy search algorithm shown in FIG. 3 as the initial user sequence index.
  • the cooperative device may also initialize the number of cyclic search, the number of times of the initialization cycle search is at most the number of cells L, of course, the number of cyclic search times may also be less than L times, except that the initial sequence is the same and the exchanged cell search order is the same.
  • the user sequence obtained by the search with less than L times of cyclic search times is generally inferior to the pilot scheduling effect of the user sequence obtained by the number of cyclic search times, and the pilot scheduling of the user sequence obtained by the maximum number of cyclic search times is L times. The effect is the same.
  • the embodiment of the present invention describes the method of the present invention L times.
  • the collaborative device can also initialize the number of neighborhood drops Niter.
  • the value of Niter is not specifically limited. From the balance of computational overhead and final effects, the value of Niter can be set to K.
  • the cyclic search count variable l is set to 0, and the historical optimal user sequence P* is set to index.
  • the loop search count variable is incremented by one.
  • the loop search count variable l is incremented by one.
  • a counter can be used to perform a loop search count.
  • step 514 is performed.
  • step 505 is performed.
  • the collaborative device may assign the historical optimal user sequence P* to the target optimization sequence P, and at the same time, set the contraindication table T to null, and set the neighborhood iteration count variable c to 0.
  • the coordinated device may select the first cell of the L cells as the cell to be exchanged of the UE, and the cell may be specifically the L cells, and the previous search is selected as any cell other than the switching cell.
  • the neighborhood iteration count variable is incremented by one.
  • the neighborhood iteration count variable c is cumulatively incremented by one.
  • a neighborhood counter iteration count can be performed.
  • step 503 is performed.
  • step 508 is performed.
  • all neighborhood sequences of P can be obtained. For example, when there are K UEs in the cell to be exchanged, when the UE exchanges, the UEs of the cell to be exchanged can be exchanged at the location of the two users in the P, and the P can be obtained. a neighborhood sequence; if the UE of the cell to be exchanged exchanges the positions of the three user combinations in P, the P can be obtained. a neighborhood sequence; if the UE of the cell to be exchanged is exchanged at the location of the X user combinations in P, P can be obtained. A neighborhood sequence, where X is a positive integer greater than one and no greater than K.
  • the number of all neighbor sequences obtained may be smaller than the corresponding value.
  • a value of 2 for X ensures a good user sequence, and the sequence of neighbors that need to be acquired is not too much.
  • the value of the embodiment X of the present invention takes 2 as an example.
  • the optimal neighborhood sequence of P is taken out.
  • Each user sequence corresponds to a rate and rate of the system.
  • formula (1.4) The system and rate of all neighborhood sequences of P can be obtained, and then the optimal neighborhood sequence of P can be obtained.
  • step 510 is performed.
  • step 511 is performed.
  • the physical meaning is that the system and rate of the optimal neighborhood sequence of the target optimization sequence is greater than the system and rate of the historical optimal user sequence.
  • the optimal neighborhood sequence not in the contraindication table T can be extracted and assigned to P ⁇ , which can be expressed by formula (5.2):
  • the above steps 501 to 514 are a specific implementation of performing a tabu search on the user sequences of the L cells.
  • the steps can also be adjusted.
  • the step of setting the contraindication table to be empty can be placed after the end of each cycle, the counter can be counted from 0, and so on.
  • the steps implemented based on the idea similar to the steps of FIG. 5 are also within the scope of the present invention.
  • the middle neighborhood sequence N(P) calculates the objective function value corresponding to all neighborhood sequences. As shown in Table 5-1.
  • the special criterion is satisfied, which is better than the historical optimal solution.
  • the optimal solution P* is unchanged, and P ⁇ is added to the taboo table T.
  • FIG. 8 is a comparison diagram of effects between tabu search scheduling and random scheduling according to an embodiment of the present invention.
  • the number of cells is 2
  • the comparison effect of a rate difference between a single user under tabu search (TS) scheduling and random scheduling is as shown in FIG. 8.
  • FIG. 9 is another effect comparison diagram of tabu search scheduling and random scheduling according to an embodiment of the present invention.
  • the number of cells is 3, the comparison effect of a rate difference between a single user under taboo (TS) scheduling and random scheduling is shown in FIG.
  • TS single user under taboo
  • each of the L cells includes K UEs, and the greedy search part of the greedy TS scheduling is directed to All user combinations of L cells, and the number of loop executions of the greedy search part is K times.
  • the method in the embodiment of the present invention has a significant improvement in user rate compared to the random scheduling, and the number of users in the cell is increased. The more obvious the effect.
  • Table 5-5 compares the complexity of the exhaustive search scheduling with the TS search scheduling and the greedy TS search scheduling of the present invention.
  • the cooperative device may also search for the user sequence of the L cells by using other pilot scheduling optimization criteria, for example, a criterion for maximizing the minimum rate of users in the L cells, and the like. There are no restrictions here.
  • FIG. 10 is a schematic structural diagram of a cooperative device 1000 of a multi-input and output system according to an embodiment of the present invention.
  • the system includes L cells, and each of the L cells has a maximum of K UEs.
  • the collaboration device 1000 may include an acquisition unit 1001, a search unit 1002, and a scheduling unit 1003.
  • the obtaining unit 1001 is configured to acquire multiple user combinations of the L cells.
  • the user combination of the multiple user combinations is used to form a user sequence of the L cells, and each user sequence includes K user combinations of the L cells, and each user combination includes at most L UEs, and the K
  • the UEs in each of the user combinations belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations.
  • the user combinations of L cells are a total of K L user combinations.
  • the multiple user combinations may be all or part of user combinations of L cells.
  • the UE mentioned in the embodiment of the present invention is a single antenna UE, and the multi-antenna UE can be regarded as multiple single antenna UEs.
  • a four-antenna UE can be considered as four single-antenna UEs in the embodiment of the present invention.
  • the searching unit 1002 is configured to perform greedy search on the plurality of user combinations of the L cells, and add the first user combination in each search in the greedy search process to the optimized user sequence.
  • the first user combination is a user combination that can join the initial user sequence and meet the search condition in each search, and any one of the optimized user sequences exists only in one user combination of the optimized user sequence.
  • the scheduling unit 1003 is configured to perform pilot scheduling on the UEs in the L cells according to the optimized user sequence.
  • the UEs of the same user combination belonging to the optimized user sequence among the L cells share the same pilot sequence.
  • the scheduling unit 1003 may send the optimized user sequence to the base station to which the L cells belong, so that the base station performs pilot scheduling on the UE of the cell under the jurisdiction of the base station according to the optimized user sequence.
  • the scheduling unit 1003 may send, to the base station to which the first cell of the L cells belongs, a partial sequence corresponding to the first cell in the optimized user sequence, so that the base station to which the first cell belongs corresponds to the first cell in the optimized user sequence.
  • the partial sequence performs pilot scheduling on the UE of the first cell.
  • the collaborative device 1000 searches for a plurality of user combinations of the L cells to obtain an optimized user sequence of L cells, and performs pilot scheduling on the UEs of the L cells according to the optimized user sequence.
  • Count A better pilot scheduling effect is achieved when the complexity of the method is low, and the computational overhead and integrated pilot effect of the multi-input and output system are balanced.
  • the cooperative device 1000 may be a base station to which one of the L cells belongs, or may be a network element device that administers L cells, or a network element that is independent of a base station to which any one of the L cells belongs. device.
  • the searching unit 1002 is specifically configured to perform greedy search on the multiple user combinations of the L cells according to the pilot scheduling optimization criterion, and add the first user combination in the greedy search process to the optimized user sequence, the search The condition is used to conform the optimized user sequence to the pilot scheduling optimization criterion, the pilot scheduling optimization criterion including a criterion of: a criterion for maximizing a system and a rate of the L cells; or making a UE of the L cells The criterion of the minimum rate maximum; or the criterion that maximizes the system and rate of the L cells and satisfies the rate requirement of the quality of service QoS of the UEs of the L cells.
  • the pilot scheduling optimization criterion may also be other
  • the searching unit 1002 is specifically configured to use a criterion pair that maximizes the system and rate of the L cells.
  • the plurality of users of the L cells perform a greedy search, and add the first user combination in the greedy search process to the optimized user sequence, wherein the first user combination can join the optimized user sequence in each search and Combined with the highest rate user.
  • the acquiring unit 1001 is further configured to acquire a large-scale fading factor of a base station of each of the L cells to other cells of the L cell, and acquire multiple user combinations of the L cells according to the large-scale fading factor. And a rate to form a sum rate set of a plurality of users of the L cells, wherein a sum rate in the sum rate set is in one-to-one correspondence with a plurality of user combinations of the L cells;
  • the system and the highest rate criterion of the cell perform greedy search for the multiple user combinations of the L cells, and add the first user combination in the greedy search process to the optimized user sequence, and the search unit 1002 is specifically configured to repeatedly perform step 10.1. And the preset number of times in step 10.2, wherein the preset number of times is not greater than K:
  • the searching unit 1002 is further configured to: after the greedy search is completed, if the number C of user combinations that join the optimized user sequence is less than K, select KC user combinations from the L cells to join the optimized user sequence. To form the optimized user sequence.
  • the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and each dimension of the Rate table corresponds to one of the L cells, and the first dimension of the Rate table is The subscripts respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. It should be noted that the first cell may be any one of L cells.
  • the cooperative device 1000 can also perform the method of FIG. 2 and implement the functions of the collaborative device in the embodiment shown in FIG. 2 and FIG. 3 .
  • the cooperative device 1000 can also perform the method of FIG. 2 and implement the functions of the collaborative device in the embodiment shown in FIG. 2 and FIG. 3 .
  • FIG. 11 is a schematic structural diagram of a cooperative device 1100 of a multiple input/output system according to an embodiment of the present invention.
  • the system includes L cells, and each of the L cells has a maximum of K user equipment UEs.
  • the collaborative device 1100 can include a determining unit 1101, a searching unit 1102, and a scheduling unit 1103.
  • the determining unit 1101 is configured to determine an initial user sequence.
  • the initial user sequence is a user sequence of the L cells, and each user sequence includes K user combinations of the L cells, and each user combination includes at most L UEs, and each of the K user combinations
  • the UEs in the user combination belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations.
  • the UE mentioned in the embodiment of the present invention is a single antenna UE, and the multi-antenna UE can be regarded as multiple Single antenna UE.
  • a four-antenna UE can be considered as four single-antenna UEs in the embodiment of the present invention.
  • the searching unit 1102 is configured to perform a tabu search according to the initial user sequence to obtain an optimized user sequence.
  • the optimized user sequence is a user sequence of the L cells.
  • the tabu search is performed based on the initial user sequence.
  • the switching cells in the L cells it is first necessary to specify the switching cells in the L cells, and then exchange the UEs belonging to the switching cells in the initial user sequence to obtain a neighborhood sequence, and obtain a historical optimal user sequence by comparison.
  • the neighborhood switching can be performed multiple times; there can also be multiple cells performing the switching.
  • one of the L cells may be designated as a switching cell one by one.
  • the scheduling unit 1103 is configured to perform pilot scheduling on the UEs in the L cells according to the optimized user sequence, where UEs of the same user combination belonging to the optimized user sequence share the same pilot sequence in the L cells .
  • the scheduling unit 1103 may send the optimized user sequence to the base station to which the L cells belong, so that the base station performs pilot scheduling on the UE of the cell under the jurisdiction of the base station according to the optimized user sequence.
  • the scheduling unit 1103 may send, to the base station to which the first cell of the L cells belongs, a partial sequence corresponding to the first cell in the optimized user sequence, so that the base station to which the first cell belongs corresponds to the first cell in the optimized user sequence.
  • the partial sequence performs pilot scheduling on the UE of the first cell.
  • the cooperative device 1100 searches for the initial user sequence of the L cells to obtain an optimized user sequence, and performs pilot scheduling on the L cells according to the optimized user sequence, which can be performed in a case where the algorithm complexity is low. A better pilot scheduling effect is achieved, and the computational overhead and integrated pilot effects of the multi-input and output system are balanced.
  • the cooperative device 1100 may be a base station to which one of the L cells belongs, or may be a network element device that administers L cells, or a network element that is independent of a base station to which any one of the L cells belongs. device.
  • the performing search unit 1102 is configured to perform a tabu search according to the pilot scheduling optimization criterion to obtain an optimized user sequence according to the initial user sequence for performing a tabu search to obtain an optimized user sequence.
  • the frequency scheduling optimization criterion includes one of the following criteria: a criterion that maximizes the system and rate of the L cells; or a criterion that maximizes the minimum rate of UEs in the L cells; or maximizes the system and rate of the L cells.
  • the pilot scheduling optimization criterion may also be other pilot scheduling optimization criteria, which is not limited by the embodiment of the present invention.
  • the collaboration device 1100 may further include an obtaining unit 1104.
  • An acquiring unit 1104 configured to acquire a large-scale fading factor of a base station of each of the L cells to other cells of the L cell, where each of the L cells reaches a base station of another cell of the L cell
  • the large-scale fading factor is used to determine the rate of the UEs in the L cells, and the sum rate of the system is the sum of the rates of all UEs in the L cells.
  • the pilot scheduling optimization criterion is a criterion for maximizing the system and rate of the L cells
  • the tabu search is performed to obtain an optimized user sequence, and the search unit 1102 is specifically used to:
  • the historical user sequence is taken as the optimized user sequence, and N is a positive integer not greater than L, wherein
  • step 11.2 performing the predetermined number of times for step 11.2.1 and step 11.2.2, wherein
  • 11.2.1. Perform location switching of the UE to be exchanged by any X user combinations in the current user sequence to obtain multiple neighborhood sequences of the current user sequence, and obtain corresponding multiple sequence sequences of the current user sequence.
  • Each selected cell is different, the system and rate are determined by the large-scale fading factor, and X is a positive integer greater than 1 and not greater than K;
  • the first neighborhood sequence satisfies the special criterion of the tabu search, assign the first neighborhood sequence to the historical user sequence and the current user sequence, and add the first neighborhood sequence to the search contraindication.
  • the second neighbor sequence that is not in the search contrain table and has the largest system and rate is assigned to the current user sequence in the multiple neighborhood sequences of the current user sequence, and Adding the second neighborhood sequence to the search forbidden table, where the feature criterion is that the system and rate of the first neighborhood sequence is greater than a sum rate of the historical user sequence, or the feature criterion is the first neighborhood sequence The system and rate are greater than or equal to the sum rate of the historical user sequence.
  • the value of X is 2, that is, only two UEs in the same cell are exchanged when acquiring the neighborhood of the user sequence.
  • the predetermined number of times is K times.
  • the maximum number of iterations of a round of tabu search is set to K times, which can achieve a certain balance in the pilot scheduling effect and algorithm performance.
  • rate( ⁇ opt ) represents the system and rate corresponding to the user sequence of the L cells
  • ⁇ k represents the kth user combination in the user sequence of the L cells
  • rate( ⁇ k ) represents the kth user combination
  • ⁇ jk1 represents a large-scale fading factor of the UE belonging to the cell 1 in the L cells and belonging to the k-th user combination to the base station to which the cell j belongs.
  • the large-scale fading factor ⁇ jkl of the L cells can be expressed by the formula (11.2):
  • r jk1 represents the distance from the kth UE in the cell 1 to the base station in the cell j
  • ⁇ is the attenuation index
  • z jk1 represents the base station logarithm of the kth UE to the cell j in the cell 1
  • Normal random variable satisfying 10log 10 (z jkl ) ⁇ CN(0, ⁇ shadow )
  • ⁇ jkl represents the large-scale fading factor of the kth UE in the cell 1 to the base station in the cell j
  • CN(0, ⁇ shadow ) represents zero mean and variance ⁇ shadow Gaussian distribution
  • ⁇ shadow indicates the number of shadow fading variance Gaussian distribution.
  • the determining unit 1101 in determining an initial user sequence, is specifically configured to randomly determine a user sequence of the L cells as the initial user sequence.
  • the searching unit 1102 is further configured to perform greedy search on the multiple user combinations of the L cells according to a principle that the system and the rate corresponding to the greedy search user sequence are maximized, and the greedy search process is performed.
  • the first user combination in each of the searches is added to the greedy search user sequence, wherein the first user combination is a combination of users who can join the greedy search user sequence and the highest rate in each search, the greedy search user Any one of the UEs in the sequence exists only in one user combination of the greedy search user sequence; the determining unit 1101 is specifically configured to determine the greedy search user sequence as the initial user sequence.
  • the obtaining unit 1104 is further configured to acquire, according to the large-scale fading factor, a sum rate of the plurality of user combinations to form a sum rate set of the plurality of user combinations, where the sum rate in the sum rate set and the L numbers Residential area Multiple user combinations are one-to-one correspondence; greedy search is performed on a plurality of user combinations of the L cells according to a principle that the system and rate corresponding to the greedy search user sequence are maximized, and each search in the greedy search process is performed The first user combination is added to the greedy search user sequence, and the search unit 1102 is specifically configured to repeatedly perform the preset number of steps 11.3 and 11.4, wherein the preset number of times is not greater than K:
  • the determining unit 1101 is further configured to: after the greedy search is completed, if the number C of user combinations that join the initial user sequence is less than K, select KC user combinations from the L cells to join the initial user sequence. Medium to form the initial user sequence.
  • the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and each dimension of the Rate table corresponds to one of the L cells, and the first dimension of the Rate table is The subscripts respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. It should be noted that the first cell may be any one of L cells.
  • multiple users of the L cells are combined into all or part of the user combinations of all user combinations of the L cells.
  • the cooperative device 1100 can also perform the method of FIG. 4 and implement the functions of the collaborative device in the embodiment shown in FIG. 4 and FIG. 5 .
  • the specific implementation can refer to the embodiment shown in FIG. 4 and FIG. 5 . This will not be repeated here.
  • FIG. 12 is a schematic structural diagram of a collaboration device 1200 according to an embodiment of the present invention.
  • the system includes L cells, and each of the L cells has a maximum of K UEs.
  • the collaborative device 1200 can include a receiver 1201, a transmitter 1203, a processor 1202, and a memory 1204.
  • Receiver 1201, transmitter 1203, processor 1202, and memory 1204 are interconnected by a bus 1205 system.
  • the bus 1205 may be an ISA bus, a PCI bus, or an EISA bus.
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one double-headed arrow is shown in Figure 12, but it does not mean that there is only one bus or one type of bus.
  • the memory 1204 is configured to store a program.
  • the program can include program code, the program code including computer operating instructions.
  • Memory 1204 can include read only memory and random access memory and provides instructions and data to processor 1202.
  • the memory 1204 may include a high speed RAM memory and may also include a non-volatile memory such as at least one disk memory.
  • the processor 1202 executes a program stored in the memory 1204, and is configured to acquire multiple user combinations of the L cells, perform greedy search on multiple user combinations of the L cells, and perform each search in the greedy search process.
  • the first user combination is added to the optimized user sequence, and pilot scheduling is performed on the UEs in the L cells according to the optimized user sequence.
  • the user combination of the multiple user combinations is used to form a user sequence of the L cells, and each user sequence includes K user combinations of the L cells, and each user combination includes at most L UEs.
  • the UEs in each of the K user combinations belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations.
  • the first user combination is a user combination that can join the initial user sequence and satisfy the search condition in each search, and any one of the optimized user sequences exists only in one user combination of the optimized user sequence.
  • the UEs of the same user combination belonging to the optimized user sequence among the L cells share the same pilot sequence.
  • the UE mentioned in the embodiment of the present invention is a single antenna UE, and the multi-antenna UE can be regarded as multiple single antenna UEs.
  • a four-antenna UE can be considered as four single-antenna UEs in the embodiment of the present invention.
  • the user combinations of L cells are a total of K L user combinations.
  • the multiple user combinations may be all or part of user combinations of L cells.
  • the processor 1202 may send the optimized user sequence to the base station to which the L cells belong, so that the base station bases the base station according to the optimized user sequence.
  • the UE of the cell in the jurisdiction performs pilot scheduling; or the processor 1202 may send, to the base station to which the first cell of the L cells belongs, a partial sequence corresponding to the first cell in the optimized user sequence, so that the base station to which the first cell belongs Pilot scheduling is performed on the UE of the first cell according to the partial sequence corresponding to the first cell in the optimized user sequence.
  • the first cell may be any one of L cells. Of course, there may be other scheduling modes, which are not limited herein.
  • Processor 1202 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method may be completed by an integrated logic circuit of hardware in the processor 1202 or an instruction in a form of software.
  • the processor 1202 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP processor, etc.), or a digital signal processor (DSP), an application specific integrated circuit. (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component.
  • the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
  • the steps of the method disclosed in the embodiments of the present invention may be directly implemented by the hardware decoding processor, or may be performed by a combination of hardware and software modules in the decoding processor.
  • the software module can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like.
  • the storage medium is located in the memory 1204, and the processor 1202 reads the information in the memory 1204 and completes the steps of the above method in combination with its hardware.
  • the collaborative device 1200 searches for a plurality of user combinations of the L cells to obtain an optimized user sequence of L cells, and performs pilot scheduling on the UEs of the L cells according to the optimized user sequence.
  • a better pilot scheduling effect is achieved when the algorithm complexity is low, and the computational overhead and integrated pilot effect of the multi-input and output system are balanced.
  • the collaboration device 1200 may be a base station to which one of the L cells belongs, or may be a network element device that administers L cells, or a network element that is independent of a base station to which any one of the L cells belongs. device.
  • the greedy search is performed on the multiple user combinations of the L cells to obtain an optimized user sequence
  • the processor 1202 is specifically configured to perform greedy on the multiple user combinations of the L cells according to the pilot scheduling optimization criterion.
  • Searching, and adding the first user combination in the greedy search process to the optimized user sequence the search condition is used to make the optimized user sequence conform to the pilot scheduling optimization criterion
  • the pilot scheduling optimization criterion includes the following criteria: a criterion for maximizing the system and rate of the L cells; or a criterion for maximizing a minimum rate of UEs in the L cells; or a criterion for maximizing the system and rate of the L cells and satisfying UEs of the L cells Guidelines for rate requirements for quality of service QoS.
  • the pilot scheduling optimization criterion may also be other pilot scheduling optimization criteria, which is not limited by the embodiment of the present invention.
  • the pilot scheduling optimization criterion is a criterion for maximizing a system and a rate of the L cells
  • the multiple user combinations of the L cells are used according to pilot scheduling optimization criteria.
  • the greedy search, and the first user combination in the greedy search process is added to the optimized user sequence, and the processor 1202 is specifically configured to combine the multiple users of the L cells according to the criterion that the system and the rate of the L cells are the largest.
  • a greedy search is performed, and the first user combination in the greedy search process is added to the optimized user sequence, wherein the first user combination is a combination of users who can join the optimized user sequence and the highest rate in each search.
  • the processor 1202 may acquire, by the receiver 1201 and the transmitter 1203, a large-scale fading factor of each base station of the L cells to other base stations of the L cell, and acquire the L according to the large-scale fading factor.
  • the sum rate of the multiple users of the cell is combined to form a sum rate set of a plurality of users of the L cells, wherein the sum rate in the sum rate set is in one-to-one correspondence with the plurality of user combinations of the L cells.
  • the system and the rate-maximizing criterion of the cell perform greedy search for the multiple user combinations of the L cells, and add the first user combination in the greedy search process to the optimized user sequence, and the processor 1202 is specifically configured to repeatedly perform step 12.1. And the preset number of times in step 12.2, wherein the preset number of times is not greater than K:
  • the processor 1202 is further configured to: after the greedy search is completed, if the number C of user combinations that join the optimized user sequence is less than K, select KC user combinations from the L cells to join the optimized user sequence. To form the optimized user sequence.
  • the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and each dimension of the Rate table corresponds to one of the L cells, and the first dimension of the Rate table is The subscripts respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. It should be noted that the first cell may be any one of L cells.
  • the processor 1202 is specifically configured to acquire all or a part of user combinations of the most K L user combinations of the L cells.
  • the cooperative device 1200 can also perform the method of FIG. 2 and implement the functions of the cooperative device in the embodiment shown in FIG. 2 and FIG. 3 .
  • the cooperative device 1200 can also perform the method of FIG. 2 and implement the functions of the cooperative device in the embodiment shown in FIG. 2 and FIG. 3 .
  • FIG. 13 is a schematic structural diagram of a collaboration device 1300 according to an embodiment of the present invention.
  • the system includes L cells, and each of the L cells has a maximum of K user equipment UEs.
  • the collaborative device 1300 can include a receiver 1301, a transmitter 1303, a processor 1302, and a memory 1304.
  • Receiver 1301, transmitter 1303, processor 1302, and memory 1304 are interconnected by a bus 1305 system.
  • the bus 1305 can be an ISA bus, a PCI bus, or an EISA bus.
  • the bus can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one double-headed arrow is shown in Figure 13, but it does not mean that there is only one bus or one type of bus.
  • the memory 1304 is configured to store a program.
  • the program can include program code, the program code including computer operating instructions.
  • Memory 1304 can include read only memory and random access memory and provides instructions and data to processor 1302.
  • Memory 1304 may include high speed RAM memory and may also include non-volatile memory, such as at least one disk memory.
  • the processor 1302 is configured to execute a program stored in the memory 1304, configured to determine an initial user sequence, perform a tabu search according to the initial user sequence to obtain an optimized user sequence, and perform piloting on the UEs in the L cells according to the optimized user sequence. Scheduling.
  • the initial user sequence is a user sequence of the L cells, and each user sequence includes K user combinations of the L cells, and each user combination includes at most L UEs, and each of the K user combinations
  • the UEs in the user combination belong to different cells in the L cells, and the UEs in each of the L cells belong to different user combinations in the K user combinations
  • the optimized user sequence is the L users.
  • a user sequence of a cell in which UEs belonging to the same user combination of the optimized user sequence share the same pilot sequence.
  • the UE mentioned in the embodiment of the present invention is a single antenna UE, and the multi-antenna UE can be regarded as multiple single antenna UEs.
  • a four-antenna UE can be considered as four single-antenna UEs in the embodiment of the present invention.
  • the tabu search is performed based on the initial user sequence.
  • the switching cells in the L cells it is first necessary to specify the switching cells in the L cells, and then exchange the UEs belonging to the switching cells in the initial user sequence to obtain a neighborhood sequence, and obtain a historical optimal user sequence by comparison.
  • the neighborhood switching can be performed multiple times; there can also be multiple cells performing the switching.
  • one of the L cells may be designated as a switching cell one by one.
  • the optimized user sequence may be sent to the base station to which the L cells belong, so that the base station performs pilot scheduling on the UE of the cell under the jurisdiction of the base station according to the optimized user sequence.
  • the processor 1302 may, when performing pilot scheduling on the UEs in the L cells according to the optimized user sequence, send, to the base station to which the first cell of the L cells belongs, the part corresponding to the first cell in the optimized user sequence.
  • the sequence is such that the base station to which the first cell belongs performs pilot scheduling on the UE of the first cell according to the partial sequence corresponding to the first cell in the optimized user sequence.
  • the pilot scheduling optimization criterion may also be other pilot scheduling optimization criteria, which is not limited by the embodiment of the present invention.
  • the method performed by the cooperative device disclosed in any of the embodiments of FIG. 4 and FIG. 5 of the present invention may be applied to the processor 1302 or implemented by the processor 1302.
  • the processor 1302 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method may be completed by an integrated logic circuit of hardware in the processor 1302 or an instruction in a form of software.
  • the processor 1302 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP Processor, etc.), or a digital signal processor (DSP), an application specific integrated circuit. (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component.
  • CPU central processing unit
  • NP Processor network processor
  • DSP digital signal processor
  • ASIC application specific integrated circuit
  • FPGA off-the-shelf programmable gate array
  • the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
  • the steps of the method disclosed in the embodiments of the present invention may be directly implemented by the hardware decoding processor, or may be performed by a combination of hardware and software modules in the decoding processor.
  • the software module can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like.
  • the storage medium is located in memory 1304, and processor 1302 reads the information in memory 1304 and, in conjunction with its hardware, performs the steps of the above method.
  • the collaborative device 1300 searches for the initial user sequence of the L cells to obtain an optimized user sequence, and performs pilot scheduling on the L cells according to the optimized user sequence, which can be performed when the algorithm complexity is low. A better pilot scheduling effect is achieved, and the computational overhead and integrated pilot effects of the multi-input and output system are balanced.
  • the cooperative device 1300 may be a base station to which one of the L cells belongs, or may be a network element device that administers L cells, or a network element that is independent of a base station to which any one of the L cells belongs. device.
  • the processor 1302 is specifically configured to perform, according to the initial user sequence, a tabu search according to a pilot scheduling optimization criterion to obtain an optimized user sequence.
  • the frequency scheduling optimization criterion includes one of the following criteria: a criterion that maximizes the system and rate of the L cells; or a criterion that maximizes the minimum rate of UEs in the L cells; or maximizes the system and rate of the L cells.
  • the pilot scheduling optimization criterion may also be other pilot scheduling optimization criteria, which is not limited by the embodiment of the present invention.
  • the processor 1302 may further acquire, by the receiver 1301 and the transmitter 1303, a large-scale fading factor of a base station of each of the L cells to other cells of the L cell, where the L The large-scale fading factor of the base station of each of the cells in the cell to the other cell of the L cell is used to determine the rate of the UE in the L cells, and the sum rate of the system is the sum of the rates of all the UEs in the L cells. .
  • the processor 1302 is specifically configured to:
  • the historical user sequence is taken as the optimized user sequence, and N is a positive integer not greater than L, wherein
  • the first neighborhood sequence satisfies the special criterion of the tabu search, assign the first neighborhood sequence to the historical user sequence and the current user sequence, and add the first neighborhood sequence to the search taboo.
  • the value of X is 2, that is, only two UEs in the same cell are exchanged when acquiring the neighborhood of the user sequence.
  • the predetermined number of times is K times.
  • the maximum number of iterations of a round of tabu search is set to K times, which can achieve a certain balance in the pilot scheduling effect and algorithm performance.
  • rate( ⁇ opt ) represents the system and rate corresponding to the user sequence of the L cells
  • ⁇ k represents the kth user combination in the user sequence of the L cells
  • rate( ⁇ k ) represents the kth user combination
  • ⁇ jk1 represents a large-scale fading factor of the UE belonging to the cell 1 in the L cells and belonging to the k-th user combination to the base station to which the cell j belongs.
  • the large-scale fading factor ⁇ jkl of the L cells can be expressed by the formula (13.2):
  • r jk1 represents the distance from the kth UE in the cell 1 to the base station in the cell j
  • ⁇ is the attenuation index
  • z jk1 represents the base station logarithm of the kth UE to the cell j in the cell 1
  • Normal random variable satisfying 10log 10 (z jkl ) ⁇ CN(0, ⁇ shadow )
  • ⁇ jkl represents the large-scale fading factor of the kth UE in the cell 1 to the base station in the cell j
  • CN(0, ⁇ shadow ) represents zero mean and variance ⁇ shadow Gaussian distribution
  • ⁇ shadow indicates the number of shadow fading variance Gaussian distribution.
  • the processor 1302 in determining an initial user sequence, is specifically configured to randomly determine a user sequence of the L cells as the initial user sequence.
  • the processor 1302 in determining an initial user sequence, is specifically configured to perform greedy search on multiple user combinations of the L cells according to a principle that the system and the rate corresponding to the initial user sequence are maximized. And adding a first user combination in each search in the greedy search process to the initial user sequence, wherein the first user combination is a user combination that can join the initial user sequence and the highest rate in each search , Any one of the initial user sequences exists only in one user combination of the initial user sequence.
  • the processor 1302 is further configured to acquire, according to the large-scale fading factor, respective sum rates of the plurality of user combinations to form a sum rate set of the multiple user combinations, where the sum rate in the sum rate set and the L numbers Multiple user combinations of cells are one-to-one correspondence; greedy search is performed on multiple user combinations of the L cells according to the principle that the system and rate corresponding to the initial user sequence are maximized, and each time in the greedy search process
  • the first user combination in the search is added to the initial user sequence, and the processor 1302 is specifically configured to repeatedly perform the preset times of the step 13.3 and the step 13.4, wherein the preset number of times is not greater than K:
  • the processor 1302 is further configured to: after the greedy search is completed, if the number C of user combinations joining the initial user sequence is less than K, select KC from the L cells. User combinations are added to the initial user sequence to form the initial user sequence.
  • the sum rate set is represented by a Rate table, where the Rate table is an L-dimensional array, and each dimension of the Rate table corresponds to one of the L cells, and the first dimension of the Rate table is The subscripts respectively correspond to the UEs of the first cell in the L cells, and the first dimension corresponds to the first cell. It should be noted that the first cell may be any one of L cells.
  • multiple users of the L cells are combined into all or part of the user combinations of all user combinations of the L cells.
  • the cooperative device 1300 can also perform the method of FIG. 4 and implement the functions of the collaborative device in the embodiment shown in FIG. 4 and FIG. 5, and the specific implementation can refer to the embodiment shown in FIG. 4 and FIG. This will not be repeated here.
  • the disclosed systems, devices, and methods may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be in an electrical, mechanical or other form.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the functions may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a standalone product.
  • the technical solution of the present invention which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including
  • the instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention.
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本发明实施例提供了一种多输入输出系统的导频调度方法及协同设备，该方法包括确定初始用户序列；根据该初始用户序列进行禁忌搜索以获得优化用户序列；根据该优化用户序列对该L个小区中的UE进行导频调度，其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。本发明实施例中，通过对该L个小区的多个用户组合进行搜索以得到L个小区的优化用户序列，并根据优化用户序列对L个小区的UE进行导频调度，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。

Description

多输入输出系统的导频调度方法及协同设备 技术领域
本发明实施例涉及通信领域，并且更具体地，涉及一种多输入输出系统的导频调度方法及协同设备。
背景技术
大规模MIMO(Very Large MIMO或Massive MIMO)以其特有的优点：获得更高倍数的信道容量，更低的能量消耗，十分精准的空间区分度，相对廉价的硬件实现等，获得了无线通信领域的相当关注。
现有的一种参考信号指示(CSI)反馈模式中，其反馈量随着天线数线性增长，随着基站处天线数目的大量增加，当天线数目很大时，反馈所需的时间将会远大于信道相干时间。因此，在大规模MIMO系统中主要通过利用信道互易性来获得信道状态信息。
但是由于导频信号空间的维数总是有限的，所以不可避免的总是存在不同小区的用户采用相同导频同时发射，从而导致基站无法区分，形成所谓的“导频污染”。当基站不存在协作时，随着基站天线数的无限增加，非相关噪声和快衰落效应都能被平均掉，影响系统性能的主要是由于导频污染所致的小区间干扰，而且无论上行还是下行链路，等效信干噪比都仅与大尺度衰落因子相关；并且当存在导频污染时，提高上行导频的发射功率对提升信道估计性能是没有任何意义的。
一种考虑基站间无协作的Massive MIMO多小区多用户系统，含L个小区，每小区含K个单天线用户(多天线用户可视为多个单天线用户)，进行全网频率复用，所有这些用户位于相同的时频资源块上，依次从每个小区中选择一个用户组成一个含L个用户的组合
Figure PCTCN2014090531-appb-000001
是小区l中选择出的对应于组合Ωk的用户指数，这样的用户组共有K个，记为Ω＝{Ω1,Ω2,…,ΩK}。组合内用户使用同一导频序列进行信道估计，组合间用户使用相互正交的导频序列，这样可以保证L个小区共享总个数为K的正交导频序列，且小区内各用户互不干扰。对使用同一导频序列的用户组合Ωk进行优化配对，可以减小导频污染带来的干扰。导频调度是一个全局优化问题，不同的用户组合情况将影响到和速率的不同。
如果采用穷举搜索法，对所有可能的用户组合进行遍历搜索，选择出使得系统和速率最大的最优用户组，搜索量非常大，用户数量和小区数稍微增加将带来搜索量的急剧猛增，在实际应用中不可行。
发明内容
本发明实施例提供一种多输入输出系统的导频调度方法及协同设备，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
第一方面，提供了一种多输入输出系统的导频调度方法，其特征在于，其中该系统包括L个小区，该L个小区的每一个小区最多存在K个用户设备UE，该方法包括：确定初始用户序列，该初始用户序列为该L个小区的一种用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合；根据该初始用户序列进行禁忌搜索以获得优化用户序列，其中，该优化用户序列为该L个小区的一种用户序列；根据该优化用户序列对该L个小区中的UE进行导频调度，其中，该L个小区中属于该优化用户序列的同一个用户组合 的UE共享相同的一段导频序列。
结合第一方面，在第一种可能的实现方式中，根据该初始用户序列进行禁忌搜索以获得优化用户序列具体实现为，根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。。
结合第一方面的第一种可能的实现方式，在第二种可能的实现方式中，该方法还包括：获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，其中，该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子用于确定该L个小区中的UE的速率，该系统的和速率为该L个小区中所有接入UE的速率之和。
结合第一方面的第二种可能的实现方式，在第三种可能的实现方式中，具体实现为：当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列具体实现为：将该初始用户序列赋值给历史用户序列和当前用户序列；对步骤a、b循环执行N次后将该历史用户序列作为该优化用户序列，N为不大于L的正整数，其中
a、将搜索禁忌表置空；
b、对步骤b1、b2执行预定的次数，其中
b1、将待交换小区在该当前用户序列中任意X个用户组合的UE进行位置交换得到该当前用户序列的多个邻域序列，并获得该当前用户序列的多个邻域序列对应的系统和速率，并取出使得系统和速率最大的第一邻域序列，其中，该待交换小区为循环执行N次的过程中的第l次循环执行选定的小区，该N次的循环执行的每一次选定的小区均不相同，该系统和速率由该大尺度衰落因子确定，X为大于1且不大于K的正整数；
b2、如果该第一邻域序列满足该禁忌搜索的特赦准则，则将该第一邻域序列赋值给该历史用户序列和当前用户序列，并将该第一邻域序列加入该搜索禁忌表中，或者如果该第一邻域序列不满足禁忌搜索的特赦准则，则将该当前用户序列的多种邻域序列中不在该搜索禁忌表中且系统和速率最大的第二邻域序列赋值给当前用户序列，并将该第二邻域序列加入该搜索禁忌表中，其中，该特赦准则为该第一邻域序列的系统和速率大于该历史用户序列的和速率，或者该特赦准则为该第一邻域序列的系统和速率大于或等于该历史用户序列的和速率。
结合第一方面的第三种可能的实现方式，在第四种可能的实现方式中，具体实现为：X的值为2。
结合第一方面的第三种可能的实现方式或第一方面的第四种可能的实现方式，在第五种可能的实现方式中，具体实现为：该预定的次数为K次。
结合第一方面的第二种可能的实现方式至第一方面的第五种可能的实现方式中任一种可能的实现方式，在第六种可能的实现方式中，具体实现为：
该L个小区的用户序列对应的系统和速率由如下公式表示：
Figure PCTCN2014090531-appb-000002
其中，rate(Ωopt)表示该L个小区的用户序列对应的系统和速率，Ωk表示该L个小区的用户序列中第k个用户组合，rate(Ωk)表示该第k个用户组合的和速率，
Figure PCTCN2014090531-appb-000003
βjkl表示属于该L个小区中的小区l且属于该第k个用户组合的UE到小区j所属基站的大尺度衰落因子。
结合第一方面或第一方面的第一种可能的实现方式至第一方面的第六种可能的实现方式中任一种可能的实现方式，在第七种可能的实现方式中，确定初始用户序列具体实现为随机确定该L个小区的一个用户序列作为该初始用户序列。
结合第一方面的第三种可能的实现方式至第一方面的第六种可能的实现方式中任一种可能的实现方式，在第八种可能的实现方式中，确定初始用户序列具体实现为：根据使得该初始用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到该初始用户序列，其中该第一用户组合为该每一次搜索中能够加入该初始用户序列并且和速率最大的用户组合，该初始用户序列中的任一个UE只存在于该初始用户序列的一个用户组合中。
结合第一方面的第八种可能的实现方式，在第九种可能的实现方式中，根据使得该初始用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索得到的局部最优用户组合加入到该初始用户序列具体实现为，根据该大尺度衰落因子获取该多个用户组合各自的和速率以形成该多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；重复执行步骤c和步骤d预设的次数，其中该预设的次数不大于K：
c、将该第一用户组合加入到该初始用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
d、将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
结合第一方面的第八种可能的实现方式或第一方面的第九种可能的实现方式，在第十种可能的实现方式中，确定初始用户序列具体还包括：在该贪婪搜索完成之后，如果加入该初始用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该初始用户序列中，以形成该初始用户序列。
结合第一方面的第九种可能的实现方式，在第十一种可能的实现方式中，具体实现为：该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。
第二方面，提出了一种多输入输出系统的导频调度方法，其特征在于，其中该系统包括L个小区，该L个小区的每一个小区最多存在K个用户设备UE，该方法包括：获取该L个小区的多个用户组合，其中该多个用户组合中的任一个用户组合用于构成该L个小区的用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合；对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到优化用户序列，其中该第一用户组合为该每一次搜索中能够加入该初始用户序列并且满足搜索条件的用户组合，该优化用户序列中的任一个UE只存在于该优化用户序列的一个用户组合中；根据该优化用户序列对该L个小区中的UE进行导频调度，其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
结合第二方面，在第一种可能的实现方式中，对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中最符合搜索条件的用户组合加入到优化用户序列具体实现为，按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并 将该贪婪搜索过程中的第一用户组合加入到优化用户序列，该搜索条件用于使该优化用户序列符合该导频调度优化准则，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。
结合第二方面的第一种可能的实现方式，在第二种可能的实现方式中，具体实现为：当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列具体实现为：按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，其中该第一用户组合为该每一次搜索中能够加入该优化用户序列并且和速率最大的用户组合。
结合第二方面的第二种可能的实现方式，在第三种可能的实现方式中，按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列具体实现为，获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，并根据该大尺度衰落因子获取该L个小区的多个用户组合的和速率以形成该L个小区的多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；重复执行步骤c和步骤d预设的次数，其中该预设的次数不大于K：
c、将该第一用户组合加入到该优化用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
d、将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
结合第二方面的第二种可能的实现方式或第二方面的第三种可能的实现方式，在第四种可能的实现方式中，该方法还包括：在贪婪搜索完成之后，如果加入该优化用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该优化用户序列中，以形成该优化用户序列。
结合第二方面的第三种可能的实现方式，在第五种可能的实现方式中，具体实现为：该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。
第三方面，提出了一种多输入输出系统的协同设备，其特征在于，其中该系统包括L个小区，该L个小区的每一个小区最多存在K个用户设备UE，该协同设备包括：确定单元，用于确定初始用户序列，该初始用户序列为该L个小区的一种用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合；搜索单元，根据该初始用户序列进行禁忌搜索以获得优化用户序列，其中，该优化用户序列为该L个小区的一种用户序列；调度单元，用于根据该优化用户序列对该L个小区中的UE进行导频调度，其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
结合第三方面，在第一种可能的实现方式中，该确定单元具体实现为根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。
结合第三方面的第一种可能的实现方式，在第二种可能的实现方式中，该协同设备还包括：获取单元，用于获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，其中，该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子用于确定该L个小区中的UE的速率，该系统的和速率为该L个小区中所有接入UE的速率之和。结合第三方面的第二种可能的实现方式，在第三种可能的实现方式中，具体实现为：当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，该搜索单元具体用于将该初始用户序列赋值给历史用户序列和当前用户序列；对步骤a、b循环执行N次后将该历史用户序列作为该优化用户序列，N为不大于L的正整数，其中
a、将搜索禁忌表置空；
b、对步骤b1、b2执行预定的次数，其中
b1、将待交换小区在该当前用户序列中任意X个用户组合的UE进行位置交换得到该当前用户序列的多个邻域序列，并获得该当前用户序列的多个邻域序列对应的系统和速率，并取出使得系统和速率最大的第一邻域序列，其中，该待交换小区为循环执行N次的过程中的第l次循环执行选定的小区，该N次的循环执行的每一次选定的小区均不相同，该系统和速率由该大尺度衰落因子确定，X为大于1且不大于K的正整数；
b2、如果该第一邻域序列满足该禁忌搜索的特赦准则，则将该第一邻域序列赋值给该历史用户序列和当前用户序列，并将该第一邻域序列加入该搜索禁忌表中，或者如果该第一邻域序列不满足禁忌搜索的特赦准则，则将该当前用户序列的多种邻域序列中不在该搜索禁忌表中且系统和速率最大的第二邻域序列赋值给当前用户序列，并将该第二邻域序列加入该搜索禁忌表中，其中，该特赦准则为该第一邻域序列的系统和速率大于该历史用户序列的和速率，或者该特赦准则为该第一邻域序列的系统和速率大于或等于该历史用户序列的和速率。
结合第三方面的第三种可能的实现方式，在第四种可能的实现方式中，具体实现为：X的值为2。
结合第三方面的第三种可能的实现方式或第三方面的第四种可能的实现方式，在第五种可能的实现方式中，具体实现为：该预定的次数为K次。
结合第三方面的第二种可能的实现方式至第三方面的第五种可能的实现方式中任一种可能的实现方式，在第六种可能的实现方式中，具体实现为：该L个小区的用户序列对应的系统和速率由如下公式表示：
Figure PCTCN2014090531-appb-000004
其中，rate(Ωopt)表示该L个小区的用户序列对应的系统和速率，Ωk表示该L个小区的用户序列中第k个用户组合，rate(Ωk)表示该第k个用户组合的和速率，
Figure PCTCN2014090531-appb-000005
βjkl表示属于该L个小区中的小区l且属于该第k个用户组合的UE到小区j所属基站的大尺度衰落因子。
结合第三方面或第三方面的第一种可能的实现方式至第三方面的第六种可能的实现方式中任一种可能的实现方式，在第七种可能的实现方式中，该确定单元具体用于随机确定该L个小区的一个用户序列作为该初始用户序列。
结合第三方面的第三种可能的实现方式至第三方面的第六种可能的实现方式中任一种可能的实现方式，在第八种可能的实现方式中，具体实现为：该搜索单元还用于根据使得贪婪搜索用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪 搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到该贪婪搜索用户序列，其中该第一用户组合为该每一次搜索中能够加入该贪婪搜索用户序列并且和速率最大的用户组合，该贪婪搜索用户序列中的任一个UE只存在于该贪婪搜索用户序列的一个用户组合中；该确定单元具体用于确定该贪婪搜索用户序列为该初始用户序列。
结合第三方面的第八种可能的实现方式，在第九种可能的实现方式中，具体实现为，该获取单元还用于根据该大尺度衰落因子获取该多个用户组合各自的和速率以形成该多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；在用于根据使得贪婪搜索用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到贪婪搜索用户序列，该搜索单元具体用于重复执行步骤c和步骤d预设的次数，其中该预设的次数不大于K：
c、将该第一用户组合加入到该初始用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
d、将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
结合第三方面的第八种可能的实现方式或第三方面的第九种可能的实现方式，在第十种可能的实现方式中，具体实现为：该确定单元还用于在该贪婪搜索完成之后，如果加入该初始用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该初始用户序列中，以形成该初始用户序列。
结合第三方面的第九种可能的实现方式，在第十一种可能的实现方式中，具体实现为：该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。
第四方面，提出了一种多输入输出系统的协同设备，其特征在于，其中该系统包括L个小区，该L个小区的每一个小区最多存在K个用户设备UE，该协同设备包括：获取单元，用于获取该L个小区的多个用户组合，其中该多个用户组合中的任一个用户组合用于构成该L个小区的用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合；搜索单元，用于对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到优化用户序列，其中该第一用户组合为该每一次搜索中能够加入该初始用户序列并且满足搜索条件的用户组合，该优化用户序列中的任一个UE只存在于该优化用户序列的一个用户组合中；调度单元，用于根据该优化用户序列对该L个小区中的UE进行导频调度，其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
结合第四方面，在第一种可能的实现方式中，具体实现为：该搜索单元具体用于按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，该搜索条件用于使该优化用户序列符合该导频调度优化准则，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。
结合第四方面的第一种可能的实现方式，在第二种可能的实现方式中，具体实现为：当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，在用于按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，该搜索单元具体用于按照使得该L个小区的系统和速率最大的 准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，其中该第一用户组合为该每一次搜索中能够加入该优化用户序列并且和速率最大的用户组合。
结合第四方面的第二种可能的实现方式，在第三种可能的实现方式中，具体实现为，该获取单元还用于获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，并根据该大尺度衰落因子获取该L个小区的多个用户组合的和速率以形成该L个小区的多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；在用于按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，该搜索单元具体用于重复执行步骤c和步骤d预设的次数，其中该预设的次数不大于K：
c、将该第一用户组合加入到该优化用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
d、将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
结合第四方面的第二种可能的实现方式或第四方面的第三种可能的实现方式，在第四种可能的实现方式中，该搜索单元还用于在贪婪搜索完成之后，如果加入该优化用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该优化用户序列中，以形成该优化用户序列。
结合第四方面的第三种可能的实现方式，在第五种可能的实现方式中，具体实现为：该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。
基于以上技术方案，本发明实施例的多输入输出系统的导频调度方法及协同设备，通过对该L个小区的多个用户组合进行搜索以得到L个小区的优化用户序列，并根据优化用户序列对L个小区的UE进行导频调度，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
附图说明
为了更清楚地说明本发明实施例的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本发明实施例的一种应用场景示意图。
图2是本发明实施例贪婪搜索调度的方法流程图。
图3是本发明实施例贪婪搜索的具体流程图。
图4是本发明实施例禁忌搜索调度的方法流程图。
图5是本发明实施例禁忌搜索的具体流程图。
图6是本发明实施例贪婪搜索调度与随机调度的一个效果对比图。
图7是本发明实施例贪婪搜索调度与随机调度的另一个效果对比图。
图8是本发明实施例禁忌搜索调度与随机调度的一个效果对比图。
图9是本发明实施例禁忌搜索调度与随机调度的另一个效果对比图。
图10是本发明实施例协同设备的一个结构示意图。
图11是本发明实施例协同设备的另一个结构示意图。
图12是本发明实施例协同设备的再一个结构示意图。
图13是本发明实施例协同设备的再一个结构示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
本发明的技术方案，可以应用于各种通信系统，例如：全球移动通讯系统(GSM，Global System of Mobile communication)，码分多址(CDMA，Code Division Multiple Access)系统，宽带码分多址(WCDMA，Wideband Code Division Multiple Access Wireless)，通用分组无线业务(GPRS，General Packet Radio Service)，长期演进(LTE，Long Term Evolution)等。
用户设备(UE，User Equipment)，也可称之为移动终端(Mobile Terminal)、移动用户设备等，可以经无线接入网(例如，RAN，Radio Access Network)与一个或多个核心网进行通信，用户设备可以是移动终端，如移动电话(或称为“蜂窝”电话)和具有移动终端的计算机，例如，可以是便携式、袖珍式、手持式、计算机内置的或者车载的移动装置，它们与无线接入网交换语言和/或数据。
基站，可以是GSM或CDMA中的基站(BTS，Base Transceiver Station)，也可以是WCDMA中的基站(NodeB)，还可以是LTE中的演进型基站(eNB或e-NodeB，evolutional Node B)，本发明并不限定，但为描述方便，下述实施例以eNB为例进行说明。
需要说明的是，本发明实施例中，相同的字母正体和斜体没有区别，均代表相同的含义。
图1是本发明实施例的一种应用场景示意图。如图1所示，在一个由L个小区组成的多输入输出系统，每小区含一个配备M根天线的基站和最多K个单天线UE。对于多天线UE，可视为多个单天线UE。当然，图1仅仅是本发明实施例的一种应用场景，本发明实施例的方法和设备并不限于图1所示应用场景下的应用。需要说明的是，这里提到的UE，均指当前接入到L个小区中某一个小区的用户。
每一个用户的大尺度衰落因子的公式如公式(1.1)所示：
Figure PCTCN2014090531-appb-000006
  公式(1.1)
其中，rjkl表示所述L个小区中的小区l中的第k个UE到小区j中基站的距离，γ为衰减指数，zjkl表示小区l中第k个UE到小区j中基站的对数正态随机变量，满足10log10(zjkl)～CN(0,σshadow)，βjkl表示小区l中的第k个UE到小区j所属基站的大尺度衰落因子，CN(0,σshadow)表示均值为0，方差为σshadow的高斯分布，σshadow表示对数高斯分布的阴影衰落的方差。
实际系统中，基站端的天线数目M是个定值，底噪必然存在，此时SINR其实不难获得，可以用SINR来计算系统的容量。当基站端的天线数目无限大，即M→∞时，噪声的影响可以忽略不计，小区j中第k个UE的等效信干比(Signal-to-Interference Ratio，SIR)如公式(1.2)所示
Figure PCTCN2014090531-appb-000007
  公式(1.2)
其中，
Figure PCTCN2014090531-appb-000008
表示小区j中第k个UE的等效信干比。
为简便起见，本发明实施例以SIR为例计算系统的容量。
对L个小区内的UE进行分组以得到该L个小区的一个用户序列。其中，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。这样可以保证L个小区共享总个数为K个的正交导频序列，且小区内各用户互不干扰。当然，在实际系统中，很多时候还可以采用正交性较好但不满足完全正交的准正交序列，这里就不做区分了。
在优化导频调度时，可以基于不同的优化目的获取不同的用户序列，再根据用户序列对L个小区的UE进行导频调度。不妨将实现优化目的的准则称为导频调度优化准则。
一种方式，协同设备可按照使得L个小区的系统和速率最大的准则搜索L个小区的用户序列。
根据用户序列的划分，进而，可以得到整个系统的和速率公式(1.3)：
Figure PCTCN2014090531-appb-000009
  公式(1.3)
系统的和速率可视为K个用户组合的和速率之和。同一用户组合的UE共享相同的一段导频序列，可以保证L个小区共享总个数为K个的正交导频序列，且小区内各用户互不干扰，可减少导频污染。
其中，用户序列中第m个用户组合Ωm的和速率可用公式(1.4)表示：
Figure PCTCN2014090531-appb-000010
  公式(1.4)
其中，
Figure PCTCN2014090531-appb-000011
表示Ωm的和速率。
此时导频调度优化准则是选择一种最优的用户序列Ωopt＝{Ω1,Ω2,…,ΩK}，使得系统的和速率rate(Ωopt)最大，该导频调度优化准则可用公式(1.5)表示：
Figure PCTCN2014090531-appb-000012
  公式(1.5)
根据公式(1.3)、公式(1.4)和公式(1.5)，协同设备可按照使得L个小区的系统和速率最大的准则，搜索L个小区的用户序列。
另一种方式，协同设备可根据使得L个小区内最小用户速率最大的准则搜索L个小区的用户序列。根据公式(1.2)求得的SIR，可求得每个用户的速率，一旦用户序列划分确定，就可以得到该种方式下的最小用户速率；遍历所有用户的序列组合，找到使得最小用户速率最大的那一种序列组合，即可将之确认为所求序列组合。
其中，用户序列中第m个用户组合Ωm的中的UE的最小速率可用公式(1.4)表示：
Figure PCTCN2014090531-appb-000013
  公式(1.6)
此时，导频调度优化准则是选择一种最优用户序列Ωopt＝{Ω1,Ω2,…,ΩK}，使得L 个小区内用户的最小速率最大，该导频调度优化准则可用公式(1.7)表示：
rate(Ωopt)＝argmax(minl＝1,2,…,Krate(Ωm))公式(1.7)
根据公式(1.6)和公式(1.7)，协同设备可按照使得L个小区内最小用户速率最大的准则，搜索L个小区的用户序列。
再一种方式，协同设备可根据满足L个小区内业务质量(Quality of Service，QoS)需求和用户最大速率需求的准则搜索L个小区的用户序列。
例如，在考虑QoS时，考虑用户所需要的最大速率，公式(1.4)可修改为公式(1.8)：
Figure PCTCN2014090531-appb-000014
  公式(1.8)
此时，导频调度优化准则的判决准则仍然采用公式(1.5)。与公式(1.4)相比，公式(1.8)引入了用户QoS的考虑因素。此处只举了一个最简单的例子，即用户所需要的最大速率R。每个用户所需求的最大速率不一样。比如，针对满负载的业务来说，R就是无穷大；而针对语音用户来说，R就是一个较小的固定值，不需要有非常高的信干比就可以满足需求。相应地，协同设备可按照用户的QoS需求和使得L个小区内用户和速率最大的准则来搜索L个小区的用户序列。
当然，协同设备还可根据其它的准则，搜索L个小区的用户序列，本发明实施例在此不作限定。
另外，需要说明的是，上述公式的计算均以L个小区中包含K个UE的例子进行说明，在实际的应用中本领域的技术人员可以对上述公式进行调整以得到合适的计算公式。
表1.1是本发明实施例L个小区的一种用户序列。
Figure PCTCN2014090531-appb-000015
如表1.1所示，其中Ω表示用户序列，其内包含用户组合Ω1、Ω2、…、Ωk、…、ΩK共K个用户组合。p_1、p_2、…、p_l、…、p_L分别表示该L个小区。小区包含的UE为表格中表示小区的符号右侧的元素的集合；组合中包含的UE为表格中表示小区的符号下方的元素的集合。
Figure PCTCN2014090531-appb-000016
表示属于第l个小区且属于用户序列的第k个用户组合的UE，其中1<l<L，1<k<K。
图2是本发明实施例贪婪搜索调度的方法流程图，图2的方法由多输入输出系统的协同设备执行。其中该系统包括L个小区，该L个小区的每一个小区最多存在K个UE。该方法包括：
201，获取该L个小区的多个用户组合。
其中该多个用户组合中的任一个用户组合用于构成该L个小区的用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合。
当L个小区下每个小区都存在K个UE时，L个小区的用户组合总共为KL个用户组合。
本发明实施例中，该多个用户组合可以是当L个小区的全部或部分用户组合。需要特别指出的是，本发明实施例中提到的UE为单天线UE，多天线UE可视为多个单天线UE。例如，四天线UE在本发明实施例中可视为4个单天线UE。
202，对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到优化用户序列。
其中该第一用户组合为该每一次搜索中能够加入该初始用户序列并且满足搜索条件的用户组合，该优化用户序列中的任一个UE只存在于该优化用户序列的一个用户组合中。
可选地，步骤202具体实现为，按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，该搜索条件用于使该优化用户序列符合该导频调度优化准则，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。当然，导频调度优化准则还可以是其它的导频调度优化准则，本发明实施例对此不作限制。
本发明中，该搜索条件为能够满足导频调度优化目的的条件。例如导频调度优化准则 为使得该L个小区的系统和速率最大的准则时，该搜索条件用于选出多个用户组合中和速率最大的用户组合。其中，用户组合的和速率为用户组合中所有UE的速率之和。又例如，导频调度优化准则为使得该L个小区中UE的最小速率最大的准则时，该搜索条件用于选出多个用户组合中UE的最小速率最大的用户组合。
203，根据该优化用户序列对该L个小区中的UE进行导频调度。
其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
具体地，可向L个小区所属的基站发送该优化用户序列以便基站根据该优化用户序列对基站管辖的小区的UE进行导频调度。或者，可向L个小区中的第一小区所属的基站发送该优化用户序列中第一小区对应的部分序列，以便该第一小区所属的基站根据该优化用户序列中第一小区对应的部分序列对第一小区的UE进行导频调度。需要说明的是，该第一小区可以是L个小区中的任一个小区。当然，还可能存在其它的调度方式，本发明实施例在此不作限制。
本发明实施例中，通过对该L个小区的多个用户组合进行搜索以得到L个小区的优化用户序列，并根据优化用户序列对L个小区的UE进行导频调度，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
可选地，作为一个实施例，当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，该按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列具体实现为：按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，其中该第一用户组合为该每一次搜索中能够加入该优化用户序列并且和速率最大的用户组合。
进一步地，按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列具体实现为：获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，并根据该大尺度衰落因子获取该L个小区的多个用户组合的和速率以形成该L个小区的多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；
重复执行步骤2.1和步骤2.2预设的次数，其中该预设的次数不大于K：
2.1.将该第一用户组合加入到该优化用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
2.2.将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
进一步地，该方法还包括：在贪婪搜索完成之后，如果加入该优化用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该优化用户序列中，以形成该优化用户序列。其中，当优化用户序列中的用户组合为K个时，优化用户序列的UE的合集等于L个小区的UE的合集。
具体地，该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。需要说明的是，该第一小区可以是L个小区中的任一个小区。
下面，将结合具体的实施例，对本发明实施例的方法做进一步的描述。
图3是本发明实施例贪婪搜索的具体流程图。图3的方法由多输入输出系统的协同设备执行。该协同设备可以设置在多输入输出系统的某个基站上，也可以是独立于基站的某个网元设备。图3所示的方法的一种应用场景可参见图1的描述，本发明实施例在此不再赘述。本发明实施例中，以使得L个小区内系统的和速率最大为准则为例，对L个小区的 用户组合进行贪婪搜索以得到优化用户序列。
301，获取大尺度衰落因子。
协同设备可采用合适的手段获取L个小区中每个小区的K个UE的大尺度衰落因子。例如，通过指示基站测量获得，或通过计算模型结合地图导航等计算获得，或者是其它方式，本发明实施例在此不作限制。
302，获取L个小区的多个用户组合的和速率值。
当L个小区中每个小区都存在K个UE时，按照每个小区选出一个UE形成一个用户组合的准则，L个小区中存在KL个不同的用户组合。
如果L个小区中某个小区存在小于K个的UE，例如存在S个UE，S<K，则此时该小区可视为还存在K-S个信号为0的UE，然后按照每个小区选出一个UE形成一个用户组合的准则，选择出KL个用户组合，再将用户组合中信号为0的UE剔除，并去掉重复的用户组合，可得到L个小区的所有用户组合。因此，L个小区中最多存在KL种不同的用户组合。
在进行贪婪搜索时，可对L个小区的所有用户组合进行贪婪搜索，也可对L个小区的部分用户组合进行贪婪搜索。
需要注意的是，如果是对L个小区的部分用户组合进行贪婪搜索，则最终可能只能够得到优化用户序列的部分用户组合，优化用户序列的其它用户组合需要通过其它方式确定，例如，从L个小区中不在优化用户序列当前组合的UE中随机选择UE形成优化用户序列的其它用户组合。
当导频调度优化准则是使得L个小区内系统的和速率最大的准则时，在对L个小区的多个用户组合进行贪婪搜索时，需要计算该多个用户组合的和速率值。该多个用户组合的和速率值可构成一个和速率集合，该和速率集合中的元素(和速率)与该多个用户组合中的用户组合一一对应。
本发明实施例中，可以用一个K维的Rate表存储该和速率集合的和速率。其中，该Rate表可以是一个K维数组，该Rate表的每个维度对应于L个小区中的一个小区，例如第1个维度对应于第1个小区，第l个维度对应于第l个小区，第L个维度对应于第L个小区，等等。每一个维度的一个下标对应于该维度所对应的小区的一个UE，例如第1个小区有K个UE，则第1个维度就有与该K个UE对应的K个下标，第l个小区有K-2个UE，则第l个维度就有与该K-2个UE对应的K-2个下标，等等。
此时，可根据用户组合中的UE，将用户组合的和速率值放到Rate表中对应的位置中。
当然，也不排除用其它形式的数据结构存储该和速率集合，例如指针链表，等等。
303，确定循环执行次数N、循环计数变量n置为0
确定需要执行贪婪搜索的循环执行次数N。如果N＝K，则该贪婪搜索为完整的贪婪搜索，最终搜索结果只有一个用户序列，如果N小于K，则最终搜索结果可能存在多个，可从中选择一个作为贪婪搜索的结果。
将循环计数变量n置为0。
304，n++
循环计数变量累计加1。
具体地，可用一个计数器进行循环计数。
305，判断循环计数变量是否大于循环执行次数。
如果n>N，则执行步骤308。
如果n≤N，则执行步骤306。
306，取出当前和速率集合中和速率最大值对应的第一用户组合，加入到优化用户序列中。
以Rate表为例，通过比较Rate表中存储的和速率值，可获得和速率最大的元素Ωm，即：
Figure PCTCN2014090531-appb-000017
取出Ωm对应的第一用户组合，即为本轮迭代取出的用户组合，加入优化用户序列中。
307，将和速率集合中包含第一用户组合的任一个UE的第二用户组合对应的和速率值删除或置为0。
如步骤306所示，第一用户组合为
Figure PCTCN2014090531-appb-000018
如果Rate表中的元素对应的第二用户组合包含
Figure PCTCN2014090531-appb-000019
中的一个或多个UE，则将该第二元素删除或置为0。
具体的，在Rate表中，可将该Rate表中包含该Ωm中的每一个UE对应的维下标的元素删除或置为0。
执行步骤304。
308，输出优化用户序列。
如果得到的优化用户序列中包含K个用户组合，则此时的优化用户序列为完整的用户序列，可直接输出。
如果得到的优化用户序列中用户组合小于K个，则此时的优化用户序列为不完整的用户序列，还需要通过其它方式获取优化用户序列中剩余的用户组合。例如，可从L个小区中随机选择第一用户组合加入优化用户序列，直到优化用户序列中的用户组合为K个，其中第一用户组合中的UE尚未出现在优化用户序列的UE中。当然，在选择用户组合时，要需要满足一个条件，即当优化用户序列中的用户组合为K个时，优化用户序列的UE的合集等于L个小区的UE的合集。
得到优化用户序列后，可根据优化用户序列对该L个小区进行导频调度。其中，L个小区中属于同一用户组合的UE共享相同的一段导频序列。
当然，图3只是本发明实施例的一种具体实现方式，还可对图3的步骤部分调整以得到搜索的结果，例如，步骤303可放在步骤301、302之前，循环计数器可从0开始计数，等等，本发明实施例在此并不做限制。
图6是本发明实施例贪婪搜索调度与随机调度的一个效果对比图。当小区个数为2时，贪婪搜索调度与随机调度下单个用户的一种速率差值比较效果如图6所示。
另外，贪婪搜索调度与随机调度下单个用户的速率差值的一个示例的具体效果如表3-1所示，其中，小区个数为2，UE个数分别为2、4、8、16、32。
表3-1：
Figure PCTCN2014090531-appb-000020
图7是本发明实施例贪婪搜索调度与随机调度的另一个效果对比图。当小区个数为3时，贪婪搜索调度与随机调度下单个用户的一种速率差值比较效果如图7所示。
另外，贪婪搜索调度与随机调度下单个用户的速率差值的一个示例的具体效果如表3-2所示，其中，小区个数为3，UE个数分别为2、4、8、16、32。表3-2：
Figure PCTCN2014090531-appb-000021
Figure PCTCN2014090531-appb-000022
需要说明的是图6、图7、表3-1、表3-2所示的效果中，L个小区中每个小区都包含K个UE，且贪婪搜索是针对L个小区的所有用户组合，贪婪搜索的循环执行次数都是K次。
从图6、图7、表3-1、表3-2可以看出，本发明实施例的方法，相对于随机调度在用户速率上有较明显的提升，小区内的用户数越大，提升效果越显著。
表3-3是本发明穷举搜索调度与贪婪搜索调度的复杂度比较。
表3-3：
Figure PCTCN2014090531-appb-000023
从表3-3可以看出，贪婪搜索调度的复杂度远远低于穷举搜索调度的复杂度。
另外，本发明实施例中，协同设备也可以其它导频调度优化准则对L个小区的用户序列进行搜索，例如，使得L个小区内的用户最小速率最大的准则，等等，本发明实施例在此不作限制。
图4是本发明实施例禁忌搜索调度的方法流程图，图4的方法由多输入输出系统的协同设备执行。其中该系统包括L个小区，该L个小区的每一个小区最多存在K个UE。该方法包括：
401，确定初始用户序列。
其中，该初始用户序列为该L个小区的一种用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合。
需要特别指出的是，本发明实施例中提到的UE为单天线UE，多天线UE可视为多个单天线UE。例如，四天线UE在本发明实施例中可视为4个单天线UE。
402，根据该初始用户序列进行禁忌搜索以获得优化用户序列。
其中，该优化用户序列为该L个小区的一种用户序列。
本发明实施例中，以初始用户序列为基础，进行禁忌搜索。
禁忌搜索中，首先需要指定L个小区中的交换小区，然后对初始用户序列中属于交换小区的UE进行交换以获得邻域序列，通过比较得到一个历史最优用户序列。每一个交换小区，邻域交换可以重复执行多次；执行交换的小区也可以有多个。例如，在本发明实施例中，可逐一指定该L个小区中的一个小区作为交换小区。
可选地，步骤402具体实现为：根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。当然，导频调度优化准则还可以是其它的导频调度优化准则，本发明实施例对此不作限制。
403，根据该优化用户序列对该L个小区中的UE进行导频调度，其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
具体地，可向L个小区所属的基站发送该优化用户序列以便基站根据该优化用户序列对基站管辖的小区的UE进行导频调度。或者，可向L个小区中的第一小区所属的基站发送该优化用户序列中第一小区对应的部分序列，以便该第一小区所属的基站根据该优化用户序列中第一小区对应的部分序列对第一小区的UE进行导频调度。当然，还可能存在其它的调度方式，本发明实施例在此不作限制。
本发明实施例中，通过对L个小区的初始用户序列进行搜索以得到优化用户序列，并根据优化用户序列对L个小区进行导频调度，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
可选地，作为一个实施例，该方法还包括：获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，其中，该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子用于确定该L个小区中的UE的速率，该系统的和速率为该L个小区中所有接入UE的速率之和。
可选地，作为一个实施例，当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，该根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列具体可以实现为：
将该初始用户序列赋值给历史用户序列和当前用户序列；
对步骤4.1、步骤4.2循环执行N次后将该历史用户序列作为该优化用户序列，N为不大于L的正整数，其中
4.1、将搜索禁忌表置空；
4.2、对步骤4.2.1、4.2.2执行预定的次数，其中
4.2.1、将待交换小区在该当前用户序列中任意X个用户组合的UE进行位置交换得到该当前用户序列的多个邻域序列，并获得该当前用户序列的多个邻域序列对应的系统和速率，并取出使得系统和速率最大的第一邻域序列，其中，该待交换小区为循环执行N次的过程中的第l次循环执行选定的小区，该N次的循环执行的每一次选定的小区均不相同，该系统和速率由该大尺度衰落因子确定，X为大于1且不大于K的正整数；
4.2.2、如果该第一邻域序列满足该禁忌搜索的特赦准则，则将该第一邻域序列赋值给该历史用户序列和当前用户序列，并将该第一邻域序列加入该搜索禁忌表中，或者如果该第一邻域序列不满足禁忌搜索的特赦准则，则将该当前用户序列的多种邻域序列中不在该搜索禁忌表中且系统和速率最大的第二邻域序列赋值给当前用户序列，并将该第二邻域序列加入该搜索禁忌表中，其中，该特赦准则为该第一邻域序列的系统和速率大于该历史用户序列的和速率，或者该特赦准则为该第一邻域序列的系统和速率大于或等于该历史用户序列的和速率。
进一步地，X的取值为2，即在获取用户序列的邻域时只对同一小区的2个UE进行交换。进一步地，该预定的次数为K次。将一轮禁忌搜索的最大迭代次数设为K次，可在导频调度效果和算法性能中折中取得一定平衡。
具体地，该L个小区的用户序列对应的系统和速率由如下公式(4.1)表示：
Figure PCTCN2014090531-appb-000024
  公式(4.1)，
其中，rate(Ωopt)表示该L个小区的用户序列对应的系统和速率，Ωk表示该L个小区的用户序列中第k个用户组合，rate(Ωk)表示该第k个用户组合的和速率，
Figure PCTCN2014090531-appb-000025
βjkl表示属于该L个小区中的小区l且属于该第 k个用户组合的UE到小区j所属基站的大尺度衰落因子。
具体地，L个小区的大尺度衰落因子βjkl可用公式(4.2)表示:
Figure PCTCN2014090531-appb-000026
  公式(4.2)，
其中，rjkl表示所述L个小区中小区l中的第k个UE到小区j中基站的距离，γ为衰减指数，zjkl表示小区l中的第k个UE到小区j中基站对数正态随机变量，满足10log10(zjkl)～CN(0,σshadow)，βjkl表示小区l中的第k个UE到小区j中基站的大尺度衰落因子，CN(0,σshadow)表示均值为0，方差为σshadow的高斯分布，σshadow表示对数高斯分布的阴影衰落的方差。
使得用户序列对应的系统和速率最大的准则可用公式(4.3)表示：
Figure PCTCN2014090531-appb-000027
  公式(4.3)。
可选地，作为一个实施例，确定初始用户序列具体实现为：随机确定该L个小区的一个用户序列作为该初始用户序列。
可选地，作为另一个实施例，确定初始用户序列具体实现为：根据使得该初始用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到该初始用户序列，其中该第一用户组合为该每一次搜索中能够加入该初始用户序列并且和速率最大的用户组合，该初始用户序列中的任一个UE只存在于该初始用户序列的一个用户组合中。
具体地，根据使得该初始用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索得到的局部最优用户组合加入到该初始用户序列可实现为：
根据该大尺度衰落因子获取该多个用户组合各自的和速率以形成该多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；
重复执行步骤4.3和步骤4.4预设的次数，其中该预设的次数不大于K：
4.3.将该第一用户组合加入到该初始用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
4.4将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
进一步地，该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。需要说明的是，该第一小区可以是L个小区中的任一个小区。
可选地，确定初始用户序列还可实现为：在该贪婪搜索完成之后，如果加入该初始用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该初始用户序列中，以形成该初始用户序列。
可选地，L个小区的多个用户组合为该L个小区的所有用户组合中的全部或部分用户组合。
下面，将结合具体的实施例，对本发明实施例的方法做进一步的描述。
图5是本发明实施例禁忌搜索的具体流程图。图5的方法由多输入输出系统的协同设备执行。该协同设备可以设置在多输入输出系统的某个基站上，也可以是独立于基站的某个网元设备。图5所示的方法的一种应用场景可参见图1的描述，本发明实施例在此不再赘述。本发明实施例中，以使得L个小区内系统的和速率最大为准则为例，对L个小区的用户序列进行搜索。
501，获取大尺度衰落因子。
协同设备可根据公式(1.1)
Figure PCTCN2014090531-appb-000028
获取L个小区内每个UE的大尺度衰落因子。
当然，协同设备也可通过其它方式获取该大尺度衰落因子，本发明实施例在此不作限制。
502，初始化参数。
具体地，协同设备可初始化禁忌搜索的初始用户序列(index)。
协同设备可随机指定L个小区的一个用户序列作为初始用户序列index，或者，协同设备也可将图3所示根据贪婪搜索算法得出的用户序列作为初始用户序列index。
另外，协同设备还可初始化循环搜索次数，该初始化循环搜索次数最多为小区的个数L，当然，循环搜索次数也可以小于L次，只是在初始序列相同，交换小区搜索顺序相同的情况下，循环搜索次数小于L次的搜索得到的用户序列一般情况下会比循环搜索次数为L次得到的用户序列的导频调度效果差，最多与循环搜索次数为L次得到的用户序列的导频调度效果相同。本发明实施例以L次对本发明的方法进行描述。
另外，协同设备还可初始化邻域跌代次数Niter。Niter的取值并不作具体的限制。从计算开销和最终效果两方面均衡考虑，Niter取值可设为K。
另外，循环搜索计数变量l置为0，历史最优用户序列P*置为index。
503，循环搜索计数变量加1。
每次循环搜索，循环搜索计数变量l累计加1。
具体的，可用一个计数器进行循环搜索计数。
504，判断循环搜索计数变量是否大于循环搜索次数。
如果l>L，则禁忌搜索结束，执行步骤514。
如果l≤L，则执行步骤505。
505，初始化每轮循环搜索的目标优化序列、禁忌表及邻域迭代计数变量。
具体的，协同设备可将历史最优用户序列P*赋值给目标优化序列P，同时，将禁忌表T置为空，将邻域迭代计数变量c置为0。
另外，协同设备可选中L个小区中的第l个小区作为UE的待交换小区，该小区具体可以是L个小区中，之前搜索选中作为交换小区以外的任一个小区。
506，邻域迭代计数变量加1。
每次进行邻域交换，邻域迭代计数变量c累计加1。
与步骤503类似，可用一个计数器进行邻域迭代计数。
507，判断邻域迭代计数变量是否大于邻域迭代次数。
如果c>Niter，则本次循环搜索结束，执行步骤503。
如果c≤Niter，则执行步骤508。
508，将P的最优邻域序列赋值给P^。
首先，可获取P的所有邻域序列。以待交换小区存在K个UE为例，在进行UE交换时，可将待交换小区的UE在P中2个用户组合的位置进行交换，此时可得到P的
Figure PCTCN2014090531-appb-000029
种邻域序列；如果将待交换小区的UE在P中3个用户组合的位置进行交换，可得到P的
Figure PCTCN2014090531-appb-000030
种邻域序列；如果将待交换小区的UE在P中X个用户组合的位置进行交 换，可得到P的
Figure PCTCN2014090531-appb-000031
种邻域序列，其中X为大于1且不大于K的正整数。当然，如果待交换小区存在的UE小于K，则得到的所有邻域序列个数可能小于上述对应的个数值。一般情况下，X取值为2即可保证获得较好的用户序列，且需要获取的邻域序列也不会太多。本发明实施例X取值以2为例。
其次，取出P的最优邻域序列。每一个用户序列对应于系统的一种和速率。根据公式(1.3)
Figure PCTCN2014090531-appb-000032
或公式(1.4)
Figure PCTCN2014090531-appb-000033
可获取P的所有邻域序列的系统和速率，进而可取得P的最优邻域序列。
将P的最优邻域序列赋值给P^
509，判断P^是否满足特赦准则。
如果P^满足特赦准则，则执行步骤510。
如果P^不满足特赦准则，则执行步骤511。
其中，特赦准则用公式(5.1)表示：
Figure PCTCN2014090531-appb-000034
  公式(5.1)
其物理含义为，目标优化序列的最优邻域序列的系统和速率大于历史最优用户序列的系统和速率。
510，将P^赋值给目标优化序列和历史最优用户序列。
将P^赋值给目标优化序列P，作为下一次邻域迭代的初始序列。
将P^赋值给历史最优用户序列P*。
执行步骤513。
511，将P的邻域序列中不在禁忌表T的最优邻域序列赋值给P^。
根据步骤508中的所有邻域序列，可取出不在禁忌表T的最优邻域序列赋值给P^，具体可用公式(5.2)表示：
Figure PCTCN2014090531-appb-000035
  公式(5.2)
512，将P^赋值给目标优化序列。
将P^赋值给目标优化序列P，作为下一次邻域迭代的初始序列。
执行步骤513。
513，将P^加入禁忌表T。
将P^加入禁忌表T，以便在下一次邻域迭代将P^排除。
执行步骤506。
上述步骤501至步骤514是对L个小区的用户序列进行禁忌搜索的一种具体实现。当然，也可对其中的步骤进行调整，例如，可将禁忌表置为空的步骤放在每一次循环结束之后，计数器可从0开始计数，等等。基于与图5的步骤类似的思路实现的步骤也属于本发明保护的范围。
下面以三小区导频分配为例，假定L＝3，K＝8。初始化所有小区的用户指数向量都是：p_l＝[1，2，3，4，5，6，7，8]，l＝{1，……，L}，令Τ＝Φ。从第一个小区开始进行逐小区优化，以第一个小区为例，优化目标向量为P＝p_l，迭代过程中保持其他小区指数向量 不变，计算此时的系统和速率f(index)＝120，令历史最优解P*＝P^，交换P中任意两个用户指数的位置得到P的
Figure PCTCN2014090531-appb-000036
中邻域序列N(P)，计算所有邻域序列对应的目标函数值。如表5-1所示。
表5-1：
Figure PCTCN2014090531-appb-000037
当前邻域中最优序列P^＝[8，2，3，4，5，6，7，1]，对应适配值135，满足以下公式：
Figure PCTCN2014090531-appb-000038
即满足特赦准则，优于历史最优解，不论P^是否在禁忌表中，都更新历史最优解P*＝P^和当前解P＝P^，将P加入禁忌表T中；第一次迭代后的当前解P＝[8，2，3，4，5，6，7，1]，邻域候选解状态如图2-2所示，此时当前邻域中最优序列P^＝[7，2，3，4，5，6，8，1]，对应适配值125小于135，不满足特赦准则且不在禁忌表中，那么只更新当且解P＝P^，保持历史最优解P*不变，将P^加入禁忌表T中。按照上述方式进行迭代，直到达到预先设定的最大迭代次数Niter，迭代过程中，如果P^已经在禁忌表中，那么应当选择次优于P^且不在T中的候选解来更新当前解。
第一次迭代后的邻域候选解状态如表5-2所示。
表5-2：
Figure PCTCN2014090531-appb-000039
图8是本发明实施例禁忌搜索调度与随机调度的一个效果对比图。当小区个数为2时，禁忌搜索(TS)调度与随机调度下单个用户的一种速率差值比较效果如图8所示。
另外，禁忌搜索调度、贪婪的禁忌搜索调度与随机调度下单个用户的速率差值的一个示例的具体效果如表5-3所示，其中，小区个数为2，UE个数分别为2、4、8、16、32。
表5-3：
Figure PCTCN2014090531-appb-000040
图9是本发明实施例禁忌搜索调度与随机调度的另一个效果对比图。当小区个数为3时，禁忌(TS)调度与随机调度下单个用户的一种速率差值比较效果如图9所示。
另外，禁忌搜索调度、贪婪的禁忌搜索调度与随机调度下单个用户的速率差值的一个示例的具体效果如表5-4所示，其中，小区个数为3，UE个数分别为2、4、8、16、32。
表5-4：
Figure PCTCN2014090531-appb-000041
另外，需要说明的是，图8、图9、表5-3、表5-4所示的效果中，L个小区中每个小区都包含K个UE，贪婪的TS调度中贪婪搜索部分针对的是L个小区的所有用户组合，且贪婪搜索部分的循环执行次数都是K次。
从图8、图9、表5-3、表5-4可以看出，本发明实施例的方法，相对于随机调度在用户速率上有较明显的提升，小区内的用户数越大，提升效果越显著。
表5-5是本发明穷举搜索调度与TS搜索调度、贪婪的TS搜索调度的复杂度比较。
表5-5：
Figure PCTCN2014090531-appb-000042
从表5-5可以看出，TS搜索调度、贪婪的TS搜索调度的复杂度远远低于穷举搜索调度的复杂度。
另外，本发明实施例中，协同设备也可以其它导频调度优化准则对L个小区的用户序列进行搜索，例如，使得L个小区内的用户最小速率最大的准则，等等，本发明实施例在此不作限制。
图10是本发明实施例多输入输出系统的协同设备1000的结构示意图。其中该系统包括L个小区，该L个小区的每一个小区最多存在K个UE。协同设备1000可包括：获取单元1001、搜索单元1002和调度单元1003。
获取单元1001，用于获取该L个小区的多个用户组合。
其中该多个用户组合中的任一个用户组合用于构成该L个小区的用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合。
当L个小区下每个小区都存在K个UE时，L个小区的用户组合总共为KL个用户组合。
本发明实施例中，该多个用户组合可以是当L个小区的全部或部分用户组合。需要特别指出的是，本发明实施例中提到的UE为单天线UE，多天线UE可视为多个单天线UE。例如，四天线UE在本发明实施例中可视为4个单天线UE。
搜索单元1002，用于对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到优化用户序列。
其中该第一用户组合为该每一次搜索中能够加入该初始用户序列并且满足搜索条件的用户组合，该优化用户序列中的任一个UE只存在于该优化用户序列的一个用户组合中。
调度单元1003，用于根据该优化用户序列对该L个小区中的UE进行导频调度。
其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
具体地，调度单元1003可向L个小区所属的基站发送该优化用户序列以便基站根据该优化用户序列对基站管辖的小区的UE进行导频调度。或者，调度单元1003可向L个小区中的第一小区所属的基站发送该优化用户序列中第一小区对应的部分序列，以便该第一小区所属的基站根据该优化用户序列中第一小区对应的部分序列对第一小区的UE进行导频调度。当然，还可能存在其它的调度方式，本发明实施例在此不作限制。
本发明实施例中，协同设备1000通过对该L个小区的多个用户组合进行搜索以得到L个小区的优化用户序列，并根据优化用户序列对L个小区的UE进行导频调度，能够在算 法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
可选地，协同设备1000可以是L个小区中某一小区所属的基站，也可以是管辖L个小区的一个网元设备，或者是独立于L个小区中任意一个小区所属的基站的网元设备。可选地，搜索单元1002具体用于按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，该搜索条件用于使该优化用户序列符合该导频调度优化准则，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。当然，导频调度优化准则还可以是其它的导频调度优化准则，本发明实施例对此不作限制。
可选地，作为一个实施例，当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，搜索单元1002具体用于按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，其中该第一用户组合为该每一次搜索中能够加入该优化用户序列并且和速率最大的用户组合。
进一步地，获取单元1001还用于获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，并根据该大尺度衰落因子获取该L个小区的多个用户组合的和速率以形成该L个小区的多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；在用于按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，搜索单元1002具体用于重复执行步骤10.1和步骤10.2预设的次数，其中该预设的次数不大于K：
10.1.将该第一用户组合加入到该优化用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
10.2.将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
进一步地，搜索单元1002还用于在贪婪搜索完成之后，如果加入该优化用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该优化用户序列中，以形成该优化用户序列。
具体地，该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。需要说明的是，该第一小区可以是L个小区中的任一个小区。
协同设备1000还可执行图2的方法，并实现协同设备在图2、图3所示实施例中的功能，其具体实现可参考图2、图3所示的实施例，本发明实施例在此不再赘述。
图11是本发明实施例多输入输出系统的协同设备1100的结构示意图。其中该系统包括L个小区，该L个小区的每一个小区最多存在K个用户设备UE。协同设备1100可包括确定单元1101、搜索单元1102和调度单元1103。
确定单元1101，用于确定初始用户序列。
其中，该初始用户序列为该L个小区的一种用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合。
需要特别指出的是，本发明实施例中提到的UE为单天线UE，多天线UE可视为多个 单天线UE。例如，四天线UE在本发明实施例中可视为4个单天线UE。
搜索单元1102，用于根据该初始用户序列进行禁忌搜索以获得优化用户序列。
其中，该优化用户序列为该L个小区的一种用户序列。
本发明实施例中，以初始用户序列为基础，进行禁忌搜索。
禁忌搜索中，首先需要指定L个小区中的交换小区，然后对初始用户序列中属于交换小区的UE进行交换以获得邻域序列，通过比较得到一个历史最优用户序列。每一个交换小区，邻域交换可以重复执行多次；执行交换的小区也可以有多个。例如，在本发明实施例中，可逐一指定该L个小区中的一个小区作为交换小区。
调度单元1103，用于根据该优化用户序列对该L个小区中的UE进行导频调度，其中，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
具体地，调度单元1103可向L个小区所属的基站发送该优化用户序列以便基站根据该优化用户序列对基站管辖的小区的UE进行导频调度。
或者，调度单元1103可向L个小区中的第一小区所属的基站发送该优化用户序列中第一小区对应的部分序列，以便该第一小区所属的基站根据该优化用户序列中第一小区对应的部分序列对第一小区的UE进行导频调度。
本发明实施例中，协同设备1100通过对L个小区的初始用户序列进行搜索以得到优化用户序列，并根据优化用户序列对L个小区进行导频调度，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
可选地，协同设备1100可以是L个小区中某一小区所属的基站，也可以是管辖L个小区的一个网元设备，或者是独立于L个小区中任意一个小区所属的基站的网元设备。
可选地，在用于根据该初始用户序列进行禁忌搜索以获得优化用户序列，搜索单元1102具体用于根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。当然，导频调度优化准则还可以是其它的导频调度优化准则，本发明实施例对此不作限制。
可选地，作为一个实施例，协同设备1100还可包括获取单元1104。获取单元1104，用于获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，其中，该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子用于确定该L个小区中的UE的速率，该系统的和速率为该L个小区中所有UE的速率之和。
进一步地，当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，在用于根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，搜索单元1102具体用于：
将该初始用户序列赋值给历史用户序列和当前用户序列；
对步骤11.1、步骤11.2循环执行N次后将该历史用户序列作为该优化用户序列，N为不大于L的正整数，其中
11.1、将搜索禁忌表置空；
11.2、对步骤11.2.1、步骤11.2.2执行预定的次数，其中
11.2.1、将待交换小区在该当前用户序列中任意X个用户组合的UE进行位置交换得到该当前用户序列的多个邻域序列，并获得该当前用户序列的多个邻域序列对应的系统和速率，并取出使得系统和速率最大的第一邻域序列，其中，该待交换小区为循环执行N次的过程中的第l次循环执行选定的小区，该N次的循环执行的每一次选定的小区均不相同，该系统和速率由该大尺度衰落因子确定，X为大于1且不大于K的正整数；
11.2.2、如果该第一邻域序列满足该禁忌搜索的特赦准则，则将该第一邻域序列赋值给该历史用户序列和当前用户序列，并将该第一邻域序列加入该搜索禁忌表中，或者如果该 第一邻域序列不满足禁忌搜索的特赦准则，则将该当前用户序列的多种邻域序列中不在该搜索禁忌表中且系统和速率最大的第二邻域序列赋值给当前用户序列，并将该第二邻域序列加入该搜索禁忌表中，其中，该特赦准则为该第一邻域序列的系统和速率大于该历史用户序列的和速率，或者该特赦准则为该第一邻域序列的系统和速率大于或等于该历史用户序列的和速率。
进一步地，X的取值为2，即在获取用户序列的邻域时只对同一小区的2个UE进行交换。进一步地，该预定的次数为K次。将一轮禁忌搜索的最大迭代次数设为K次，可在导频调度效果和算法性能中折中取得一定平衡。
具体地，该L个小区的用户序列对应的系统和速率由如下公式(11.1)表示：
Figure PCTCN2014090531-appb-000043
  公式(11.1)，
其中，rate(Ωopt)表示该L个小区的用户序列对应的系统和速率，Ωk表示该L个小区的用户序列中第k个用户组合，rate(Ωk)表示该第k个用户组合的和速率，
Figure PCTCN2014090531-appb-000044
βjkl表示属于该L个小区中的小区l且属于该第k个用户组合的UE到小区j所属基站的大尺度衰落因子。
具体地，L个小区的大尺度衰落因子βjkl可用公式(11.2)表示:
Figure PCTCN2014090531-appb-000045
  公式(11.2)，
其中，rjkl表示所述L个小区中小区l中的第k个UE到小区j中基站的距离，γ为衰减指数，zjkl表示小区l中的第k个UE到小区j中基站对数正态随机变量，满足10log10(zjkl)～CN(0,σshadow)，βjkl表示小区l中的第k个UE到小区j中基站的大尺度衰落因子，CN(0,σshadow)表示均值为0，方差为σshadow的高斯分布，σshadow表示对数高斯分布的阴影衰落的方差。
使得用户序列对应的系统和速率最大的准则可用公式(11.3)表示：
  公式(11.3)。
可选地，作为一个实施例，在用于确定初始用户序列，确定单元1101具体用于随机确定该L个小区的一个用户序列作为该初始用户序列。
可选地，作为另一个实施例，搜索单元1102还用于根据使得贪婪搜索用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到该贪婪搜索用户序列，其中该第一用户组合为该每一次搜索中能够加入该贪婪搜索用户序列并且和速率最大的用户组合，该贪婪搜索用户序列中的任一个UE只存在于该贪婪搜索用户序列的一个用户组合中；确定单元1101具体用于确定该贪婪搜索用户序列为该初始用户序列。
进一步地，获取单元1104还用于根据该大尺度衰落因子获取该多个用户组合各自的和速率以形成该多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的 多个用户组合一一对应；在用于根据使得贪婪搜索用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到贪婪搜索用户序列，搜索单元1102具体用于重复执行步骤11.3和步骤11.4预设的次数，其中该预设的次数不大于K：
11.3.将该第一用户组合加入到该初始用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
11.4将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
进一步地，确定单元1101还用于在该贪婪搜索完成之后，如果加入该初始用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该初始用户序列中，以形成该初始用户序列。
具体地，该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。需要说明的是，该第一小区可以是L个小区中的任一个小区。
可选地，L个小区的多个用户组合为该L个小区的所有用户组合中的全部或部分用户组合。
协同设备1100还可执行图4的方法，并实现协同设备在图4、图5所示实施例中的功能，其具体实现可参考图4、图5所示的实施例，本发明实施例在此不再赘述。
图12是本发明实施例协同设备1200的结构示意图。其中该系统包括L个小区，该L个小区的每一个小区最多存在K个UE。协同设备1200可包括接收器1201、发射器1203、处理器1202和存储器1204。
接收器1201、发射器1203、处理器1202和存储器1204通过总线1205系统相互连接。总线1205可以是ISA总线、PCI总线或EISA总线等。所述总线可以分为地址总线、数据总线、控制总线等。为便于表示，图12中仅用一个双向箭头表示，但并不表示仅有一根总线或一种类型的总线。
存储器1204，用于存放程序。具体地，程序可以包括程序代码，所述程序代码包括计算机操作指令。存储器1204可以包括只读存储器和随机存取存储器，并向处理器1202提供指令和数据。存储器1204可能包含高速RAM存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个磁盘存储器。
处理器1202，执行存储器1204所存放的程序，用于获取该L个小区的多个用户组合，对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到优化用户序列，并根据该优化用户序列对该L个小区中的UE进行导频调度。
其中，该多个用户组合中的任一个用户组合用于构成该L个小区的用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合。另外，该第一用户组合为该每一次搜索中能够加入该初始用户序列并且满足搜索条件的用户组合，该优化用户序列中的任一个UE只存在于该优化用户序列的一个用户组合中。并且，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
需要特别指出的是，本发明实施例中提到的UE为单天线UE，多天线UE可视为多个单天线UE。例如，四天线UE在本发明实施例中可视为4个单天线UE。
当L个小区下每个小区都存在K个UE时，L个小区的用户组合总共为KL个用户组合。
本发明实施例中，该多个用户组合可以是当L个小区的全部或部分用户组合。
具体地，在用于根据该优化用户序列对该L个小区中的UE进行导频调度时，处理器1202可向L个小区所属的基站发送该优化用户序列以便基站根据该优化用户序列对基站管辖的小区的UE进行导频调度；或者，处理器1202可向L个小区中的第一小区所属的基站发送该优化用户序列中第一小区对应的部分序列，以便该第一小区所属的基站根据该优化用户序列中第一小区对应的部分序列对第一小区的UE进行导频调度。需要说明的是，该第一小区可以是L个小区中的任一个小区。当然，还可能存在其它的调度方式，本发明实施例在此不作限制。
上述如本发明图2、图3任一实施例揭示的协同设备执行的方法可以应用于处理器1202中，或者由处理器1202实现。处理器1202可能是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法的各步骤可以通过处理器1202中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器1202可以是通用处理器，包括中央处理器(Central Processing Unit，简称CPU)、网络处理器(Network Processor，简称NP)等；还可以是数字信号处理器(DSP)、专用集成电路(ASIC)、现成可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本发明实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本发明实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器1204，处理器1202读取存储器1204中的信息，结合其硬件完成上述方法的步骤。
本发明实施例中，协同设备1200通过对该L个小区的多个用户组合进行搜索以得到L个小区的优化用户序列，并根据优化用户序列对L个小区的UE进行导频调度，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
可选地，协同设备1200可以是L个小区中某一小区所属的基站，也可以是管辖L个小区的一个网元设备，或者是独立于L个小区中任意一个小区所属的基站的网元设备。
可选地，在用于对该L个小区的多个用户组合进行贪婪搜索以得到优化用户序列，处理器1202具体用于按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，该搜索条件用于使该优化用户序列符合该导频调度优化准则，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。当然，导频调度优化准则还可以是其它的导频调度优化准则，本发明实施例对此不作限制。
可选地，作为一个实施例，当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，在用于按照导频调度优化准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，处理器1202具体用于按照使得该L个小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，其中该第一用户组合为该每一次搜索中能够加入该优化用户序列并且和速率最大的用户组合。
进一步地，处理器1202可通过接收器1201和发射器1203获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，并根据该大尺度衰落因子获取该L个小区的多个用户组合的和速率以形成该L个小区的多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应。在用于在用于按照使得该L个 小区的系统和速率最大的准则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的第一用户组合加入到优化用户序列，处理器1202具体用于重复执行步骤12.1和步骤12.2预设的次数，其中该预设的次数不大于K：
12.1.将该第一用户组合加入到该优化用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
12.2.将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
进一步地，处理器1202还用于在贪婪搜索完成之后，如果加入该优化用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该优化用户序列中，以形成该优化用户序列。
具体地，该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。需要说明的是，该第一小区可以是L个小区中的任一个小区。
可选地，在用于获取该L个小区的多个用户组合，处理器1202具体用于获取该L个小区的最多KL个用户组合中的全部或部分用户组合。
协同设备1200还可执行图2的方法，并实现协同设备在图2、图3所示实施例中的功能，其具体实现可参考图2、图3所示的实施例，本发明实施例在此不再赘述。
图13是本发明实施例协同设备1300的结构示意图。其中该系统包括L个小区，该L个小区的每一个小区最多存在K个用户设备UE。协同设备1300可包括接收器1301、发射器1303、处理器1302和存储器1304。
接收器1301、发射器1303、处理器1302和存储器1304通过总线1305系统相互连接。总线1305可以是ISA总线、PCI总线或EISA总线等。所述总线可以分为地址总线、数据总线、控制总线等。为便于表示，图13中仅用一个双向箭头表示，但并不表示仅有一根总线或一种类型的总线。
存储器1304，用于存放程序。具体地，程序可以包括程序代码，所述程序代码包括计算机操作指令。存储器1304可以包括只读存储器和随机存取存储器，并向处理器1302提供指令和数据。存储器1304可能包含高速RAM存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个磁盘存储器。
处理器1302，执行存储器1304所存放的程序，用于确定初始用户序列，根据该初始用户序列进行禁忌搜索以获得优化用户序列，并根据该优化用户序列对该L个小区中的UE进行导频调度。
其中，该初始用户序列为该L个小区的一种用户序列，每个用户序列包括该L个小区的K个用户组合，每个用户组合最多包含L个UE，该K个用户组合中的每个用户组合中的UE分别属于该L个小区中不同的小区，该L个小区中的每个小区中的UE分别属于该K个用户组合中不同的用户组合，该优化用户序列为该L个小区的一种用户序列，该L个小区中属于该优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
需要特别指出的是，本发明实施例中提到的UE为单天线UE，多天线UE可视为多个单天线UE。例如，四天线UE在本发明实施例中可视为4个单天线UE。
本发明实施例中，以初始用户序列为基础，进行禁忌搜索。
禁忌搜索中，首先需要指定L个小区中的交换小区，然后对初始用户序列中属于交换小区的UE进行交换以获得邻域序列，通过比较得到一个历史最优用户序列。每一个交换小区，邻域交换可以重复执行多次；执行交换的小区也可以有多个。例如，在本发明实施例中，可逐一指定该L个小区中的一个小区作为交换小区。
具体地，处理器1302在根据该优化用户序列对该L个小区中的UE进行导频调度时， 可向L个小区所属的基站发送该优化用户序列以便基站根据该优化用户序列对基站管辖的小区的UE进行导频调度。或者，处理器1302在根据该优化用户序列对该L个小区中的UE进行导频调度时，可向L个小区中的第一小区所属的基站发送该优化用户序列中第一小区对应的部分序列，以便该第一小区所属的基站根据该优化用户序列中第一小区对应的部分序列对第一小区的UE进行导频调度。当然，导频调度优化准则还可以是其它的导频调度优化准则，本发明实施例对此不作限制。
上述如本发明图4、图5任一实施例揭示的协同设备执行的方法可以应用于处理器1302中，或者由处理器1302实现。处理器1302可能是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法的各步骤可以通过处理器1302中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器1302可以是通用处理器，包括中央处理器(Central Processing Unit，简称CPU)、网络处理器(Network Processor，简称NP)等；还可以是数字信号处理器(DSP)、专用集成电路(ASIC)、现成可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本发明实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本发明实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器1304，处理器1302读取存储器1304中的信息，结合其硬件完成上述方法的步骤。
本发明实施例中，协同设备1300通过对L个小区的初始用户序列进行搜索以得到优化用户序列，并根据优化用户序列对L个小区进行导频调度，能够在算法复杂度较低的情况下取得较好的导频调度效果，均衡考虑多输入输出系统的计算开销及综合导频效果。
可选地，协同设备1300可以是L个小区中某一小区所属的基站，也可以是管辖L个小区的一个网元设备，或者是独立于L个小区中任意一个小区所属的基站的网元设备。
可选地，在用于根据该初始用户序列进行禁忌搜索以获得优化用户序列，处理器1302具体用于根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，该导频调度优化准则包括以下一种准则：使得该L个小区的系统和速率最大的准则；或使得该L个小区中UE的最小速率最大的准则；或使得该L个小区的系统和速率最大的准则且满足该L个小区的UE的服务质量QoS的速率需求的准则。当然，导频调度优化准则还可以是其它的导频调度优化准则，本发明实施例对此不作限制。
可选地，作为一个实施例，处理器1302还可通过接收器1301和发射器1303获取该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子，其中，该L个小区中每个UE到该L小区的其它小区的基站的大尺度衰落因子用于确定该L个小区中的UE的速率，该系统的和速率为该L个小区中所有UE的速率之和。
进一步地，作为一个实施例，当该导频调度优化准则为使得该L个小区的系统和速率最大的准则，在用于根据该初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，处理器1302具体用于：
将该初始用户序列赋值给历史用户序列和当前用户序列；
对步骤13.1、步骤13.2循环执行N次后将该历史用户序列作为该优化用户序列，N为不大于L的正整数，其中
13.1、将搜索禁忌表置空；
13.2、对步骤13.2.1、步骤13.2.2执行预定的次数，其中，
13.2.1、将待交换小区在该当前用户序列中任意X个用户组合的UE进行位置交换得到该当前用户序列的多个邻域序列，并获得该当前用户序列的多个邻域序列对应的系统和速率，并取出使得系统和速率最大的第一邻域序列，其中，该待交换小区为循环执行N次 的过程中的第l次循环执行选定的小区，该N次的循环执行的每一次选定的小区均不相同，该系统和速率由该大尺度衰落因子确定，X为大于1且不大于K的正整数；
13.2.2、如果该第一邻域序列满足该禁忌搜索的特赦准则，则将该第一邻域序列赋值给该历史用户序列和当前用户序列，并将该第一邻域序列加入该搜索禁忌表中，或者如果该第一邻域序列不满足禁忌搜索的特赦准则，则将该当前用户序列的多种邻域序列中不在该搜索禁忌表中且系统和速率最大的第二邻域序列赋值给当前用户序列，并将该第二邻域序列加入该搜索禁忌表中，其中，该特赦准则为该第一邻域序列的系统和速率大于该历史用户序列的和速率，或者该特赦准则为该第一邻域序列的系统和速率大于或等于该历史用户序列的和速率。
进一步地，X的取值为2，即在获取用户序列的邻域时只对同一小区的2个UE进行交换。进一步地，该预定的次数为K次。将一轮禁忌搜索的最大迭代次数设为K次，可在导频调度效果和算法性能中折中取得一定平衡。
具体地，该L个小区的用户序列对应的系统和速率由如下公式(13.1)表示：
Figure PCTCN2014090531-appb-000047
  公式(13.1)，
其中，rate(Ωopt)表示该L个小区的用户序列对应的系统和速率，Ωk表示该L个小区的用户序列中第k个用户组合，rate(Ωk)表示该第k个用户组合的和速率，
Figure PCTCN2014090531-appb-000048
βjkl表示属于该L个小区中的小区l且属于该第k个用户组合的UE到小区j所属基站的大尺度衰落因子。
具体地，L个小区的大尺度衰落因子βjkl可用公式(13.2)表示:
Figure PCTCN2014090531-appb-000049
公式(13.2)，
其中，rjkl表示所述L个小区中小区l中的第k个UE到小区j中基站的距离，γ为衰减指数，zjkl表示小区l中的第k个UE到小区j中基站对数正态随机变量，满足10log10(zjkl)～CN(0,σshadow)，βjkl表示小区l中的第k个UE到小区j中基站的大尺度衰落因子，CN(0,σshadow)表示均值为0，方差为σshadow的高斯分布，σshadow表示对数高斯分布的阴影衰落的方差。
使得用户序列对应的系统和速率最大的准则可用公式(13.3)表示：
Figure PCTCN2014090531-appb-000050
  公式(13.3)。
可选地，作为一个实施例，在用于确定初始用户序列，处理器1302具体用于随机确定该L个小区的一个用户序列作为该初始用户序列。
可选地，作为另一个实施例，在用于确定初始用户序列，处理器1302具体用于根据使得初始用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到该初始用户序列，其中该第一用户组合为该每一次搜索中能够加入该初始用户序列并且和速率最大的用户组合， 该初始用户序列中的任一个UE只存在于该初始用户序列的一个用户组合中。
具体地，处理器1302还用于根据该大尺度衰落因子获取该多个用户组合各自的和速率以形成该多个用户组合的和速率集合，其中该和速率集合中的和速率与该L个小区的多个用户组合一一对应；在用于根据使得初始用户序列对应的系统和速率最大的原则对该L个小区的多个用户组合进行贪婪搜索，并将该贪婪搜索过程中的每一次搜索中的第一用户组合加入到该初始用户序列，处理器1302具体用于重复执行步骤13.3和步骤13.4预设的次数，其中该预设的次数不大于K：
13.3.将该第一用户组合加入到该初始用户序列中，其中该第一用户组合为当前该和速率集合中最大的和速率所对应的用户组合；
13.4将该和速率集合中的第二和速率删除或者置为0，其中该第二和速率对应的用户组合中包含该第一用户组合中的至少一个UE。
进一步地，在用于确定该初始用户序列，处理器1302还用于在该贪婪搜索完成之后，如果加入该初始用户序列的用户组合个数C小于K个，则从该L个小区中挑选K-C个用户组合加入到该初始用户序列中，以形成该初始用户序列。
具体地，该和速率集合用Rate表来表示，其中，该Rate表为L维数组，该Rate表的每一个维度分别对应于该L个小区中的一个小区，该Rate表中第一维度的下标分别对应于该L个小区中第一小区的UE，该第一维度对应于该第一小区。需要说明的是，该第一小区可以是L个小区中的任一个小区。
可选地，L个小区的多个用户组合为该L个小区的所有用户组合中的全部或部分用户组合。
协同设备1300还可执行图4的方法，并实现协同设备在图4、图5所示实施例中的功能，其具体实现可参考图4、图5所示的实施例，本发明实施例在此不再赘述。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本申请所提供的几个实施例中，应该理解到，所揭露的系统、装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本发明各个实施例所述方法的全部或部分步骤。而 前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应所述以权利要求的保护范围为准。

Claims (36)

  1. 一种多输入输出系统的导频调度方法，其特征在于，其中所述系统包括L个小区，所述L个小区的每一个小区最多存在K个用户设备UE，所述方法包括：
    确定初始用户序列，所述初始用户序列为所述L个小区的一种用户序列，每个用户序列包括所述L个小区的K个用户组合，每个用户组合最多包含L个UE，所述K个用户组合中的每个用户组合中的UE分别属于所述L个小区中不同的小区，所述L个小区中的每个小区中的UE分别属于所述K个用户组合中不同的用户组合；
    根据所述初始用户序列进行禁忌搜索以获得优化用户序列，其中，所述优化用户序列为所述L个小区的一种用户序列；
    根据所述优化用户序列对所述L个小区中的UE进行导频调度，其中，所述L个小区中属于所述优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
  2. 如权利要求1所述的方法，其特征在于，所述根据所述初始用户序列进行禁忌搜索以获得优化用户序列包括：
    根据所述初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，所述导频调度优化准则包括以下一种准则：
    使得所述L个小区的系统和速率最大的准则；或
    使得所述L个小区中UE的最小速率最大的准则；或
    使得所述L个小区的系统和速率最大的准则且满足所述L个小区的UE的服务质量QoS的速率需求的准则。
  3. 如权利要求2所述的方法，其特征在于，所述方法还包括：
    获取所述L个小区中每个UE到所述L小区的其它小区的基站的大尺度衰落因子，其中，所述L个小区中每个UE到所述L小区的其它小区的基站的大尺度衰落因子用于确定所述L个小区中的UE的速率，所述系统的和速率为所述L个小区中所有接入UE的速率之和。
  4. 如权利要求3所述的方法，其特征在于，当所述导频调度优化准则为使得所述L个小区的系统和速率最大的准则，所述根据所述初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列包括：
    将所述初始用户序列赋值给历史用户序列和当前用户序列；
    对步骤a、b循环执行N次后将所述历史用户序列作为所述优化用户序列，N为不大于L的正整数，其中
    a、将搜索禁忌表置空；
    b、对步骤b1、b2执行预定的次数，其中
    b1、将待交换小区在所述当前用户序列中任意X个用户组合的UE进行位置交换得到所述当前用户序列的多个邻域序列，并获得所述当前用户序列的多个邻域序列对应的系统和速率，并取出使得系统和速率最大的第一邻域序列，其中，所述待交换小区为循环执行N次的过程中的第l次循环执行选定的小区，所述N次的循环执行的每一次选定的小区均不相同，所述系统和速率由所述大尺度衰落因子确定，X为大于1且不大于K的正整数；
    b2、如果所述第一邻域序列满足所述禁忌搜索的特赦准则，则将所述第一邻域序列赋值给所述历史用户序列和当前用户序列，并将所述第一邻域序列加入所述搜索禁忌表中，或者如果所述第一邻域序列不满足禁忌搜索的特赦准则，则将所述当前用户序列的多种邻域序列中不在所述搜索禁忌表中且系统和速率最大的第二邻域序列赋值给当前用户序列，并将所述第二邻域序列加入所述搜索禁忌表中，其中，所述特赦准则为所述第一邻域序列的系统和速率大于所述历史用户序列的和速率，或者所述特赦准则为所述第一邻域序列的系统和速率大于或等于所述历史用户序列的和速率。
  5. 如权利要求4所述的方法，X的值为2。
  6. 如权利要求4或5所述的方法，所述预定的次数为K次。
  7. 如权利要求3至6任一项所述的方法，其特征在于，
    所述L个小区的用户序列对应的系统和速率由如下公式表示：
    Figure PCTCN2014090531-appb-100001
    其中，rate(Ωopt)表示所述L个小区的用户序列对应的系统和速率，Ωk表示所述L个小区的用 户序列中第k个用户组合，rate(Ωk)表示所述第k个用户组合的和速率，
    Figure PCTCN2014090531-appb-100002
    βjkl表示属于所述L个小区中的小区l且属于所述第k个用户组合的UE到小区j所属基站的大尺度衰落因子。
  8. 如权利要求1至7任一项所述的方法，其特征在于，所述确定初始用户序列包括：随机确定所述L个小区的一个用户序列作为所述初始用户序列。
  9. 如权利要求4至7任一项所述的方法，其特征在于，所述确定初始用户序列包括：
    根据使得所述初始用户序列对应的系统和速率最大的原则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的每一次搜索中的第一用户组合加入到所述初始用户序列，其中所述第一用户组合为所述每一次搜索中能够加入所述初始用户序列并且和速率最大的用户组合，所述初始用户序列中的任一个UE只存在于所述初始用户序列的一个用户组合中。
  10. 如权利要求9所述的方法，其特征在于，所述根据使得所述初始用户序列对应的系统和速率最大的原则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的每一次搜索得到的局部最优用户组合加入到所述初始用户序列包括：
    根据所述大尺度衰落因子获取所述多个用户组合各自的和速率以形成所述多个用户组合的和速率集合，其中所述和速率集合中的和速率与所述L个小区的多个用户组合一一对应；
    重复执行步骤c和步骤d预设的次数，其中所述预设的次数不大于K：
    c、将所述第一用户组合加入到所述初始用户序列中，其中所述第一用户组合为当前所述和速率集合中最大的和速率所对应的用户组合；
    d、将所述和速率集合中的第二和速率删除或者置为0，其中所述第二和速率对应的用户组合中包含所述第一用户组合中的至少一个UE。
  11. 如权利要求9或10所述的方法，其特征在于，所述确定初始用户序列还包括：
    在所述贪婪搜索完成之后，如果加入所述初始用户序列的用户组合个数C小于K个，则从所述L个小区中挑选K-C个用户组合加入到所述初始用户序列中，以形成所述初始用户序列。
  12. 如权利要求10所述的方法，其特征在于，所述和速率集合用Rate表来表示，其中，所述Rate表为L维数组，所述Rate表的每一个维度分别对应于所述L个小区中的一个小区，所述Rate表中第一维度的下标分别对应于所述L个小区中第一小区的UE，所述第一维度对应于所述第一小区。
  13. 一种多输入输出系统的导频调度方法，其特征在于，其中所述系统包括L个小区，所述L个小区的每一个小区最多存在K个用户设备UE，所述方法包括：
    获取所述L个小区的多个用户组合，其中所述多个用户组合中的任一个用户组合用于构成所述L个小区的用户序列，每个用户序列包括所述L个小区的K个用户组合，每个用户组合最多包含L个UE，所述K个用户组合中的每个用户组合中的UE分别属于所述L个小区中不同的小区，所述L个小区中的每个小区中的UE分别属于所述K个用户组合中不同的用户组合；
    对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的每一次搜索中的第一用户组合加入到优化用户序列，其中所述第一用户组合为所述每一次搜索中能够加入所述初始用户序列并且满足搜索条件的用户组合，所述优化用户序列中的任一个UE只存在于所述优化用户序列的一个用户组合中；
    根据所述优化用户序列对所述L个小区中的UE进行导频调度，其中，所述L个小区中属于所述优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
  14. 如权利要求13所述的方法，其特征在于，所述对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的每一次搜索中最符合搜索条件的用户组合加入到优化用户序列包括：
    按照导频调度优化准则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的第一用户组合加入到优化用户序列，所述搜索条件用于使所述优化用户序列符合所述导频调度优化准则，所述导频调度优化准则包括以下一种准则：
    使得所述L个小区的系统和速率最大的准则；或
    使得所述L个小区中UE的最小速率最大的准则；或
    使得所述L个小区的系统和速率最大的准则且满足所述L个小区的UE的服务质量QoS的速率需求的准则。
  15. 如权利要求14所述的方法，其特征在于，当所述导频调度优化准则为使得所述L个小区的系统和速率最大的准则，所述按照导频调度优化准则对所述L个小区的多个用户组合进行贪婪搜索，并将所述 贪婪搜索过程中的第一用户组合加入到优化用户序列包括：
    按照使得所述L个小区的系统和速率最大的准则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的第一用户组合加入到优化用户序列，其中所述第一用户组合为所述每一次搜索中能够加入所述优化用户序列并且和速率最大的用户组合。
  16. 如权利要求15所述的方法，其特征在于，
    所述按照使得所述L个小区的系统和速率最大的准则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的第一用户组合加入到优化用户序列包括：
    获取所述L个小区中每个UE到所述L小区的其它小区的基站的大尺度衰落因子，并根据所述大尺度衰落因子获取所述L个小区的多个用户组合的和速率以形成所述L个小区的多个用户组合的和速率集合，其中所述和速率集合中的和速率与所述L个小区的多个用户组合一一对应；
    重复执行步骤c和步骤d预设的次数，其中所述预设的次数不大于K：
    c、将所述第一用户组合加入到所述优化用户序列中，其中所述第一用户组合为当前所述和速率集合中最大的和速率所对应的用户组合；
    d、将所述和速率集合中的第二和速率删除或者置为0，其中所述第二和速率对应的用户组合中包含所述第一用户组合中的至少一个UE。
  17. 如权利要求15或16所述的方法，其特征在于，所述方法还包括：
    在贪婪搜索完成之后，如果加入所述优化用户序列的用户组合个数C小于K个，则从所述L个小区中挑选K-C个用户组合加入到所述优化用户序列中，以形成所述优化用户序列。
  18. 如权利要求16所述的方法，其特征在于，所述和速率集合用Rate表来表示，其中，所述Rate表为L维数组，所述Rate表的每一个维度分别对应于所述L个小区中的一个小区，所述Rate表中第一维度的下标分别对应于所述L个小区中第一小区的UE，所述第一维度对应于所述第一小区。
  19. 一种多输入输出系统的协同设备，其特征在于，其中所述系统包括L个小区，所述L个小区的每一个小区最多存在K个用户设备UE，所述协同设备包括：
    确定单元，用于确定初始用户序列，所述初始用户序列为所述L个小区的一种用户序列，每个用户序列包括所述L个小区的K个用户组合，每个用户组合最多包含L个UE，所述K个用户组合中的每个用户组合中的UE分别属于所述L个小区中不同的小区，所述L个小区中的每个小区中的UE分别属于所述K个用户组合中不同的用户组合；
    搜索单元，用于根据所述初始用户序列进行禁忌搜索以获得优化用户序列，其中，所述优化用户序列为所述L个小区的一种用户序列；
    调度单元，用于根据所述优化用户序列对所述L个小区中的UE进行导频调度，其中，所述L个小区中属于所述优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
  20. 如权利要求19所述的协同设备，其特征在于，所述确定单元具体用于根据所述初始用户序列，按照导频调度优化准则进行禁忌搜索以获得优化用户序列，所述导频调度优化准则包括以下一种准则：
    使得所述L个小区的系统和速率最大的准则；或
    使得所述L个小区中UE的最小速率最大的准则；或
    使得所述L个小区的系统和速率最大的准则且满足所述L个小区的UE的服务质量QoS的速率需求的准则。
  21. 如权利要求20所述的协同设备，其特征在于，所述协同设备还包括：
    获取单元，用于获取所述L个小区中每个UE到所述L小区的其它小区的基站的大尺度衰落因子，其中，所述L个小区中每个UE到所述L小区的其它小区的基站的大尺度衰落因子用于确定所述L个小区中的UE的速率，所述系统的和速率为所述L个小区中所有接入UE的速率之和。
  22. 如权利要求21所述的协同设备，其特征在于，当所述导频调度优化准则为使得所述L个小区的系统和速率最大的准则，所述搜索单元具体用于：
    将所述初始用户序列赋值给历史用户序列和当前用户序列；
    对步骤a、b循环执行N次后将所述历史用户序列作为所述优化用户序列，N为不大于L的正整数，其中
    a、将搜索禁忌表置空；
    b、对步骤b1、b2执行预定的次数，其中
    b1、将待交换小区在所述当前用户序列中任意X个用户组合的UE进行位置交换得到所述当前用户序列的多个邻域序列，并获得所述当前用户序列的多个邻域序列对应的系统和速率，并取出使得系统和速率最大的第一邻域序列，其中，所述待交换小区为循环执行N次的过程中的第l次循环执行选定的小区，所述N次的循环执行的每一次选定的小区均不相同，所述系统和速率由所述大尺度衰落因子确定，X为大于1且不大于K的正整数；
    b2、如果所述第一邻域序列满足所述禁忌搜索的特赦准则，则将所述第一邻域序列赋值给所述历史用户序列和当前用户序列，并将所述第一邻域序列加入所述搜索禁忌表中，或者如果所述第一邻域序列不满足禁忌搜索的特赦准则，则将所述当前用户序列的多种邻域序列中不在所述搜索禁忌表中且系统和速率最大的第二邻域序列赋值给当前用户序列，并将所述第二邻域序列加入所述搜索禁忌表中，其中，所述特赦准则为所述第一邻域序列的系统和速率大于所述历史用户序列的和速率，或者所述特赦准则为所述第一邻域序列的系统和速率大于或等于所述历史用户序列的和速率。
  23. 如权利要求22所述的协同设备，X的值为2。
  24. 如权利要求22或23所述的协同设备，所述预定的次数为K次。
  25. 如权利要求21至24任一项所述的协同设备，其特征在于，
    所述L个小区的用户序列对应的系统和速率由如下公式表示：
    Figure PCTCN2014090531-appb-100003
    其中，rate(Ωopt)表示所述L个小区的用户序列对应的系统和速率，Ωk表示所述L个小区的用户序列中第k个用户组合，rate(Ωk)表示所述第k个用户组合的和速率，
    Figure PCTCN2014090531-appb-100004
    βjkl表示属于所述L个小区中的小区l且属于所述第k个用户组合的UE到小区j所属基站的大尺度衰落因子。
  26. 如权利要求19至25任一项所述的协同设备，其特征在于，所述确定单元具体用于随机确定所述L个小区的一个用户序列作为所述初始用户序列。
  27. 如权利要求22至25任一项所述的协同设备，其特征在于，
    所述搜索单元还用于根据使得贪婪搜索用户序列对应的系统和速率最大的原则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的每一次搜索中的第一用户组合加入到所述贪婪搜索用户序列，其中所述第一用户组合为所述每一次搜索中能够加入所述贪婪搜索用户序列并且和速率最大的用户组合，所述贪婪搜索用户序列中的任一个UE只存在于所述贪婪搜索用户序列的一个用户组合中；
    所述确定单元具体用于确定所述贪婪搜索用户序列为所述初始用户序列。
  28. 如权利要求27所述的协同设备，其特征在于，
    所述获取单元还用于根据所述大尺度衰落因子获取所述多个用户组合各自的和速率以形成所述多个用户组合的和速率集合，其中所述和速率集合中的和速率与所述L个小区的多个用户组合一一对应；
    在用于根据使得贪婪搜索用户序列对应的系统和速率最大的原则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的每一次搜索中的第一用户组合加入到贪婪搜索用户序列，所述搜索单元具体用于重复执行步骤c和步骤d预设的次数，其中所述预设的次数不大于K：
    c、将所述第一用户组合加入到所述贪婪搜索用户序列中，其中所述第一用户组合为当前所述和速率集合中最大的和速率所对应的用户组合；
    d、将所述和速率集合中的第二和速率删除或者置为0，其中所述第二和速率对应的用户组合中包含所述第一用户组合中的至少一个UE。
  29. 如权利要求27或28所述的协同设备，其特征在于，所述确定单元还用于在所述贪婪搜索完成之后，如果加入所述初始用户序列的用户组合个数C小于K个，则从所述L个小区中挑选K-C个用户组合加入到所述初始用户序列中，以形成所述初始用户序列。
  30. 如权利要求28所述的协同设备，其特征在于，所述和速率集合用Rate表来表示，其中，所述Rate表包含L个维度，所述Rate表的每一个维度分别对应于所述L个小区中的一个小区，所述Rate表的每一个维度的下标分别对应于所述L个小区中的UE。
  31. 一种多输入输出系统的协同设备，其特征在于，其中所述系统包括L个小区，所述L个小区的每一个小区最多存在K个用户设备UE，所述协同设备包括：
    获取单元，用于获取所述L个小区的多个用户组合，其中所述多个用户组合中的任一个用户组合用于构成所述L个小区的用户序列，每个用户序列包括所述L个小区的K个用户组合，每个用户组合最多包含L个UE，所述K个用户组合中的每个用户组合中的UE分别属于所述L个小区中不同的小区，所述L个小区中的每个小区中的UE分别属于所述K个用户组合中不同的用户组合；
    搜索单元，用于对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的每一次搜索中的第一用户组合加入到优化用户序列，其中所述第一用户组合为所述每一次搜索中能够加入所述初始用户序列并且满足搜索条件的用户组合，所述优化用户序列中的任一个UE只存在于所述优化用户序列的一个用户组合中；
    调度单元，用于根据所述优化用户序列对所述L个小区中的UE进行导频调度，其中，所述L个小区中属于所述优化用户序列的同一个用户组合的UE共享相同的一段导频序列。
  32. 如权利要求31所述的协同设备，其特征在于，所述搜索单元具体用于按照导频调度优化准则对所述L个小区的多个用户组合进行贪婪搜索以得到优化用户序列，所述导频调度优化准则包括以下一种准则：
    使得所述L个小区的系统和速率最大的准则；或
    使得所述L个小区中UE的最小速率最大的准则；或
    使得所述L个小区的系统和速率最大的准则且满足所述L个小区的UE的服务质量QoS的速率需求的准则。
  33. 如权利要求32所述的协同设备，其特征在于，当所述导频调度优化准则为使得所述L个小区的系统和速率最大的准则，在用于按照导频调度优化准则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的第一用户组合加入到优化用户序列，所述搜索单元具体用于按照使得所述L个小区的系统和速率最大的准则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的第一用户组合加入到优化用户序列，其中所述第一用户组合为所述每一次搜索中能够加入所述优化用户序列并且和速率最大的用户组合。
  34. 如权利要求33所述的协同设备，其特征在于，
    所述获取单元还用于获取所述L个小区中每个UE到所述L小区的其它小区的基站的大尺度衰落因子，并根据所述大尺度衰落因子获取所述L个小区的多个用户组合的和速率以形成所述L个小区的多个用户组合的和速率集合，其中所述和速率集合中的和速率与所述L个小区的多个用户组合一一对应；
    在用于按照使得所述L个小区的系统和速率最大的准则对所述L个小区的多个用户组合进行贪婪搜索，并将所述贪婪搜索过程中的第一用户组合加入到优化用户序列，所述搜索单元具体用于
    重复执行步骤c和步骤d预设的次数，其中所述预设的次数不大于K：
    c、将所述第一用户组合加入到所述优化用户序列中，其中所述第一用户组合为当前所述和速率集合中最大的和速率所对应的用户组合；
    d、将所述和速率集合中的第二和速率删除或者置为0，其中所述第二和速率对应的用户组合中包含所述第一用户组合中的至少一个UE。
  35. 如权利要求33或34所述的协同设备，其特征在于，所述搜索单元还用于在贪婪搜索完成之后，如果加入所述优化用户序列的用户组合个数C小于K个，则从所述L个小区中挑选K-C个用户组合加入到所述优化用户序列中，以形成所述优化用户序列。
  36. 如权利要求34所述的协同设备，其特征在于，所述和速率集合用Rate表来表示，其中，所述Rate表为L维数组，所述Rate表的每一个维度分别对应于所述L个小区中的一个小区，所述Rate表中第一维度的下标分别对应于所述L个小区中第一小区的UE，所述第一维度对应于所述第一小区。
PCT/CN2014/090531 2013-11-07 2014-11-07 多输入输出系统的导频调度方法及协同设备 WO2015067200A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310554785.2 2013-11-07
CN201310554785.2A CN104640222B (zh) 2013-11-07 2013-11-07 多输入输出系统的导频调度方法及协同设备

Publications (1)

Publication Number Publication Date
WO2015067200A1 true WO2015067200A1 (zh) 2015-05-14

Family

ID=53040918

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/090531 WO2015067200A1 (zh) 2013-11-07 2014-11-07 多输入输出系统的导频调度方法及协同设备

Country Status (2)

Country Link
CN (1) CN104640222B (zh)
WO (1) WO2015067200A1 (zh)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107735956B (zh) * 2015-06-26 2020-10-23 瑞典爱立信有限公司 提高无线通信系统容量的方法、相关设备和存储介质
CN106059728B (zh) * 2016-05-05 2019-03-01 西安交通大学 一种大规模mimo系统中的基于相移的导频设计方法
CN106027214B (zh) * 2016-05-16 2019-01-04 东南大学 一种多小区大规模mimo系统导频分配方法
CN107548094B (zh) * 2016-06-23 2020-08-25 华为技术有限公司 传输用户序列的方法、网络设备和终端设备
CN107888332B (zh) * 2016-11-28 2019-03-01 电信科学技术研究院有限公司 一种用户检测方法和装置
CN106712817B (zh) * 2016-12-26 2020-01-14 山东大学 一种低复杂度的基于用户交换的导频分配方法
CN108063656B (zh) * 2017-11-28 2020-11-03 中通服咨询设计研究院有限公司 一种适用于大规模mimo蜂窝网络的新型导频分配方法
CN108667495B (zh) * 2018-04-25 2021-06-11 东南大学 分布式mimo系统中基于互污染指标的导频调度方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101938296A (zh) * 2009-06-29 2011-01-05 华为技术有限公司 一种导频序列生成方法、用户设备、基站
CN102804887A (zh) * 2009-06-12 2012-11-28 华为技术有限公司 无线接入系统中上行链路小区间干扰协调的方法
CN103298124A (zh) * 2013-06-14 2013-09-11 东南大学 基于空间正交性的大规模mimo系统导频分配方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101882952B (zh) * 2010-06-24 2012-12-19 东南大学 统计特征模式的空分多址传输方法
US9893773B2 (en) * 2011-09-21 2018-02-13 Provenance Asset Group Llc System and method of wireless communication using large-scale antenna networks

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102804887A (zh) * 2009-06-12 2012-11-28 华为技术有限公司 无线接入系统中上行链路小区间干扰协调的方法
CN101938296A (zh) * 2009-06-29 2011-01-05 华为技术有限公司 一种导频序列生成方法、用户设备、基站
CN103298124A (zh) * 2013-06-14 2013-09-11 东南大学 基于空间正交性的大规模mimo系统导频分配方法

Also Published As

Publication number Publication date
CN104640222B (zh) 2018-06-05
CN104640222A (zh) 2015-05-20

Similar Documents

Publication Publication Date Title
WO2015067200A1 (zh) 多输入输出系统的导频调度方法及协同设备
Xu et al. Optimal power allocation scheme for non-orthogonal multiple access with $\alpha $-fairness
CN114390580A (zh) 波束上报方法、波束信息确定方法及相关设备
CN111698789B (zh) 通信系统中的调度方法、装置及存储介质
CN110365380B (zh) 数据传输的方法、通信装置及系统
CN102457951B (zh) 一种多小区协作通信中链路联合波束成形的方法和基站
CN111510946B (zh) 测量上报的方法与装置
CN105577316B (zh) 预编码的方法和基站
CN104796900A (zh) 基于拍卖理论的蜂窝网络中d2d通信资源分配方法
Perlaza et al. On the base station selection and base station sharing in self-configuring networks
CN105451268A (zh) 一种高能效的异构小区接入方法
WO2017114513A1 (zh) 一种csi反馈方法及装置
WO2022252963A1 (zh) 信道状态信息的测量方法和装置
EP4287740A1 (en) User equipment, electronic device, wireless communication method and storage medium
Kamel et al. Average downlink rate in Ultra-Dense Networks
CN105960005B (zh) 超密集网络中保障用户公平性的功率控制方法
Lin et al. Fast spectrum reuse and power control for device-to-device communication
CN102130861B (zh) 一种终端间信道相关性的确定方法及基站
CN113796127B (zh) 多频率通信网络中的小区选择
CN115913486A (zh) 信息上报方法、装置、终端及可读存储介质
JP6619295B2 (ja) スケジューリング装置
Li et al. Time allocation and optimization in time-reversal wireless powered communication networks
Zhong et al. A game theory scheme combined with the interference region in cellular D2D heterogeneous networks
Liang et al. A practical dynamic clustering scheme using spectral clustering in ultra dense network
Zhang et al. Artificial fish swarm algorithm based pilot allocation in massive MIMO system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14859571

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14859571

Country of ref document: EP

Kind code of ref document: A1