Abstract: The object of the present invention is to provide a technology of deciding CoMP point dynamically by considering the spectrum efficiency after channel-dependent scheduling with different possible combinations of CoMP points. According to the present invention, A wireless communication system, characterized in deciding number of CoMP (coordinated multipoint transmission and reception) point to a user equipment by using information related to scheduling.
[DESCRIPTION]
[Title of Invention]
COMMUUNICATION SYSTEM, COMMUNICATION METHOD, BASE STATION, USER EQUIPMENT,
AND THEREIN PROGRAM
[Technical Field]
[0001]
The present disclosure relates generally to communications and, more specifically, to a technology for CoMP (coordinated multipoint transmission and reception) point decision for coordinated multipoint transmission used in a cellular system.
[Background Art]
[0002]
To meet demand for high speed wireless access beyond 3G cellular system, E-UTRA (Evolved UMTS Terrestrial Radio Access) has been standardized as LTE (Long Term Evolution) radio access in 3GPP (3rd Generation Partnership Project). Since March 2008, 3GPP has started standardization of IMT-Advanced as LTE-Advanced. The target peak rate of LTE-Advanced is 1Gbps and 500 Mbps for downlink and uplink, respectively [1]. The higher spectrum efficiency compared to LTE is also required. At the same time, LTE-Advanced should consider backward compatibility with rel.8 LTE as it is developed as the enhancement of LTE.
[0003]
Downlink CoMP is considered as a key technique to improve the cell-edge throughput and/or to
increase system throughput for 3GPP LTE-Advanced [2].
[0004]
100 and 200 in Figure 1 show the examples of downlink (DL) transmission without CoMP (w/o CoMP) and with CoMP (w/ CoMP), respectively. In Fig. 1, only point 1(101) is evolved in transmission for the UE (user equipment) 104. The UE 104 receives the signal from point 1 but interferences from point 2(102) and point 3(103). However, in 200, three points, i.e., point 1(201), point 2(202), and point 3(203) are involved in coordinated transmission. We use "CoMP point" to indicate the point involved in coordinated transmission, which can be eNode-B (eNB), cell, or other type of node, such as remote antennas connected by optical fiber [3]. A "CoMP UE" is the UE who has more than one CoMP points.
[0005]
Joint transmission and coordinated scheduling/coordinated beamforming for downlink CoMP have been agreed as candidate schemes for LTE-Advanced [2]. In both cases, multiple CoMP points allocate resource blocks (RBs) with same spectrum positions. In case of joint transmission, multiple CoMP points transmit data simultaneously to CoMP UE 204. As shown in the example of 200, the UE 204 is a CoMP UE. The interference from point 2(202) and point 3(203) are transformed into signals to improve the received signal for the CoMP UE 204. The CoMP gain in terms of the signal-to-interference plus noise ratio (SINR) comes from the diversity gain of all CoMP points, i.e., point 1(201), point 2(202) and point 3(203). However, in case of coordinated scheduling, only CoMP point 1(201) is transmitting data to CoMP UE 204 and the interferences from CoMP point 2(202) and 3(203) are cancelled. The CoMP gain in terms of SINR comes from the interference avoidance of CoMP point 2(202) and 3(203).
[0006]
How to select CoMP points has great impact on CoMP gain in terms of cell-edge user throughput and average sector throughput. The CoMP gain may be limited when CoMP points are not properly selected to include the potential strong received signal for joint processing or to mitigate the strong interference for coordinated scheduling.
[0007]
Conventional method to select CoMP point is to use fixed CoMP points to calculate and rank the PF metric (proportional fairness metric) for channel-dependent scheduling. For example, Figure 2
illustrates the flowchart of conventional method carried out in the scheduler, which is described as follows:
[0008]
In Step 1, the scheduler receives the UE feedback. The UE feedback includes the UE measured
long-term averaged channel information, such as reference signal received power (RSRP) used in [3] or geometry in [4]. RSRP, is determined for a considered cell as the linear average over the power contributions (in [W]) of the resource elements that carry cell-specific reference signals within the considered measurement frequency bandwidth. The geometry is the average signal-to-interference plus noise ratio (SINR) in the presence of shadowing and path loss.
[0009]
Besides above information, the UE feedback also includes the channel quality indicator (CQI), which is used to indicate the variation of channel state over bandwidth the measured channel state on each RB over RS bandwidth. For example, in LTE, the feedback period for RSRP can be 120~200ms, etc.
[5]; while, the feedback period for CQI can be 2ms, 5ms, 10ms, etc. [6].
[0010]
In Step 2, the cell (or other types of points) ranking is carried out for each UE. For each UE, the cells (points) are ranked as follow.
Where X„ s is the uth UE feedback RSHP or geometry of the S th cell in dB, obtained from Step 1.
[0011]
In Step 3, the candidate CoMP point is selected for each UE and the UE who has more than one
candidate CoMP points is regarded as a candidate CoMP UE. For the uth UE, the point with highest is selected as the serving cell, which is always a CoMP point Besides the serving cell, the cell with its Xm-XMfi-Threshold is also selected as a candidate CoMP point, where Threshold is pre-defined parameter in dB. For example, with low threshold, only the UEs close to the cell edge may be set as a candidate CoMP UE, who is able to make use of CoMP to improve its throughput
[0012]
The Step 4 includes step 306, 307 and 308. In step 306, for each candidate CoMP UE, N points
among candidate CoMP points are fixed as CoMP points. Accordingly, a candidate CoMP UE is
decided as a CoMP UE when N>1 CoMP points are selected; otherwise, when N=1, a candidate
CoMP UE reduces to a non-CoMP UE.
[0013]
After deciding CoMP points, in step 307, the scheduling metric is calculated and ranked for the
preparation of channel-dependent scheduling. For a CoMP UE, the CQI improved by CoMP with fixed N CoMP points is used for metric calculation; while, for the other UE, the CQI without CoMP (N=1) is used.
[0014]
In step 308, the RBs are allocated for all UEs. The resource allocation is carried out based on the
metric ranking. When the UE has N>1 CoMP points, the channel-dependent scheduling allocates the resource blocks with same spectrum positions at each CoMP point
[0015]
In the final Step 5, the scheduler informs the scheduling result to the UEs, e.g., the positions and number of allocated RBs.
[Citation list]
[Non Patent Citation]
[0016]
CPTL 1] 3GPP, TR36.913 (V8.0.1), "Requirements for Further Advancements for
E-UTRA (LTE-Advanced)," June 2008 CPTL 2] 3GPP, TR36.814 (Vl.0.0), "Further Advancements for E-UTRA Physical Layer Aspects" [PTL 3] Rl-091484, NTT DOCOMO, 3GPP TSG-RAN WG1 #56bis, March 23, 27th 2009 [PTL 4] Rl-084173, Samsung, 3GPP TSG-RAN WG1 #55, November 10-14, 2008
[PTL 53 3GPP,. TS36.331 (V8.7.0) [PTL 6] 3GPP, TS36.213 CV8.4.0)
[Disclosure of Invention]
[Technical Problem]
[0017]
Though the number of CoMP point(s) is dynamic depending on neighboring circumstances of the
base station, conventional method sets fixed the number of CoMP point(s) for scheduling metric
ranking and resource allocation. When the UE has N>1 CoMP points, the channel-dependent
scheduling uses the improved CQI by using CoMP and allocates the RBs with the same spectrum
position at each CoMP point By using common RBs at N CoMP points, the achievable rate of CoMP UE is increased due to the improved CQI. However, when the instantaneous achievable rate with CoMP is not large enough, that is, smaller than N times of that without CoMP, the spectrum efficiency (rate-to-spectrum ratio) is decreased. Therefore, the problem of conventional method is that the fixed N CoMP point(s) doesn't consider whether the spectrum efficiency after channel-dependent scheduling is increased or not.
[0018]
We intend to dynamically decide CoMP point and efficiently allocate resource blocks by considering the spectrum efficiency after channel-dependent scheduling with different possible combinations of CoMP points.
[Solution of Reference]
[0019]
The 1st invention for solving the above-mentioned problems, which is a wireless communication
system, characterized in deciding number of CoMP (coordinated multipoint transmission and
reception) points to a user equipment by using information related to scheduling.
[0020]
The 2nd invention for solving the above-mentioned problems, which is a base station, characterized in deciding number of CoMP (coordinated multipoint transmission and reception) points to a user equipment by using information related to scheduling.
[0021]
The 3rd invention for solving the above-mentioned problems, which is a user equipment,
characterized in communicating to use decided number of CoMP (coordinated multipoint
transmission and reception) points based on using information related to scheduling by a base
station.
[0022]
The 4th invention for solving the above-mentioned problems, which is a coordinator, characterized in deciding number of CoMP (coordinated multipoint transmission and reception) points to a user equipment by using information related to scheduling.
[0023]
The 5th invention for solving the above-mentioned problems, which is a wireless communication method, characterized in deciding number of CoMP (coordinated multipoint transmission and reception) points to a user equipment by using information related to scheduling.
[0024]
The 6th invention for solving the above-mentioned problems, which is a recording medium stored a communication program, said program causing a computer to execute the processes of: deciding
number of CoMP (coordinated multipoint transmission and reception) points to a user equipment by using information related to scheduling.
[0025]
The 7th invention for solving the above-mentioned problems, which is a recording medium stored a communication program, said program causing a computer to execute the processes of:
communicating to use decided number of CoMP (coordinated muttipoint transmission and reception) points based on using information related to scheduling by a base station.
[Advantageous EFFECT]
[0026]
Dynamic CoMP point decision and efficient resource allocation assume part of or all possible
combinations of CoMP points to calculate and rank scheduling metric. The spectrum efficiency is
improved by comparing the channel-dependent scheduling result using different calculated
scheduling metrics. Comparing the proposed approach with conventional one, larger CoMP gain in terms of cell-edge user throughput and average sector throughput can be achieved.
[Brief Description of Drawings]
[0027]
For a more complete understanding of the present disclosure and its advantages, reference is now
made to the following description taken in conjunction with the accompanying drawings, in which like reference numerals represent like parts:
[Fig. 1]
Fig.1 shows an example of DL transmission without CoMP and with CoMP.
Fig. 2]
Fig.2 shows a flowchart of conventional method.
[Fig. 3]
Fig.3 shows one example of a block diagram of the present invention.
[Fig. 4]
Fig.4 shows a General flowchart of proposed method.
[Fig. 5]
Fig. 5 shows a flowchart of Example 1.
[Fig. 6]
Fig.6 shows a flowchart of Example 2.
[Fig. 7]
Fig.7 shows a flowchart of Example 3.
[Fig. 8]
Fig.8 shows a flowchart of Example 4.
[Fig. 9]
Fig.9 shows a flowchart of Example 5.
[Fig. 10]
Fig. 10 shows a flowchart of Example 6.
[Fig. 11]
Fig. 11 shows a flowchart of Example 7.
[Fig. 12]
Fig. 12 shows a flowchart of Example 8.
[Fig. 13]
Fig. 13 shows a flowchart of Example 9.
[Reference Signs List]
[0028]
310 UE (user equipment)
311 reference signal transmitter
312 data transmitter
313 controller
314 antenna
315 receiver
320 eNodeB
321 reference signal receiver
322 data receiver,
323 controller
324 scheduler
325 transmitter
326 antenna
[Best Mode For Carrying Out the Invention]
[0029]
Figure 4 through 13, discussed below, and the embodiment used to describe the principles of the
present invention are by way of illustration only and should not be construed in any way to limit the scope of the disclosure. Those skilled in the art will understand that the principles of the present disclosure may be implemented in any suitably arranged wireless network.
[0030]
At first, a configuration of this embodiment will be explained. Fig. 3 is a schematic view of a wireless communication system 300 in the present invention. The wireless communication system 300 includes a plural of UE (User Equipment) 310 and a plural of eNodeB 320.
[0031]
The UE 310 includes a UE feedback transmitter 311, a data transmitter 312, a controller 313, an
antenna 314, and a receiver 315.
[0032]
The UE feedback transmitter 311 generates a UE feedback, and transmits it to the each eNodeB 320 as serving cell via the antenna 314. The UE feedback includes the measurement long-term averaged channel information, such as RSRP for each of all eNodeB.
[0033]
The data transmitter 312 generates an actual data, and transmits it to the eNodeB 320 via the
antenna 314.
[0034]
When receives pilot signal from eNodeB, the controller 313 measures RSRP (reference signal
received power), geometry, or CQI (channel quality index) as UE feedback of the eNodeB.
[0035]
The receiver 315 receives various kinds of the data or control information being transmitted from the eNodeB 320.
[0036]
The eNodeB 320 is a point, for example, a base station. The eNodeB 320 includes a feedback receiver 321, a data receiver 322, a controller 323, a scheduler 324, a transmitter 325 and an antenna 326.
[0037]
The feedback receiver 321 receives the feedback information generated by the UE. The UE feedback may be exchanged between eNodeBs for coordinated transmission or reception. The data receiver 322 receives an actual data generated by the UE. The controller 323 carries out several kind of controlling for each function in eNodeB, for example, controlling of transmit power etc.. The scheduler 324 carries out resource allocation, and generates scheduling information based on resource allocation. Furthermore, carries out coordination of transmission or reception for CoMP UE, its details are as follows. The transmitter 325 transmits the scheduling information via the antenna 326.
[0038]
Next, a proposed method of this invention will be explained.
[0039]
Firstly, a general flowchart of the method is illustrated in Figure 4, which is described as follows.
[0040]
In Step 401, a scheduler 324 receives UE feedback information about all the points (e.g., cells) close to the UE.
[0041]
In Step 402, the cell (or other types of points) ranking is carried out for each UE. For each UE, the cells are ranked as follows.
Where X , is the uth UE feedback information of the S th cell in dB(u and S are natural 19,5
number). When is the UE feedback RSRP or geometry, the UE feedback RSRP or geometry is obtained from Step 401.
When is the CQI information of the allocated RBs, the positions of allocated RBs is obtained from Step 404 (explained later).
[0042]
In Step 403, the candidate CoMP point is selected for each UE and the UE who has more than one candidate CoMP points is regarded as a candidate CoMP UE. For the uth UE, the point with highest is selected as the serving cell, which is always a CoMP point. The cell with its is also selected as the candidate CoMP point, where Threshold is pre-defined parameter in dB.
[0043]
In Step 404, the CoMP points of candidate CoMP UEs are not fixed for metric ranking. Firstly, the number of CoMP points is initialized for the scheduling metric calculation and ranking. Next, for a CoMP candidate UE, the CoMP point decision and resource allocation are jointly carried out.
Accordingly, the decided CoMP points and allocation results are used to update metric ranking for the following resource allocation. The step 406 includes the initialization of the CoMP point(s) and the metric ranking. The scheduling metric is ranked by using initialized N CoMP points at the beginning and then update by using the decided CoMP points and scheduling results from the output of step 407. In the step 407, the CoMP point decision as well as resource allocation is carried out for each UE. The result of step 407 will be used to update metric ranking in step 406.
[0044]
In Step 405, the scheduler informs the scheduling result to the UEs, e.g., the positions and number of allocated RBs.
[0045]
Next, the following discussion gives Example 1 ~9 to illustrate proposed method. The flowcharts of 9 examples are based on general flowchart in Fig. 4 but have different Step 404. The flowchart of Step 406 and Step 407 in the Step 404 will be explained for Example 1-9 in Fig. 5-13, respectively.
Concerning decision number of CoMP point "N0", example 1-9 carry out to use information related to scheduling. The information related to scheduling is information that is used by scheduling, for example PF metric, CQI or Qos etc.. To be noted is that we calculate scheduling metric according to PF metric as an example. The scheduling metric can be based on maximum CQI criterion, quality of service (QoS) etc.. It is preferable for information related to scheduling calculated in the case of scheduling in the eNodeB. In addition, concerning allocation, example 1-9 carry out to consider the information related to scheduling in assuming par or all CoMP point.
[0046]
In the following Example 1 -3, the Step 404 in Fig. 4 is carried out to allocate resource blocks at first and then decide CoMP points. The best combination of CoMP points is selected to achieve the highest normalized sum rate of allocated RBs. Since different initialization is assumed in step 406, the corresponding process in step 407 is also different for each Example.
[0047]
In Example 4-6, the Step 404 in Fig. 4 is carried out to decide CoMP points first and then allocate resource blocks. The best combination of Compo points is selected to achieve the highest spectrum efficiency normalized rate based on the RB with highest metric belong to the candidate CoMP UE.
After that, the resource is allocated to this UE by using decided number of CoMP points.
[0048]
In Example 7-9, the Step 404 in Fig. 4 is carried out to recursively decide CoMP points and allocate resource blocks. The scheduler 324 firstly allocates a RB with highest metric and temporarily decides the number of CoMP points N based on this allocated RB. Then, the scheduler 324 tries to allocate one more RB and adjust N0 to achieve highest normalized sum rate of allocated RBs. The RBs is allocated one-by-one for a candidate CoMP UE by using the lasted updated N0. This iteration process stops until all the RBs with highest metric belong to .this UE is allocated and the number of CoMP points is also decided.
[0049]
Additionally, while a configuration in which the scheduler ranked the cells UE by UE was exemplified in the above-mentioned explanation, a configuration in which the cells are ranked in the UE side and transmitted to the eNode B may be employed.
[0050]
The example 1 is explained below. The scheduler 324 in the example 1 will be explained.
[0051]
The scheduler 324 calculates the metric of each resource block (RB) of all UEs. For a non-CoMP UE, e.g., a UE close to cell center, only single point transmission is used, i.e., N=1. For a candidate CoMP UE, e.g., a UE close to cell edge, the metrics are calculated by using the CQI assuming all combinations of the CoMP points. Here, the combination of the CoMP points means to select different number of CoMP points. The number of the CoMP points, N, is a numerical value ranging from "1 to Nmax", where the maximum value Nmax of the CoMP point may be pre-decided.
[0052]
The scheduler 324 ranks the calculated metrics RB by RB. Among the calculated metrics, the
scheduler 324 finds the highest metric and regards its owner UE as the watched UE, also called the highest UE. In addition, the scheduler 324 allocates the RBs to the watched UE when its metric is highest.
[0053]
When the watched UE is a candidate CoMP UE, the scheduler 324 firstly divides the allocated RBs of this watched UE into different groups, each group has the metric assuming same number of CoMP points. Next, the scheduler 324 calculates the normalized sum rate of the allocated RBs belong to each group. The normalized sum rate of a group assuming N CoMP points, T(N), is defined as the sum of estimated data rates normalized by f(N) when the RB(s) are simultaneously allocated on N CoMP points. Here, f(N) is a function of N, e.g., f(N)=max(1,a N) and 0
The example 2 will be explained. The explanation of a configuration similar to that of the
above-mentioned example is omitted.
[0072]
The scheduler 324 calculates the metric of each RB by assuming that the number of CoMP points is simply initialized as single point, i.e., N=1 for candidate CoMP UEs as well as for other non-CoMP UEs.
[0073]
The scheduler 324 calculates the metrics of all UEs on each RB and ranks the calculated metrics RB by RB.
[0074]
Among the calculated metrics, the scheduler 324 finds the highest metric and regards its owner UE as the watched UEV also called the highest. U.E. According to, the metric ranking on each unallocated. RBs, the scheduler can allocate the RB to the highest UE when the top metric is belong to this UE.
[0075]
When the watched UE is a candidate CoMP UE, the scheduler 324 calculates T(N=1) of highest RB based on the CQI assuming N=1 CoMP point. Further, the scheduler 324 calculates T(N+1) by using combined CQI when N+1 CoMP points are used for synchronous transmission on the allocated RBs of the watched UE.
[0076]
The scheduler 324 compares the calculated T(N=1) and T(N+1). If T(N+1)>T(N), the scheduler
increases N as N=N+1. This iteration process will stop only if the maximum T(N0) is achieved.
Therefore, the number of the CoMP points is decided as N=N0. Then, CoMP points are decided too, since above process have already assumed combination of CoMP point.
[0077]
After CoMP point decision, the scheduler 324 updates the metric ranking when N0 is unequal to the initialized N (N=1 in this example). Firstly, the metrics of the watched UE with initialized N is deleted and then replaced by the metric as N=N0. The metric ranking is updated by using the new metric with N=N0 of the watched UE.
[0078]
The scheduler 324 performs the above-mentioned process until all the RBs are allocated or all UEs are allocated, where the candidate CoMP UEs have also decided their CoMP point(s).
[0079]
The above-mentioned process is carried out cell-by-cell. After all cells have finished the above
mentioned process, the scheduler 324 coordinates the resource allocation for the CoMP UEs. When a CoMP UE is allocated different RB(s) from its CoMP points, the scheduler 324 makes coordination for this uncommonly allocated RB. One coordination method is to release the uncommonly allocated RB to another UE, who can be the UE with lower metric than the allocated UE in the metric ranking list of this RB. When a CoMP UE is not allocated the RB of CoMP points as a result of coordination, the scheduler will allocate the CoMP UE in the next scheduling time.
[0080]
An operation of this example is described below. For Example 2, the detailed flowchart of step 406 and step 407 in the Step 404 is shown in Fig. 6 and described as follows.
[0081]
In step 406 of Fig. 6, there are two steps, step 601 and 602. In step 601, the number of CoMP points is initialized as N=1 and the cell index starts from 0. Next, in step 602, the metric calculation and ranking use CQI assuming initialized N=1 CoMP point for CoMP UE candidates as well as for other non-CoMP UE to carry out channel-dependent scheduling. Here, the PF metric is explained in Math.
1.
[0082]
In step 407 of Fig. 6, the step 603 firstly allocates each unallocated RB to the UE with highest metric.
Among the allocated RBs, the scheduler 324 finds the RB with highest metric and regards its owner UE-as the watched UE,also called the, highest UE..Then,Jhe step .604 checks whether the highfst UE (the UE with highest PF metric) is a candidate CoMP UE or not. If yes, go to step 605; otherwise, go to step 609. The step 605 checks whether this candidate CoMP UE has already decided N or not.
[0083]
If yes, go to step 609; otherwise, go to step 606. In step 606, the achievable T(N) of this candidate
CoMP UE is calculated based on CQI of the allocated RB(s). T(N) and T(N+1) are calculated by
assuming N and N+1 CoMP points, respectively. Next, the step 607 compares T(N) and T(N+1) and checks whether NT(N) or not.
[0084]
If yes, go to step 608, where N is increased as N=N+1; otherwise, go to step 609. Accordingly, the number of CoMP points is decided as N=N0 to achieve highest T(N0). Then, CoMP points are decided too. After the CoMP point decision, the step 609 checks whether all RBs or UEs are allocated.
[0085]
If yes, go to step 610; otherwise, go to step 602 in step 406 to update the metric ranking. The step
610 checks whether the current cell is the last cell. If yes, go to step 611; otherwise, go to step 612 to move to the next UE and then go to step 602 in step 606 to update the metric ranking. In step 611, the coordination after resource allocation for each cell is carried out. For each CoMP UE, if the allocated RB is not commonly allocated at each CoMP point, the scheduler will replace the owner of this RB by the next UE with lower metric. The output of step 611 is the information about allocated RBs for each UE.
[0086]
Making a configuration like this example enables to improve both average sector throughput and
cell-edge user throughput.
[0087]
The scheduler 324 in the example 3 will be explained. The explanation of a configuration similar to that of the above-mentioned example is omitted.
[0088]
The scheduler 324 calculates the metric of each RB of the UE by assuming that the number of CoMP points is simply initialized as single point, i.e., N=Nmax (Nmax: maximum number of CoMP points) for candidate CoMP UEs but N=1 for other non-CoMP UE.
[0089]
The scheduler 324 calculates the metrics of all UEs on each RB and ranks the calculated metrics RB by RB. Among the allocated RBs, the scheduler 324 finds the RB with highest metric and regards its owner UE as the watched UE, also called the highest UE. According to the metric ranking on each unallocated RBs, the scheduler can allocate the RB to the highest UE when the top metric is belong to this UE.
[0090]
Over the allocated RBs, the scheduler 324 calculates T(N=Nmax) of the watched UEs based on the CQI assuming Nmax CoMP point. Further, the scheduler 324 calculates T(N-1) by using combined CQI when N-.1 CoMP.points !are,used for „synchronous-transmission, on the allocated RB.s of.the watched UE.
[0091]
The scheduler 324 compares the calculated T(N) and T(N-1). If N>1 and T(N)1 and T(N)
The scheduler 324 in the example 4 will be explained. The explanation of a configuration similar to that of the above-mentioned example is omitted.
[0103]
The scheduler 324 calculates the metric of each of the resource blocks (RBs) of all UEs. For a
candidate CoMP UE, the metrics are calculated by using the CQl assuming all combinations of the CoMP points.
[0104]
The scheduler 324 ranks the calculated metrics RB by RB.
[0105]
Among the cafcufated metrics, the scheduler 324 finds the RB with the highest metric and regards its owner UE as the watched UE, also called the highest UE.
[0106]
When the watched UE is a candidate CoMP UE, the scheduler 324 decides the number of CoMP
points "N0" for this watched UE based on the RB with the highest metric. Then, CoMP points are
decided too, since above process have already assumed combination of CoMP point.
[0107]
After CoMP point decision, the scheduler 324 deletes the metric with N unequal to N0 and updates the metric ranking by using the metric of N=N0. In addition, the scheduler 324 allocates the RBs to the watched UE when the top metric belongs to this UE.
[0108]
The scheduler 324 perform the above mentioned process until.all the RBs are allocated, or all UEst are allocated, where the candidate CoMP UEs have also decided their CoMP point(s).
[0109]
The above-mentioned process is carried out cell-by-cell. After all cells have finished the above
mentioned process, the scheduler 324 coordinates the resource allocation for the CoMP UEs. When a CoMP UE is allocated different RB(s) from its CoMP points, the scheduler 324 coordinates different owner UEs on this uncommonly allocated RB. One coordination method is to release the uncommonly allocated RB to another UE, who can be the UE with lower metric than the allocated UE in the metric ranking list of this RB. When a CoMP UE is not allocated the RB of CoMP points as a result of coordination, the scheduler will allocate the CoMP UE in the next scheduling time.
[0110]
An operation of this example is described below. For Example 4, the detailed flowchart of step 406 and 407 in the Step 404 is shown in Fig. 8 and described as follows. In step 406 of Fig 8, there are two steps, step 801 and 802. In step 407 of Fig. 8, there are ten steps, step 803 to 812.
[0111]
In step 801, the number of CoMP points of candidate CoMP UEs is initialized as N=1~Nmax, while for other UEs, N=1. Here, the cell index starts from 0. The starts cell index is optional.
[0112]
Next, in step 802, the metric calculation and ranking use CQI of different UEs with initialized N CoMP points for channel-dependent scheduling. Here, the PF metric is explained in Math. 1.
[0113]
In step 803, the scheduler firstly finds the RB with the highest metric and regards its owner UE as the watched UE called the highest UE.
[0114]
In step 804, the scheduler checks whether the highest UE is a candidate CoMP UE or not. If yes, go to step 805; otherwise, go to step 808.
[0115]
In step 805, the scheduler checks whether this candidate CoMP UE has already decided N0 or not. If yes, go to step 808; otherwise, go to step 806.
[0116]
In step 806, the number N0 of CoMP points is decided as N=N0 since N0 CoMP points can achieve the highest metric. Then, CoMP points are decided too.
[0117]
After the CoMP point decision, in step 807, the scheduler deletes the UE's metric with N unequal to N0 from the metric ranking list. If the highest metric of unallocated RB belongs to the current UE, this RB is allocated in the step 808.
[01181
In step 809, scheduler checks whether all RBs or UEs are allocated. If yes, go to step 810; otherwise, go to step 802 in block 406 to update the metric ranking.
[0119]
After the-reseurce-allocation and point decision have been completed in the current cell.the step 810.. _ of the scheduler checks whether the current cell is the last cell. If yes, go to step 811; otherwise, go to step 812 to move to the next cell and then go to step 802 in block 406 to calculate and rank the metrics of the UEs belong to the cell.
[0120]
In step 811, the scheduler coordinates after resource allocation for each cell. For each CoMP UE, if the allocated RB is not commonly allocated at each CoMP point, the scheduler will replace the owner of this RB by the next UE with lower PF metric. The output of step 811 is the information about allocated RBs for each UE.
[0121]
Making a configuration like this example enables to improve both average sector throughput and cell-edge user throughput.
[0122]
The scheduler 324 in the example 5 will be explained. The explanation of a configuration similar to that of the above-mentioned example is omitted.
[0123]
The scheduler 324 calculates the metric of each RB by assuming that the number of CoMP points is simply initialized as single point, i.e., N=1 for candidate CoMP UEs as well as for other non-CoMP UEs.
[0124]
Among the calculated metrics, the scheduler 324 finds the RB with highest metric and regards its owner UE as the watched UE, also called the highest UE.
[0125]
When the watched UE is a candidate CoMP UE, the scheduler 324 firstly finds out the RB with the highest metric belong to this watched UE and calculates the T(N) based on the CQI assuming N=1 CoMP point. Further, the scheduler 324 calculates T(N+1) by using combined CQI when N+1 CoMP points are used for synchronous transmission on this RB.
[0126]
The scheduler 324 compares the calculated T(N) and T(N+1). If the rate T(N+1)>T(N), the scheduler increases N as N=N+1. This iteration process will stop only if the maximum T(N) is achieved. Therefore, the number of the CoMP points is decided as N=N0. Then, CoMP points are decided too, since above process have already assumed combination of CoMP point.
[0127]
After CoMP point decision, the scheduler 324 releases the allocated RBs with metric of N unequal to No of watched UE. In addition, the metric ranking is updated by deleting the metrics of the watched UE or calculating the metric as N=N0. After updating the metric ranking, the scheduler 324 allocates the RBs to the watched UE of which the metric is highest.
[0128] The scheduler 324 performs the above-mentioned process until all the RBs are allocated or all UEs are allocated, where the candidate CoMp LIEs have decided their CoM P. point(s).,., „:. _
[0129]
The above-mentioned process is carried out cell-by-ce!l. After all cells have finished the above
mentioned process, the scheduler 324 coordinates the resource allocation for the CoMP UEs. When a CoMP UE is allocated different RB(s) from its CoMP points, the scheduler 324 makes coordination for this uncommonly allocated RB. One coordination method is to release the uncommonly allocated RB to another UE, who can be the UE with lower metric than the allocated UE in the metric ranking list of this RB. When a CoMP UE is not allocated the RB of CoMP points as a result of coordination, the scheduler will allocate the CoMP UE in the next scheduling time.
[0130]
An operation of this example is described below. For Example 5, the detailed flowchart of block 406 and 407 in the Step 404 is shown in Fig. 9 and described as follows. In block 406 of Fig 9, there are two steps, step 901 and 902. In block 407 of Fig. 9, there are eleven steps, step 903 to 912.
[0131]
In step 901, the number of CoMP points is initialized as N=1 and the cell index starts from 0. Next, in step 902, the metric calculation and ranking use CQI assuming initialized N=1 CoMP points for CoMP UE candidates as well as for other non-CoMP UE to carry out channel-dependent scheduling. Here, the PF metric is explained in Math. 1.
[0132]
In block 407 of Fig. 9, the step 903 firstly finds the RB with highest metric. Among the allocated RBs, the scheduler 324 finds the RB with highest metric and regards its owner UE as the watched UE, also called the highest UE. Then, the step 904 checks whether the highest UE (the UE with highest PF metric) is a candidate CoMP UE or not. If yes, go to step 905; otherwise, go to step 909. The step 905 checks whether this candidate CoMP UE has already decided N0 or not. If yes, go to step 909; otherwise, go to step 906. In step 906, the achievable normalized sum rate of this candidate CoMP UE is calculated based on the allocated RB(s). T(N) and T(N+1) are calculated by assuming N and N+1 CoMP points, respectively. Next, the step 907 compares T(N) and T(N+1) and checks whether NT(N) or not.
[0133]
If yes, go to step 908, where N is increased as N=N+1; otherwise, go to step 909. Accordingly, the number of CoMP points is decided as N=N0 to achieve highest metric. If the highest metric of
unallocated RB is the metric of the current UE with N0 CoMP points, this RB is allocated in the step 909.
[0134]
After the allocation, the step 910 checks whether all RBs or UEs are allocated. If yes, go to step 911;
otherwise, go to step 902 in block 406 to update the metric ranking.
[0135]
After the resource allocation and point decision have been completed in the current cell, the step 911 checks whether the current cell is the last cell. If yes, go to step 912; otherwise, go to step 913 to move to the next cell and then go to step 902 in block 406 to calculate and rank the metrics of the UEs belong to the, cell.
[0136]
In step 912, the scheduler coordinates after resource allocation for each cell. For each CoMP UE, if the allocated RB is not commonly allocated at each CoMP point, the scheduler will replace the owner of this RB by the next UE with lower PF metric. The output of step 912 is the information about allocated RBs for each UE.
[0137]
Making a configuration like this example enables to improve both average sector throughput and
cell-edge user throughput.
[0138]
The scheduler 324 in the example 6 will be explained. The explanation of a configuration similar to that of the above-mentioned example is omitted.
[0139]
The scheduler 324 calculates the metric of each RB by assuming that the number of CoMP points is simply initialized as single point, i.e., N= Nmax for candidate CoMP UEs but N=1 for other non-CoMP UE.
[0140]
The scheduler 324 ranks the calculated metrics RB by RB. Among the calculated metrics, the
scheduler 324 finds the RB with highest metric and regards its owner UE as the watched UE, also
called the highest UE.
[0141]
When the watched UE is a candidate CoMP UE, the scheduler 324 firstly finds the RB with highest metric belong to this watched UE. The scheduler 324 calculates T(N=Nmax) of the watched UEs based on the CQI assuming Nmax CoMP point. Further, the scheduler 324 calculates a sum rate T(N-1) by using combined CQI when N-1 CoMP points are used for synchronous transmission on the allocated RBs of the watched UE.
[0142]
The scheduler 324 compares the calculated T(N) and T(N-1). If N>1 and T(N)1 and T(N)
The example 7 is explained below. The scheduler 324 in the example 7 will be explained. The
explanation of a configuration similar to that of the above-mentioned example is omitted.
[0156]
The scheduler 324 calculates the metric of each of the resource blocks (RBs) of all UEs. For a
non-CoMP UE, e.g., a UE close to cell-center, only single point transmission is used, i.e., N=1. For a candidate CoMP UE, the metrics are calculated by using the CQI assuming all combinations of the CoMP points. Here, the combination of the CoMP points means to select different number of CoMP points. The number of the CoMP points, N, is a numerical value ranging from "1 to Nmax°, where the maximum value Nmax of the CoMP point may be pre-decided.
[0157]
The scheduler 324 ranks the calculated metrics RB by RB. Among the calculated metrics, the
scheduler 324 finds and allocates the RB with the highest metric and regards its owner UE as the
watched UE, also called the highest UE.
[0158]
When the watched UE is a candidate CoMP UE, the scheduler 324 temporarily decides number of
CoMP points based on this allocated RB only. The temporarily decided number of CoMP points is called N0. Next, the scheduler 324 deletes the metrics assuming N unequal to N0 in the metric ranking list and then finds out the unallocated RBs for the following allocation, whose highest metric is assuming N equal to N0. Then, the scheduler 324 tries to allocate one more RB to the watched UE and adjust N0 to achieve highest normalized sum rate of allocated RBs. The RBs are allocated one-by-one for a candidate CoMP UE and the N0 is adjusted as followings.
[0159]
Next, the scheduler 324 calculates T(N0) of all the allocated RBs based on the CQI assuming N0 CoMP points. Further, the scheduler 324 calculates the sum rate T(N0-1) by using combined CQI when N0-1 CoMP points are used for synchronous transmission on the allocated RBs of the watched UE. The scheduler 324 compares the calculated rate T(N0) and T(N0-1). If N0>1 and the rate T(No)1 and T(N)
The example 8 is explained below. The scheduler 324 in the example 8 will be explained. The explanation of a configuration similar to that of the above-mentioned example is omitted.
[0175]
The scheduler 324 calculates the metric of each RB of all UEs. For a candidate CoMP UE, the metrics are calculated by using the CQI assuming N=1.
[0176]
The scheduler 324 ranks the calculated metrics RB by RB.
[0177]
Among the calculated metrics, the scheduler 324 finds and allocates the RB with the highest metric and regards its owner UE as the watched UE, also called the highest UE.
[0178]
When the watched UE is a candidate CoMP UE, the scheduler 324 firstly calculates T(N=1) of the watched UEs based on the CQI assuming N=1 CoMP point. Further, the scheduler 324 caicuiates T(N+1) by using combined CQI when N+1 CoMP points are used for synchronous transmission on the allocated RBs of the watched UE. The scheduler 324 compares the calculated T(N=1) and T(N+1). If T(N+1)>T(N), the scheduler increases N as N=N+1. This iteration process will stop only if the maximum T(N0) is achieved. Therefore, N is temporally decided as N0. Next, the scheduler 324 ..-.,,...,-.,updates-the .metric ranking list by,using.metric ktefcle-for the watched .UE and then finds ,.finds out ..the unallocated RBs with the highest metric belongs to the watched UE. Then, the scheduler 324 tries to allocate one more RB to the watched UE and adjust N0 to achieve highest sum rate of allocated RBs. The RBs are allocated one-by-one for a candidate CoMP UE and the N0 is adjusted as followings.
[0179]
Next, the scheduler 324 calculates T(N0) of the allocated RBs based on the CQI assuming N0 CoMP points. Further, the scheduler 324 calculates T(N0-1) by using combined CQI when N0-1 CoMP points are used for synchronous transmission on the allocated RBs of the watched UE. The scheduler 324 compares the calculated T(N0) and T(N0-1). If N0>1 and T(N0)T(N) or not. If yes, go to step 1208, where N is increased as N=N+1; otherwise, N is temporally decided as N0 in step 1209.
[0189]
When the allocated RB is not the first one, the step 1210 calculates the achievable normalized sum rate of this candidate CoMP based on CQI of all the allocated RBs. T(N0) and T(N0-1) are calculated by assuming N0 and N0-1 CoMP points, respectively. Next, the step 1211 compares T(N0) and T(N0-1) and checks whether N0>1 and T(N0)
The scheduler 324 in the example 9 will be explained. The explanation of a configuration similar to that of the above-mentioned example is omitted.
[0194]
The scheduler 324 calculates the metric of each RB of the UE by assuming that the number of CoMP points is simply initialized as single point, i.e., N=Nmax (Nmax: maximum number of CoMP points) for eandidate-CoMP-UEs but N=1 for other non-CoMP~UE
[0195]
The scheduler 324 calculates the metrics of all UEs on each RBs and ranks the calculated metrics
RB by RB.
[0196]
Among the allocated RBs, the scheduler 324 finds and allocates the RB with highest metric and
regards its owner UE as the watched UE, also called the highest UE.
[0197]
When the watched UE is a candidate CoMP UE, the scheduler 324 calculates T(N=Nmax) of the
watched UEs based on the CQI assuming Nmax CoMP point. Further, the scheduler 324 calculates T(N-1) by using combined CQI when N-1 CoMP points are used for synchronous transmission on the allocated RBs of the watched UE. The scheduler 324 compares the calculated T(N) and T(N-1). If N>1 and T(N)1 and T(N0)1 and T(N)1 and T(N0)
| # | Name | Date |
|---|---|---|
| 1 | 2364-CHENP-2012 POWER OF ATTORNEY 15-03-2012.pdf | 2012-03-15 |
| 2 | 2364-CHENP-2012 PCT 15-03-2012.pdf | 2012-03-15 |
| 3 | 2364-CHENP-2012 FORM-5 15-03-2012.pdf | 2012-03-15 |
| 4 | 2364-CHENP-2012 FORM-3 15-03-2012.pdf | 2012-03-15 |
| 5 | 2364-CHENP-2012 FORM-2 15-03-2012.pdf | 2012-03-15 |
| 6 | 2364-CHENP-2012 FORM-18 15-03-2012.pdf | 2012-03-15 |
| 7 | 2364-CHENP-2012 FORM-1 15-03-2012.pdf | 2012-03-15 |
| 8 | 2364-CHENP-2012 DRAWINGS 15-03-2012.pdf | 2012-03-15 |
| 9 | 2364-CHENP-2012 DESCRIPTION (COMPLETE) 15-03-2012.pdf | 2012-03-15 |
| 10 | 2364-CHENP-2012 CORRESPONDENCE OTHERS 15-03-2012.pdf | 2012-03-15 |
| 11 | 2364-CHENP-2012 CLAIMS 15-03-2012.pdf | 2012-03-15 |
| 12 | 2364-CHENP-2012 ABSTRACT 15-03-2012.pdf | 2012-03-15 |
| 13 | 2364-CHENP-2012 FORM-13 15-03-2012.pdf | 2012-03-15 |
| 14 | 2364-CHENP-2012 FORM-1 11-09-2012.pdf | 2012-09-11 |
| 15 | 2364-CHENP-2012 CORRESPONDENCE OTHERS 11-09-2012.pdf | 2012-09-11 |
| 16 | 2364-CHENP-2012 FORM -3 08-10-2012.pdf | 2012-10-08 |
| 17 | 2364-CHENP-2012 CORRESPONDENCE OTHERS 08-10-2012.pdf | 2012-10-08 |
| 18 | 2364-CHENP-2012 CORRESPONDENCE OTHERS 15-07-2013.pdf | 2013-07-15 |
| 19 | 2364-CHENP-2012-FER.pdf | 2018-03-26 |
| 20 | 2364-CHENP-2012-AbandonedLetter.pdf | 2018-10-12 |
| 1 | 2364_CHENP_2012_15-02-2018.pdf |