A 5G cell throughput optimization method in eMBB scenario

Abstract

The present invention discloses a method for optimizing the throughput of a 5G cell in an eMBB scenario. The method solves the problem that the prior art is not suitable for network optimization in an eMBB scenario, resulting in a reduction in user rate perception. The method includes normalizing the throughput performance index, obtaining a synthetic evaluation index of each cell by weighting; determining the cell to be optimized and the target cell; adjusting the aggregation level, and performing a load transfer operation according to the spatial margin of the target cell; setting a convergence condition, and performing an optimization qualification judgment on the optimized cell. The present invention obtains a synthetic evaluation index of each cell in a weighted form based on the eMBB scenario indicators of throughput, jitter and oscillation of each cell, obtains the cell that needs to be optimized the most by calculating the minimum value, and selects the target cell with balanced interaction by combining the load and throughput jitter indicators. Load optimization is implemented by adjusting the aggregation level and load transfer, thereby improving the throughput performance in the eMBB scenario and providing a guarantee for the service quality of user services.

Description

Technical Field

[0001] The present invention relates to the field of communication technologies, and in particular to a 5G cell throughput optimization method in an eMBB scenario. Background Art

[0002] Compared with 4G networks, 5G networks have been greatly improved in all aspects, showing the characteristics of higher transmission rate, lower latency, and greater connection density, meeting the requirements of multi-level heterogeneous and multi-scenario network architecture design. However, like 4G networks, as the scale of network deployment expands, 5G networks also need to be continuously optimized to improve network quality and user perception.

[0003] For network optimization methods, such as the invention patent with the patent number ZL201510188252.6, named as an LTE optimization method based on cell performance evaluation, it discloses that by analyzing various service indicators of the cell, the target cell is found by filtering and weighting, and then autonomous optimization is implemented to achieve the purpose of improving cell performance. However, this optimization method is only universal. Compared with 4G, 5G network has more slices, especially in the eMBB scenario, it pursues more service rate, so universal indicators are no longer worth promoting; in addition, the load transfer method adopted in the above patent does not take into account all the resources occupied by a user, so simply transferring part of the load will reduce the user's rate perception. Summary of the invention

[0004] The present invention mainly solves the problem that the existing technology is not suitable for network optimization in eMBB scenarios, resulting in reduced user rate perception, and provides a 5G cell throughput optimization method in eMBB scenarios. Starting from the core indicators closely related to throughput, the evaluation indicators are obtained by setting weights, and then the cells to be optimized and the target cells for load transfer are screened out, and the goal of optimizing cell performance is achieved through hierarchical optimization measures to improve customer perception.

[0005] The above technical problem of the present invention is mainly solved by the following technical solution: A 5G cell throughput optimization method in an eMBB scenario comprises the following steps:

[0006] Step 1: Obtain cell load and throughput performance indicators;

[0007] Step 2: normalize the throughput performance index and obtain the composite evaluation index of each cell by weighting;

[0008] Step 3: Determine the cell to be optimized and the target cell;

[0009] Step 4: Adjust the aggregation level and perform load transfer operations according to the spatial margin of the target cell;

[0010] Step 5: Set convergence conditions and make optimization qualification judgment on the optimized cells.

[0011] The present invention measures the throughput, jitter and oscillation of each cell, obtains the composite evaluation index of each cell in a weighted form, obtains the cell most in need of optimization by calculating the minimum value, selects the target cell with balanced interaction in combination with the load and throughput jitter index, implements load optimization by adjusting the aggregation level and load transfer, and thus improves the throughput performance in the eMBB scenario, providing guarantee for the service quality of user services.

[0012] As a preferred solution, the step 1 is specifically as follows:

[0013] NBS includes n neighboring cells n ={gNB1,gNB2,…,gNB n}, the corresponding load {Cld1, Cld2,…, Cld n}、Total number of physical resource blocks PRB={Prb1,Prb2,…,Prb n}, Throughput THR = {Thr1,Thr2,…,Thr n}、Thoutput jitter THV={Thv1,Thv2,…,Thv n} and throughput oscillation THH={Thh1,Thh2,…,Thh n}.

[0014] As a preferred solution, the step 2 specifically includes:

[0015] (2-1): Set the weights δ of throughput, throughput jitter, and throughput oscillation respectively t ∈[0.3,1),δ v ∈(0,δ t ], δ h =1-δ t -δ v Throughput is a positive indicator, while throughput jitter and throughput oscillation are negative indicators and require reverse processing.

[0016] (2-2): Calculate the maximum value of throughput jitter and throughput vibration of all cells

[0017] MTHV=max(Thv1,Thv2,…,Thv n ),

[0018] MTHH=max(Thh1,Thh2,…,Thh n );

[0019] (2-3): Calculate the reverse throughput jitter and reverse throughput oscillation for each cell

[0020] R i =MTHV-Thv i ,

[0021] R i =MTHH-Thh i ;

[0022] Calculate the mathematical expectation of the reverse throughput jitter

[0023]

[0024] Calculate the mathematical expectation of the reverse throughput oscillation

[0025]

[0026] (2-4): Calculate the mathematical expectation of the throughput indicator

[0027]

[0028] Calculate the same-direction normalized values ​​of the three indicators for each cell

[0029]

[0030]

[0031]

[0032] Computing synthetic evaluation metrics

[0033] CmpI i =δ t *OThr i +δ v *OThv i +δ h *OThh i .

[0034] As a preferred solution, the step three specifically includes:

[0035] (3-1): Set the load threshold CLd th ,CLd th ∈(0.5,1);

[0036] (3-2): Select the cell gNB with the minimum composite index value s As the cell to be optimized, select the cell with the smallest throughput jitter as the target cell gNB t ;

[0037] (3-3): Determine whether the target cell meets the condition Cld t >CLd th If yes, the cell with the smallest load is selected as the target cell gNB t .

[0038] As a preferred solution, the step 4 specifically includes:

[0039] (4-1): Improve gNB in ​​the cell to be optimized s The PDCCH channel aggregation level is increased until the aggregation level is 16; according to 3GPP38.211 Table 7.3.2.1-1, the aggregation levels of the PDCCH channels in 5G cells are {1, 2, 4, 8, 16}, which are increasing in sequence. s If the PDCCH aggregation level has reached 16, no further adjustment will be made.

[0040] (4-2): Calculate the mathematical expectation of the load of all cells

[0041]

[0042] Determine whether the conditions are met If not, cancel the load transfer; if so, in the gNB of the cell to be optimized s and target cell gNB t Load transfer between

[0043] (4-3): Obtain the gNB of the cell to be optimized s The gNB in ​​the cell to be optimized s and target cell gNB t For users in the overlapping coverage area, these users are arranged in ascending order according to the amount of resources they apply for, and the first user is selected and the physical resource block occupied by the user is set to Δ s , calculate the target cell gNB t Space margin

[0044] Sp t =CLd th -CLd t ,

[0045] Determine whether the condition Δ is met s / Prb t ≤Sp t If yes, remove the user from the gNB in ​​the cell to be optimized s Move to the target cell gNB t Update the load Cld s =Cld s -Δ s / Prb s , Cld t=Cld t +Δ s / Prb t ; If not, cancel load transfer.

[0046] As a preferred solution, step five specifically includes:

[0047] (5-1): Set the indicator convergence threshold Ids th , Ids th ∈[0.5,1), optimization times Nm opt , optimization times threshold Op th , Op th >1,

[0048] Calculate Nm opt =Nm opt +1;

[0049] (5-2): Determine whether the condition Nm is met opt ≥Op th , if yes, the optimization ends; if no, go to the next step;

[0050] (5-3): Determine the optimized cell gNB s The synthetic evaluation index CmpI s Whether the condition CmpI is met s ≥Ids th If yes, the optimization ends; if no, repeat steps 2 to 5 until the conditions are met.

[0051] Therefore, the advantages of the present invention are: according to the eMBB scenario indicators of throughput, jitter and oscillation of each cell, the composite evaluation indicator of each cell is obtained in a weighted form, the cell most in need of optimization is obtained by calculating the minimum value, and the target cell with balanced interaction is screened out by combining the load and throughput jitter indicators. Load optimization is implemented by adjusting the aggregation level and load transfer, thereby improving the throughput performance in the eMBB scenario and providing guarantee for the service quality of user services. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a schematic diagram of a process of the present invention;

[0053] Figure 2 is a cell throughput variation trend diagram during the optimization process of the present invention;

[0054] Figure 3 It is a graph showing changes in the synthetic evaluation index of the cell to be optimized in the present invention. DETAILED DESCRIPTION

[0055] The technical solution of the present invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0056] Example:

[0057] This embodiment provides a 5G cell throughput optimization method in an eMBB scenario, such as Figure 1 As shown, the following steps are included:

[0058] Step 1: Obtain cell load and throughput performance indicators;

[0059] Specifically, it includes n cells NBS which are neighboring cells. n ={gNB1,gNB2,…,gNB n}, the corresponding load {Cld1, Cld2,…, Cld n}、Total number of physical resource blocks PRB={Prb1,Prb2,…,Prb n}, Throughput THR = {Thr1,Thr2,…,Thr n}、Thoutput jitter THV={Thv1,Thv2,…,Thv n} and throughput oscillation THH={Thh1,Thh2,…,Thh n}.

[0060] Step 2: Normalize the throughput performance index and obtain the composite evaluation index of each cell by weighting; specifically, it includes:

[0061] (2-1): Set the weights δ of throughput, throughput jitter, and throughput oscillation respectively t ∈[0.3,1),δ v ∈(0,δ t ], δ g =1-δ t -δ v Throughput is a positive indicator, while throughput jitter and throughput oscillation are negative indicators and require reverse processing.

[0062] (2-2): Calculate the maximum value of throughput jitter and throughput vibration of all cells

[0063] MTHV=max(Thv1,Thv2,…,Thv n ),

[0064] MTHH=max(Thh1,Thh2,…,Thh n );

[0065] (2-3): Calculate the reverse throughput jitter and reverse throughput oscillation for each cell

[0066] R i =MTHV-Thv i,

[0067] R i =MTHH-Thh i ;

[0068] Calculate the mathematical expectation of the reverse throughput jitter

[0069]

[0070] Calculate the mathematical expectation of the reverse throughput oscillation

[0071]

[0072] (2-4): Calculate the mathematical expectation of the throughput indicator

[0073]

[0074] Calculate the same-direction normalized values ​​of the three indicators for each cell

[0075]

[0076]

[0077]

[0078] Computing synthetic evaluation metrics

[0079] CmpI i =δ t *OThr i +δ v *OThv i +δ h *OThh i .

[0080] Step 3: Determine the cell to be optimized and the target cell; specifically include:

[0081] (3-1): Set the load threshold CLd th ,CLd th ∈(0.5,1);

[0082] (3-2): Select the cell gNB with the minimum composite index value s As the cell to be optimized, select the cell with the smallest throughput jitter as the target cell gNB t ;

[0083] (3-3): Determine whether the target cell meets the condition Cld t >CLd th If yes, the cell with the smallest load is selected as the target cell gNB t .

[0084] Step 4: Adjust the aggregation level and perform load transfer operations according to the spatial margin of the target cell; specifically, include:

[0085] (4-1): Improve gNB in ​​the cell to be optimized s According to 3GPP 38.211 Table 7.3.2.1-1, the aggregation levels of 5G cell PDCCH channels are {1, 2, 4, 8, 16}, which increase in turn. s If the PDCCH aggregation level has reached 16, no further adjustment will be made.

[0086] (4-2): Calculate the mathematical expectation of the load of all cells

[0087]

[0088] Determine whether the conditions are met If not, cancel the load transfer; if so, in the gNB of the cell to be optimized s and target cell gNB t Load transfer between

[0089] (4-3): Obtain the gNB of the cell to be optimized s The gNB in ​​the cell to be optimized s and target cell gNB t For users in the overlapping coverage area, these users are arranged in ascending order according to the amount of resources they apply for, and the first user is selected and the physical resource block occupied by the user is set to Δ s , calculate the target cell gNB t Space margin

[0090] Sp t =CLd th -CLd t ,

[0091] Determine whether the condition Δ is met s / Prb t ≤Sp t If yes, remove the user from the gNB in ​​the cell to be optimized s Move to the target cell gNB t Update the load Cld s =Cld s -Δ s / Prb s , Cld t =Cld t +Δ s / Prb t ; If not, cancel load transfer.

[0092] Step 5: Set convergence conditions and make optimization qualification judgment for the optimized cells. Specifically include:

[0093] (5-1): Set the indicator convergence threshold Ids th , Ids th ∈[0.5,1), optimization times Nm opt , optimization times threshold Op th , Op th >1,

[0094] Calculate Nm opt =Nm opt +1;

[0095] (5-2): Determine whether the condition Nm is met opt ≥Op th , if yes, the optimization ends; if no, go to the next step;

[0096] (5-3): Determine the optimized cell gNB s The synthetic evaluation index CmpI s Whether the condition CmpI is met s ≥Ids th If yes, the optimization ends; if no, repeat steps 2 to 5 until the conditions are met.

[0097] The following is a specific example of this method. Taking n=10 as an example, the cell load and performance of the 5G system are shown in Table 1:

[0098] Table 1 Cells and corresponding service indicators

[0099]

[0100] The basic data are shown in Table 2:

[0101] Table 2 Basic data

[0102] project data Operating frequency (GHz) 2.6 Operating bandwidth (MHz) 100 Subcarrier bandwidth index 1 <![CDATA[Throughput weight δ t > 40％ <![CDATA[Throughput jitter weight δ v > 40％ <![CDATA[Throughput oscillation weight δ h > 20％ <![CDATA[Load threshold CLd th > 0.7 <![CDATA[Indicator convergence threshold Ids th > 0.8 <![CDATA[Number of optimization times Nm opt > 10

[0103] Step 1: Obtain cell load and throughput performance indicators;

[0104] Step 2: normalize the throughput performance index and obtain the composite evaluation index of each cell by weighting;

[0105] Calculate the maximum value of throughput jitter and throughput vibration of all cells

[0106] MTHV=max(Thv1,Thv2,…,Thv n )=0.31,

[0107] MTHH=max(Thh1,Thh2,…,Thh n )=0.97,

[0108] Calculate the reverse throughput jitter for each cell

[0109] R i =MTHV-Thv i

[0110] ={0.23,0.19,0.25,0.08,0.2,0,0.14,0.07,0.19,0.22},

[0111] Calculating reverse throughput oscillations

[0112] R i =MTHH-Thh i

[0113] ={0.45,0.09,0.18,0.32,0.21,0.15,0.63,0,0.74,0.59};

[0114] Calculate the mathematical expectation of the reverse throughput jitter

[0115]

[0116] Calculate the mathematical expectation of the reverse throughput oscillation

[0117]

[0118] Calculate the mathematical expectation of the throughput metric

[0119]

[0120] For each cell, the normalized values ​​of the three indicators are calculated, which are

[0121]

[0122]

[0123]

[0124] Computing synthetic evaluation metrics

[0125] CmpI i =δ t *OThr i +δ v *OThv i +δ h *OThh i

[0126] ={1.37,1.13,1.04,0.83,0.97,0.21,1.09,0.34,1.81,1.23}.

[0127] Step 3: Determine the cell to be optimized and the target cell;

[0128] Select the cell gNB with the minimum composite index value (0.21) s＝6 As the cell to be optimized, select the cell gNB with the smallest throughput jitter (0.06) t＝3 As a target community;

[0129] Set the load threshold CLd th =0.7, does not meet the condition Cld t =0.35>CLd th =0.7, still select cell gNB t＝3 as the target area.

[0130] Step 4: Adjust the aggregation level and perform load transfer operations according to the spatial margin of the target cell;

[0131] Increase the aggregation level of the PDCCH channel. If it has reached 16, no further adjustment will be made.

[0132] Calculate the mathematical expectation of the load of all cells

[0133]

[0134] Satisfy the conditions In the cell gNB s＝6 and gNB t＝3 Load transfer between

[0135] Get the gNB of the cell to be optimized s＝6 The gNB in ​​the cell to be optimized s＝6 and target cell gNB t＝3 For users in the overlapping coverage area, these users are arranged in ascending order according to the amount of resources they apply for, and the first user is selected and the physical resource block occupied by the user is set to Δ s , calculate the target cell gNB t Space margin

[0136] Sp t =CLd th -CLd t =0.35,

[0137] Set the condition to meet Δ s / Prb t ≤Sp t , remove the user from the gNB in ​​the cell to be optimized sMove to the target cell gNB t Update the load Cld s =Cld s -Δ s / Prb s , Cld t =Cld t +Δ s / Prb t ;

[0138] In the judgment process, if or Δ s / Prb t >Sp t , then cancel this load transfer.

[0139] Step 5: Set convergence conditions and make optimization qualification judgment on the optimized cells.

[0140] Calculate Nm opt =Nm opt +1=1;

[0141] Determine whether the condition Nm is met opt ≥Op th If yes, the optimization is finished; if no, the optimized cell gNB is determined s The synthetic evaluation index CmpI s Whether the condition CmpI is met s ≥Ids th If yes, the optimization ends; if no, repeat steps 2 to 5 until the conditions are met.

[0142] The following simulation experiment is used to illustrate

[0143] The method of the present invention is simulated on the MATLAB platform. The basic data information is shown in Tables 1 and 2 above. Certain users are randomly added and random services are configured. The results are shown in Tables 1 and 2. Figure 2 and Figure 3 .

[0144] like Figure 2 The cell throughput changes during the optimization process shown in the figure. By adjusting the aggregation level of PDCCH, the problem of DCI missed detection can be avoided. By implementing load transfer between cells, the problem of cell overload can be avoided. Both methods can effectively improve the throughput performance of 5G cells. Of course, this improvement is a spiral upward.

[0145] like Figure 3 As shown, the composite evaluation index of the cell to be optimized changes. As the number of optimizations increases, the composite evaluation index of the cell gradually increases and approaches the set index convergence threshold.

[0146] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A 5G cell throughput optimization method in an eMBB scenario, characterized in that: The following steps are involved: Step 1: Obtain cell load and throughput performance indicators; Step 2: Normalize the throughput, throughput jitter, and throughput oscillation respectively, and obtain the composite evaluation index of each cell by weighting; Step 3: Select the cell with the smallest composite index value as the cell to be optimized, and select the cell with the smallest throughput jitter as the target cell; Step 4: Increase the PDCCH channel aggregation level of the cell to be optimized until it reaches 16, obtain users in the cell to be optimized that are located in the overlapping coverage area of ​​the cell to be optimized and the target cell, select the first user in ascending order of resource application, and move the user from the cell to be optimized to the target cell if the space margin of the target cell is greater than or equal to the quotient of the physical resource blocks occupied by the user and the total number of physical resource blocks of the target cell; Step 5: Set convergence conditions and make optimization qualification judgment on the optimized cells.

2. The 5G cell throughput optimization method in an eMBB scenario according to claim 1 is characterized in that The step 1 is specifically as follows: NBS includes n neighboring cells n ={gNB1,gNB2,…,gNB n }, the corresponding load {Cld1, Cld2,…, Cld n }、Total number of physical resource blocks PRB={Prb1,Prb2,…,Prb n }, Throughput THR = {Thr1,Thr2,…,Thr n }、Thoutput jitter THV={Thv1,Thv2,…,Thv n } and throughput oscillation THH={Thh1,Thh2,…,Thh n }.

3. The 5G cell throughput optimization method in an eMBB scenario according to claim 2 is characterized in that The step 2 specifically includes: (2-1): Set the weights δ of throughput, throughput jitter, and throughput oscillation respectively t ∈[0.3,1),δ v ∈(0,δ t ], δ h =1-δ t -δ v ; (2-2): Calculate the maximum value of throughput jitter and throughput vibration of all cells MTHV=max(Thv1,Thv2,…,Thv n ), MTHH=max(Thh1,Thh2,…,Thh n ); (2-3): Calculate the reverse throughput jitter and reverse throughput oscillation for each cell RThv i =MTHV-Thv i , RThh i =MTHH-Thh i ; Calculate the mathematical expectation of the reverse throughput jitter Calculate the mathematical expectation of the reverse throughput oscillation (2-4): Calculate the mathematical expectation of the throughput indicator Calculate the same-direction normalized values ​​of the three indicators for each cell Computing synthetic evaluation metrics CmpI i =δ t *Other i +δ v *OtV i +δ h *Otthh i , 4. The 5G cell throughput optimization method in an eMBB scenario according to claim 3 is characterized in that The step three specifically includes: (3-1): Set the load threshold CLd th ,CLd th ∈(0.5,1); (3-2): Select the cell gNB with the minimum composite index value s As the cell to be optimized, select the cell with the smallest throughput jitter as the target cell gNB t ; (3-3): Determine whether the target cell meets the condition Cld t >CLd th If yes, the cell with the smallest load is selected as the target cell gNB t .

5. The 5G cell throughput optimization method in an eMBB scenario according to claim 4 is characterized in that The step 4 specifically includes: (4-1): Improve gNB in ​​the cell to be optimized s The PDCCH channel aggregation level is increased until the aggregation level is 16; (4-2): Calculate the mathematical expectation of the load of all cells Determine whether the conditions are met If not, cancel the load transfer; if so, in the gNB of the cell to be optimized s and target cell gNB t Load transfer between (4-3): Obtain the gNB of the cell to be optimized s The gNB in ​​the cell to be optimized s and target cell gNB t For users in the overlapping coverage area, these users are arranged in ascending order according to the resource application amount, and the first user is selected and the physical resource block occupied by the user is set to Δ s , calculate the target cell gNB t Space margin Sp t =CLd th -CLd t , Determine whether the condition Δ is met s / Prb t ≤Sp t If yes, remove the user from the gNB in ​​the cell to be optimized s Move to the target cell gNB t Update the load Cld s =Cld s -Δ s / Prb s , Cld t =Cld t +Δ s / Prb t ; If not, cancel load transfer.

6. The 5G cell throughput optimization method in an eMBB scenario according to claim 5 is characterized in that Step 5 specifically includes: (5-1): Set the indicator convergence threshold Ids th , Ids th ∈[0.5,1), optimization times Nm opt , optimization times threshold Op th , Op th >1, Calculate Nm opt =Nm opt +1; (5-2): Determine whether the condition Nm is met opt ≥Op th , if yes, the optimization ends; if no, go to the next step; (5-3): Determine the optimized cell gNB s The synthetic evaluation index CmpI s Whether the condition CmpI is met s ≥Ids th If yes, the optimization ends; if no, repeat steps 2 to 5 until the conditions are met.

Images