A 5G base station intelligent shutdown method based on neighbor cell energy efficiency

Abstract

The application discloses a 5G base station intelligent shutdown method based on adjacent area energy efficiency. The method solves the problem of lacking of effective method for 5G network optimization and energy consumption reduction in the prior art. The method comprises low energy efficiency cell screening, service type marking, weighted load evaluation and base station intelligent shutdown. The application screens out the base station with high energy efficiency ratio by analyzing the energy efficiency of each base station, analyzes each service in detail, calculates the performance of each service when occupying a single resource block, and then marks the service accordingly; according to the load of each adjacent base station of the base station, the weighted load is calculated, and the overall average load of the base station cluster is obtained; under the premise of guaranteeing user satisfaction, the 5G base station of low value service is shut down through the migration of adjacent base stations, so that the optimization of 5G network system energy consumption is realized.

Description

Technical Field

[0001] This invention relates to the field of 5G technology, and in particular to a method for intelligent shutdown of 5G base stations based on neighboring cell energy efficiency. Background Technology

[0002] For the telecommunications industry, with the deepening of the market economy and the gradual alignment of market operation models with international standards, competition among enterprises is becoming increasingly fierce. The growth of telecommunications operating revenue is slowing down, making cost-cutting and revenue-generating measures effective ways to improve operating profits. Major operators are increasing revenue by tapping into network potential and developing new services, while simultaneously striving to reduce operating expenses, especially electricity costs. Energy conservation and emission reduction not only align with my country's basic national conditions but also meet the survival and development strategies of enterprises. Moreover, the energy consumption of 5G equipment surpasses that of previous 3G / 4G networks. Therefore, optimizing the use of 5G networks, especially reducing energy consumption, is a crucial issue worthy of attention.

[0003] For example, Chinese invention patent ZL202110841312.5, entitled "A Method, Device, Equipment, and Storage Medium for Intelligent Shutdown of 5G Base Stations," distinguishes between services and shuts down base stations providing pure Non-GBR services and services with poor quality such as latency and packet loss, thereby achieving energy conservation. However, this patented technology still has some problems. First, and most importantly, it does not consider the energy efficiency of base station operation, which is the most direct way to assess the energy consumption of existing networks. Second, it only considers the performance of a single service and not the overall performance of the cell. Therefore, shutdown may lead to cell congestion or service dropouts. Summary of the Invention

[0004] This invention primarily addresses the lack of effective methods for optimizing 5G networks and reducing energy consumption in existing technologies, as well as the problem that existing intelligent shutdown technologies do not consider base station energy efficiency and overall base station performance, resulting in base station congestion and service dropouts even after shutdown. It provides a 5G base station intelligent shutdown method based on neighboring cell energy efficiency. Starting with analyzing the energy efficiency of neighboring base stations, it screens low-energy-efficiency target shutdown candidate stations by evaluating the percentage of neighboring base stations with higher energy efficiency than the serving base station. Combining the operational performance of various services, it eliminates high-performing service cells and selects target base stations with light weighted loads from the remaining candidate set for shutdown, thereby optimizing 5G network energy consumption.

[0005] The above-mentioned technical problems of the present invention are mainly solved by the following technical solution: a method for intelligent shutdown of 5G base stations based on neighboring cell energy efficiency, comprising:

[0006] Screen low-energy-efficiency base stations, calculate the energy efficiency of each base station, count the number of high-efficiency base stations among all its neighboring base stations, screen out base stations whose proportion of high-efficiency neighboring base stations is greater than a threshold, and generate the first set of candidates for shutdown targets.

[0007] Screen low-quality service base stations, calculate the single resource block latency and single resource block throughput for each service of each base station, select services with low single resource block latency or low throughput, mark services that are non-traffic-type and have high latency, mark services that are traffic-type and have low throughput, screen base stations with any service marked, and generate a second set of shutdown target candidates.

[0008] Screen low-weighted load base stations. Calculate the weighted load of each base station based on the average load of all its neighboring base stations. Screen base stations whose weighted load is less than the average weighted load of all base stations to generate a third set of candidate shutdown targets.

[0009] The base station at the intersection of three candidate shutdown target sets is taken as the target base station. The decision on whether to shut down the target base station is made based on the comparison of the cell overload threshold.

[0010] This invention selects base stations with lower energy efficiency than neighboring base stations within a certain proportion and includes them in the first candidate set for shutdown targets. Based on the number of resource particles allocated to each service, the unit particle latency and throughput of each service are calculated. Then, based on a threshold, high-quality services that meet the requirements are deduced, and these high-quality services are avoided as much as possible. Base stations containing low-quality services are included in the second candidate set for shutdown targets. Based on the weighted load of the base station and neighboring base stations, a weighted load is calculated, resulting in a third candidate set for shutdown. The intersection of these three candidate sets is calculated to finally determine the target base station that can be shut down. Before shutdown, all users and services of the base station need to be migrated to neighboring cells. By estimating the cell pressure brought by load migration, the target base station is finally evaluated. For those that meet the requirements, service migration is implemented all at once or in batches according to actual needs, and the source base station is shut down. The shutdown mechanism of this invention can achieve intelligent shutdown based on the needs of different users and different services, thereby achieving the goal of energy consumption optimization.

[0011] As a preferred approach, base stations that are neighboring cells are selected, and their corresponding load and power consumption information are obtained. Multiple users run on each base station, and each user carries out multiple services. The service type, service latency, number of allocated physical resource blocks, and corresponding throughput information for each service carried out by the user are set.

[0012] This scheme selects u neighboring base stations (gNBs). tar ={gNB1,gNB2,…,gNB u}, obtain the corresponding load {cLD1,cLD2,…,cLDu}, power consumption {Pwr1,Pwr2,…,Pwr} u}, each base station gNB k ∈gNB tar running k v One user, US k,i It is the i-th user. Each user conducts k vw Each business, SV k,i,j It is the j-th business. Then user US k,i The business being run is SV k,i ={SV k,i,j j∈(1,2,…,k) vw For each base station gNB k Configure user US k,i The business types of each business carried out STpe k,i,j ∈{"non-traffic class","traffic class"}, business latency SDly k,i,j (ms), the number of allocated physical resource blocks ANm k,i,j and the corresponding throughput SThr k,i,j (Mbps), where k∈[1,u], i∈[1,k] v ],j∈[1,k vw ].

[0013] As a preferred embodiment, the method of counting the number of high-efficiency base stations among all neighboring base stations of each base station includes:

[0014] Obtain all neighboring base stations for each base station;

[0015] Set an energy efficiency threshold and count the number of adjacent base stations whose throughput is not less than the sum of the current base station's throughput and the energy efficiency threshold. The throughput of each base station is the sum of the throughput of all services of all users of the base station.

[0016] throughput per base station cell SThr k =∑ i ∑ j SThr k,i,j The energy efficiency of the base station is

[0017] Set energy efficiency threshold EfV th Obtain each base station gNB k All adjacent base stations, Its number of adjacent base stations is n k For NCL k Each base station gNB zStatistically identify SThr that meets the conditions z ≥SThr k +EfV th Number of adjacent base stations k z .

[0018] As a preferred embodiment, the process of selecting base stations whose high-efficiency ratio among adjacent base stations is greater than a certain threshold includes:

[0019] The high-energy-efficiency ratio of adjacent base stations is calculated as the ratio of the number of adjacent base stations whose throughput is not less than the sum of the current base station's throughput and energy efficiency threshold to the total number of all adjacent base stations of the current base station; the high-energy-efficiency ratio of adjacent base stations Rt is obtained. k =k z / n k .

[0020] Set a ratio threshold, select base stations that meet the requirement that the proportion of high energy efficiency among adjacent base stations is greater than the ratio threshold, and form the first set of candidate base stations for shutdown targets.

[0021] Set the proportional threshold RtV th ∈(0,1); filter out Rt that satisfy the condition k >RtV th Each base station gNB k The selected base stations constitute the first set of candidate shutdown targets, gNB1. tclose .

[0022] As a preferred embodiment, the latency of each service single resource block is the ratio of the service latency of each service to the number of allocated physical resource blocks; each service SV k,i,j The latency of a single resource block is Dlpb. k,i,j =SDly k,i,j / ANm k,i,j (ms).

[0023] The throughput per service SV is the ratio of the throughput of each service to the number of allocated physical resource blocks. k,i,j The throughput of a single resource block is Thpb k,i,j =SThr k,i,j / ANm k,i,j (Mbps).

[0024] As a preferred embodiment, the services with long latency or low throughput when selecting a single resource block include:

[0025] Calculate the average latency per resource block and the average throughput per resource block for all services. The average latency per resource block for all services is the ratio of the sum of the service latencies of all services for all users at all base stations to the sum of the total number of services at all base stations. The average throughput per resource block for all services is the ratio of the sum of the throughput of all services for all users at all base stations to the total number of services at all base stations.

[0026] This solution specifically calculates the average latency per resource block for all services, DlPb = ∑ k ∑ i ∑ j Dlpb k,i,j / ∑ k (k v *k vw ), calculate the average throughput per resource block ThPb for all services = ∑ k ∑ i ∑ j Thpb k,i,j / ∑ k (k v *k vw ).

[0027] Set a marking threshold. If the latency of each service's single resource block is not less than the product of the average latency of all services' single resource blocks and the marking threshold, then the latency judgment flag for that service is set to 1; otherwise, it is marked to 0. If the throughput of each service's single resource block is not greater than the average throughput of all services' single resource blocks, then the throughput judgment flag for that service is set to 1; otherwise, it is marked to 0.

[0028] Specifically, setting the tag threshold LbV th ∈(0,1); for each business SV k,i,j If the condition Dlpb is met k,i,j ≥DlPb*LbV yh If the condition is met, the latency decision flag for this service is set to 1; otherwise, it is set to 0. k,i,j ≤ThPb*LbV th If the throughput decision flag is 1, then the throughput decision flag for that service is set to 1; otherwise, it is 0.

[0029] As a preferred embodiment, the step of marking services that are non-traffic-based and have long latency, marking services that are traffic-based and have low throughput, and selecting base stations to mark any service includes:

[0030] For each service, if the service type is non-traffic and the latency judgment flag is 1, then the service flag is set to 1; otherwise, it is set to 0. If the service type is traffic and the throughput judgment flag is 1, then the service flag is set to 1; otherwise, it is set to 0.

[0031] This solution specifically addresses each business SV. k,i,j If the service meets the service type STpe k,i,j = "Non-traffic type", and its latency decision flag is 1, then set the service flag of this service to 1; otherwise, set it to 0; if the service type STpe of this service is... k,i,j If the data type is "Traffic Class" and its throughput judgment flag is 1, then the business type of that service is marked as 1; otherwise, it is marked as 0.

[0032] Base stations whose service flag is 1 for any one of the services running on the base station are selected, and the selected base stations form the second set of shutdown target candidates.

[0033] Selected gNB k Constitutes the second set of turn-off target candidates gNB2 tclose .

[0034] As a preferred embodiment, the calculation of the weighted load for each base station includes:

[0035] Calculate the average load of all neighboring base stations for each base station; specifically, calculate the load of all its neighboring cells. Average load

[0036] Set load weights, and the weighted load of each base station is: load weight * base station load + (1 - load weight) * average load of neighboring base stations. Specifically, set the load weights... Calculate gNB for each base station k Weighted load Calculate the average weighted load of all base stations Filter out those that meet the conditions gNB base station k This constitutes the third set of candidate shutdown targets, gNB3. tclose .

[0037] As a preferred approach, a base station overload threshold is set, all target base stations are traversed, the load of all adjacent base stations of the target base station is calculated, and it is determined whether the load of all adjacent base stations exceeds the base station overload threshold. If so, the action of shutting down the target cell is canceled. If not, the base station with the lowest energy efficiency is selected from the adjacent base stations of the target base station, and it is determined whether the sum of the load of the target base station and the load of the base station with the lowest energy efficiency is greater than the base station overload threshold. If so, the action of shutting down the target cell is canceled. If not, the users of the target base station are migrated to the base station with the lowest energy efficiency, and the target base station is shut down.

[0038] The base station obtained by calculating the intersection of the three candidate sets of shutdown targets is the target base station gNB. t This constitutes the target shutdown base station set ST close =gNB1 tclose ∩gNB2 tclose ∩gNB3 tclose This scheme sets an overload threshold OvL for the residential area. thr Traverse all target base stations and calculate the target base station gNB. t The load of all adjacent base stations is checked to determine if the load of all adjacent base stations exceeds OvL. thr If yes, then cancel the action of shutting down the target cell this time; otherwise, remove it from the target base station gNB. t Neighboring base station list NCL t Select the base station with the lowest energy efficiency If the conditions are met The action to shut down the target cell will still be cancelled; otherwise, the target base station gNB will be shut down. t User migration to In the middle, the target base station gNB was completed. t The shutdown.

[0039] Therefore, the advantages of this invention are: by analyzing the energy efficiency of each base station, base stations with high energy efficiency ratios are selected; each service is analyzed in detail, and the performance of each service is calculated per unit particle occupancy, and then the services are marked accordingly; based on the load of each adjacent base station, the weighted load is calculated, and the overall average load of the base station cluster is obtained; under the premise of ensuring user satisfaction, 5G base stations with low-value services are shut down by migrating adjacent base stations, thereby optimizing the energy consumption of the 5G network system. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the process for generating the first set of shutdown target candidates in this invention;

[0041] Figure 2 This is a schematic diagram of the process for generating the second set of candidate shutdown targets in this invention;

[0042] Figure 3This is a schematic diagram of the process for generating the third set of candidate shutdown targets in this invention;

[0043] Figure 4 This is a schematic diagram of the process for intelligently shutting down the target base station in this invention;

[0044] Figure 5 This is a comparison chart of the power consumption of the present invention and other algorithms in a 5G base station with the RU in the off section;

[0045] Figure 6 This is a comparison chart of the throughput of this invention and other algorithms in a 5G base station with the RUs shut down;

[0046] Figure 7 This is a comparison chart of the average service latency of this invention and other algorithms in a 5G base station with a portion of the RUs turned off. Detailed Implementation

[0047] The technical solution of the present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0048] Example:

[0049] This embodiment presents a method for intelligent shutdown of 5G base stations based on neighboring cell energy efficiency, such as... Figures 1-4 As shown, the process includes:

[0050] Select u mutually neighboring base stations gNB tar ={gNB1,gNB2,…,gNB u}, obtain the corresponding load {cLD1,cLD2,…,cLD u}, power consumption {Pwr1,Pwr2,…,Pwr} u}, each base station gNB k ∈gNB tar running k v One user, US k,i It is the i-th user. Each user conducts k vw Each business, SV k,i,j It is the j-th business. Then user US k,i The business being run is SV k,i ={SV k,i,j j∈(1,2,…,k) vw )}.

[0051] For each base station gNB k Configure user US k,i The business types of each business carried out STpe k,i,j ∈{"non-traffic class","traffic class"}, business latency SDly k,i,j(ms), the number of allocated physical resource blocks ANm k,i,j and the corresponding throughput SThr k,i,j (Mbps), where k∈[1,u], i∈[1,k] v ],j∈[1,k vw ].

[0052] Screening of low-energy-efficiency base stations:

[0053] Calculate the energy efficiency of each base station, count the number of high-efficiency base stations among all its neighboring base stations, and select base stations whose proportion of high-efficiency neighboring base stations exceeds a threshold to generate the first set of candidate shutdown targets; the specific process includes,

[0054] (1-1): For gNB tar ={gNB1,gNB2,…,gNB u Each base station gNB in ​​} k Calculate the throughput of the base station SThr k =∑ i ∑ j SThr k,i,j Calculate the energy efficiency of the base station

[0055] (1-2): Set the energy efficiency threshold EfV th Obtain each base station gNB k All adjacent base stations, Its number of adjacent base stations is n k For NCL k Each base station gNB z Statistically identify SThr that meets the conditions z ≥SThr k +EfV th Number of adjacent base stations k z Calculate the proportion of high-energy-efficiency adjacent base stations Rt k =k z / n k .

[0056] (1-3): Set the proportional threshold RtV th ∈(0,1); filter out Rt that satisfy the condition k >RtV th Each base station gNB k The selected base stations constitute the first set of candidate shutdown targets, gNB1. tclose .

[0057] Screening of low-quality service base stations:

[0058] For each base station, calculate the latency and throughput of each service per resource block. Select services with high latency or low throughput per resource block. Mark services that are not traffic-based and have high latency, and mark services that are traffic-based and have low throughput. Select base stations for any service that are marked, and generate a second set of candidate shutdown targets. The specific process includes...

[0059] (2-1): For gNB tar ={gNB1,gNB2,…,gNB u Each base station gNB in ​​} k Calculate the SV of each business k,i,j Single resource block latency Dlpb k,i,j =SDly k,i,j / ANm k,i,j (ms), calculate each business SV k,i,j single resource block throughput (Thp) k,i,j =SThr k,i,j / ANm k,i,j (Mbps);

[0060] (2-2): Calculate the average single resource block latency of all services, i.e., the single-particle latency of all services, DlPb = ∑ k ∑ i ∑ j Dlpb k,i,j / ∑ k (k v *k vw ), calculate the single resource block throughput of all services, i.e., the single particle throughput of all services ThPb = ∑ k ∑ i ∑ j Thpb k,i,j / ∑ k (k v *k vw );

[0061] (2-3): Set the flag threshold LbV th ∈(0,1); for each business SV k,i,j If the condition Dlpb is met k,i,j ≥DlPb*LbV th If the condition is met, the latency decision flag for this service is set to 1; otherwise, it is set to 0. k,i,j ≤ThPb*LbV th If the throughput decision flag is 1, then the throughput decision flag for that service is set to 1; otherwise, it is 0.

[0062] (2-4): For each business SV k,i,j If the service meets the service type STpe k,i,j= "Non-traffic type", and its latency decision flag is 1, then set the service flag of this service to 1; otherwise, set it to 0; if the service type STpe of this service is... k,i,j If the data type is "Traffic Class" and its throughput judgment flag is 1, then the business type of that service is marked as 1; otherwise, it is marked as 0.

[0063] (2-5): For gNB tar ={gNB1,gNB2,…,gNB u Each base station gNB in ​​} k Filter out base stations whose service flag is 1 for any one of the services running on the base station, and filter out the gNBs. k Constitutes the second set of turn-off target candidates gNB2 tclose .

[0064] Screening of low-weighted load base stations:

[0065] The weighted load of each base station is calculated based on the average load of all its neighboring base stations. Base stations with a weighted load less than the average weighted load of all base stations are selected to generate a third set of candidate base station shutdown targets. Specifically, this includes (3-1): for gNB tar ={gNB1,gNB2,…,gNB u Each base station gNB in ​​} k Calculate all its neighboring cells Average load

[0066] (3-2): Set load weight Calculate gNB for each base station k Weighted load Calculate the average weighted load of all base stations Filter out those that meet the conditions gNB base station k This constitutes the third set of candidate shutdown targets, gNB3. tclose .

[0067] The base station at the intersection of three candidate shutdown target sets is identified as the target base station. Whether to shut down the target base station is determined by comparing it to a cell overload threshold. Specifically, this includes...

[0068] (4-1): Calculate the intersection of the three candidate sets of shutdown targets, and the obtained base station is the target base station gNB. t This constitutes the target shutdown base station set ST close =gNB1 tclose ∩gNB2 tclose ∩gNB3 tclose .

[0069] (4-2): Set the cell overload threshold OvL thr Traverse all target base stations and calculate the target base station gNB. t The load of all adjacent base stations is checked to determine if the load of all adjacent base stations exceeds OvL. thr If yes, then cancel the action of shutting down the target cell this time; otherwise, remove it from the target base station gNB. t Neighboring base station list NCL t Select the base station with the lowest energy efficiency If the conditions are met The action to shut down the target cell will still be cancelled; otherwise, the target base station gNB will be shut down. t User migration to In the middle, the target base station gNB was completed. t The shutdown.

[0070] The following description uses u=6 as an example to illustrate this embodiment.

[0071] The base station users and their services are shown in Table 1:

[0072] Table 1 Base Station Users and Services

[0073]

[0074] The basic data is shown in Table 2:

[0075] Table 2 Basic Data

[0076]

[0077] Screening of low-energy-efficiency base stations:

[0078] (1-1): For gNB tar Each base station gNB in ​​{gNB1, gNB2, ..., gNB6} k Calculate the throughput of the base station

[0079] SThr k =∑ i ∑ j SThr k,i,j ={50,250,155,96,32,254}(Mbps),

[0080] Calculate the energy efficiency of the base station

[0081]

[0082] (1-2): Set the energy efficiency threshold EfV th Obtain each base station gNB k All adjacent base stations, Its number of adjacent base stations is n k For NCL k Each base station gNB z Statistically identify SThr that meets the conditions z ≥SThr k +EfVt th Number of adjacent base stations k z Calculate the proportion of high-energy-efficiency adjacent base stations

[0083]

[0084] (1-3): Set the proportional threshold RtV th ∈(0,1); filter out Rt that satisfy the condition k >RtV th Each base station gNB k The selected base stations constitute the first set of candidate shutdown targets, gNB1. tcl0se gNB1 tclose ={gNB1,gNB4,gNB5}.

[0085] Screening of low-quality service base stations:

[0086] (2-1): For gNB tar ={gNB1,gNB2,…,gNB u Each base station gNB in ​​} k Calculate the SV of each business k,i,j Single resource block latency

[0087]

[0088] Calculate each business SV k,i,j single resource block throughput

[0089]

[0090] (2-2): Calculate the average single resource block latency of all services, i.e., the single-particle latency of all services, DlPb = ∑ k ∑ i ∑ j Dlpb k,i,j / ∑ k (k v *k vw =1.98ms, calculate the single resource block throughput of all services, i.e., the single particle throughput of all services ThPb = ∑ k ∑ i ∑ j Thpb k,i,j / ∑ k (l v *k vw2.39Mbps;

[0091] (2-3): Set the flag threshold LbV th ∈(0,1); for each business SV k,i,j If the condition Dlpb is met k,i,j ≥DlPb*LbV th If the delay decision flag for this service is 1, then it is 0; otherwise, it is 0. If the condition is met, Thrpb... k,i,j ≤ThPb*LbV th If the throughput decision flag is 1, then the throughput decision flag for that service is set to 1; otherwise, it is 0.

[0092] (2-4): For each business SV k,i,j If the service meets the service type STpe k,i,j = "Non-traffic type", and its latency decision flag is 1, then set the service flag of this service to 1; otherwise, set it to 0; if the service type STpe of this service is... k,i,j If a service is classified as "Traffic Class" and its throughput judgment flag is 1, then the service flag is set to 1; otherwise, it is set to 0. The final service flag result for all services is {1,{0,1,1},{1,0},1,0,{1,1,1}}.

[0093] (2-5): For gNB tar Each base station gNB in ​​{gNB1, gNB2, ..., gNB6} k Filter out base stations whose service flag is 1 for any one of the services running on the base station, and filter out the gNBs. k Constitutes the second set of turn-off target candidates gNB2 tclose gNB2 tclose ={gNB1,gNB2,gNB3,gNB4,gNB6}.

[0094] Screening of low-weighted load base stations:

[0095] (3-1): For gNB tar Each base station gNB in ​​{gNB1, gNB2, ..., gNB6} k Calculate all its neighboring cells Average load

[0096] (3-2): Set load weight Calculate gNB for each base station k Weighted load

[0097] (3-3): Calculate the average weighted load of all base stations. Filter out those that meet the conditions gNB base station k This constitutes the third set of candidate shutdown targets, gNB3. tclose gNB3 tclose ={gNB1,gNB2,gNB3}.

[0098] The base station at the intersection of three candidate shutdown target sets is identified as the target base station. Whether to shut down the target base station is determined by comparing it to a cell overload threshold. Specifically, this includes...

[0099] (4-1): Calculate the intersection of the three candidate sets of shutdown targets, and the obtained base station is the target base station gNB. t This constitutes the target shutdown base station set ST close =gNB1 tclose ∩gNB2 tclose ∩gNB3 tclose ={gNB1}.

[0100] (4-2): For ST close Base station gNB1 in the middle does not satisfy the condition that the load of all adjacent base stations exceeds OvL. thr =0.7. Select the base station with the lowest energy efficiency from the neighboring base station list NCL1 of gNB1. The conditions are not met. Therefore, users of gNB1 are migrated to gNB5, thus shutting down the gNB1 base station.

[0101] Simulation experiment:

[0102] The method of this invention (EENIC for short) was simulated on a MATLAB platform. A base station cluster consisting of six 5G base stations was used to mutually configure neighboring cells, and a certain load was set for each. This allowed some RUs (Radio Units) or DUs (Distributed Units) of the lightly loaded base stations to be intelligently shut down. The power consumption results are shown in [reference needed]. Figures 5 to 7 As shown.

[0103] like Figure 5 As shown, in a single 5G base station simulation, the power consumption of the base station decreases significantly due to the shutdown of some RUs. However, under the premise of ensuring the stable operation of existing services, the base station power consumption remains within a relatively stable range. In single-site power consumption simulation, the algorithm performance of EENIC and GSIC is similar.

[0104] like Figure 6As shown, when simulating cell throughput, EENIC will prioritize eliminating low-throughput services, especially traffic-related services. Therefore, after shutdown, only base stations with good environmental conditions and high throughput will remain, thus showing better base station performance than GSIC in the simulation, and also achieving the goal of reducing power consumption and improving performance.

[0105] like Figure 7 As shown, when simulating service latency, EENIC will prioritize eliminating services with high average latency, especially non-traffic services. Therefore, after shutdown, only base stations with good environmental conditions and low latency will remain, thus exhibiting better base station performance than GSIC in the simulation, and also achieving the goal of reducing power consumption and latency.

[0106] The specific embodiments described herein are merely illustrative of the spirit of the invention. Those skilled in the art to which this invention pertains may make various modifications or additions to the described specific embodiments or use similar methods to substitute them, without departing from the spirit of the invention or exceeding the scope defined by the appended claims.

Claims

1. A method for intelligent shutdown of 5G base stations based on neighboring cell energy efficiency, characterized in that: include: Calculate the energy efficiency of each base station, count the number of high-efficiency base stations among all its neighboring base stations, filter out base stations whose high-efficiency ratio among neighboring base stations is greater than a threshold, and generate the first set of candidates for shutdown targets. For each base station, calculate the single resource block latency and single resource block throughput for each service. The single resource block latency for each service is the ratio of the service latency of each service to the number of allocated physical resource blocks. The single resource block throughput for each service is the ratio of the throughput of each service to the number of allocated physical resource blocks. Select services with high single resource block latency or low throughput. Mark services that are non-traffic-type services with high latency and services that are traffic-type services with low throughput. Filter out base stations where any service is marked and generate a second set of candidate shutdown targets. The weighted load of each base station is calculated based on the average load of all its neighboring base stations. Base stations with a weighted load less than the average weighted load of all base stations are selected to generate a third set of candidates for shutdown targets. The base station at the intersection of three candidate shutdown target sets is the target base station. The decision on whether to shut down the target base station is based on a comparison with the cell overload threshold.

2. The intelligent shutdown method for 5G base stations based on neighboring cell energy efficiency according to claim 1, characterized in that: Select neighboring base stations, obtain their corresponding load and power consumption information, run multiple users on each base station, and each user carries out multiple services. Set the service type, service latency, number of allocated physical resource blocks, and corresponding throughput information for each service carried out by the user.

3. The intelligent shutdown method for 5G base stations based on neighboring cell energy efficiency according to claim 1, characterized in that: The process of selecting base stations whose high-efficiency ratio among adjacent base stations is greater than a certain threshold includes: The high-efficiency ratio of adjacent base stations is obtained by comparing the number of high-efficiency base stations among all adjacent base stations of each base station to the number of all adjacent base stations of the current base station. Set a ratio threshold, select base stations that meet the requirement that the proportion of high energy efficiency among adjacent base stations is greater than the ratio threshold, and form the first set of candidate base stations for shutdown targets.

4. The intelligent shutdown method for 5G base stations based on neighboring cell energy efficiency according to claim 1, characterized in that: The services with prolonged or low throughput when selecting a single resource block include: Calculate the average latency per resource block and the average throughput per resource block for all services. The average latency per resource block for all services is the ratio of the sum of the service latencies of all services for all users at all base stations to the sum of the total number of services at all base stations. The average throughput per resource block for all services is the ratio of the sum of the throughput of all services for all users at all base stations to the sum of the total number of services at all base stations. Set a marking threshold. If the latency of each service's single resource block is not less than the product of the average latency of all services' single resource blocks and the marking threshold, then the latency judgment flag for that service is set to 1; otherwise, it is marked to 0. If the throughput of each service's single resource block is not greater than the average throughput of all services' single resource blocks, then the throughput judgment flag for that service is set to 1; otherwise, it is marked to 0.

5. A method for intelligent shutdown of 5G base stations based on neighboring cell energy efficiency according to claim 4, characterized in that: The process of marking services that are not traffic-based and have long latency, marking services that are traffic-based and have low throughput, and selecting base stations that are marked for any given service includes: For each service, if the service type is non-traffic and the latency judgment flag is 1, then the service flag is set to 1; otherwise, it is set to 0. If the service type is traffic and the throughput judgment flag is 1, then the service flag is set to 1; otherwise, it is set to 0. Base stations whose service flag is 1 for any one of the services running on the base station are selected, and the selected base stations form the second set of shutdown target candidates.

6. A method for intelligent shutdown of 5G base stations based on neighboring cell energy efficiency according to claim 1 or 2, characterized in that: The calculation of the weighted load for each base station includes: Calculate the average load of all neighboring base stations for each base station; Set load weights, and the weighted load of each base station is: load weight * base station load + (1 - load weight) * average load of adjacent base stations.

7. A method for intelligent shutdown of 5G base stations based on neighboring cell energy efficiency according to claim 1 or 2, characterized in that: The determination of whether to shut down the target base station includes: Set a cell overload threshold, iterate through all target base stations, calculate the load of all adjacent base stations of the target base station, and determine whether the load of all adjacent base stations exceeds the cell overload threshold. If so, cancel the current action of shutting down the target cell. If not, select the base station with the lowest energy efficiency from the adjacent base stations of the target base station, and determine whether the sum of the load of the target base station and the load of the base station with the lowest energy efficiency is greater than the cell overload threshold. If so, cancel the current action of shutting down the target cell. If not, migrate the users of the target base station to the base station with the lowest energy efficiency and shut down the target base station.

Images