A kind of switching method, base station and user terminalTechnical field
The present invention relates to field of communication technology, a kind of switching method, base station and user terminal are particularly related to.
Background technique
In LTE (Long Term Evolution, long term evolution) system, when user is in connection status, network passes throughHandoff procedure realizes the mobile management to UE (User Equipment, user equipment).The change of UE channel condition mayTriggering switching, network may also initiate to switch due to the purpose of load balancing.Three are adjudicated and executed to switching specifically comprising measurementProcess.Handover measurement occupies an important position in handoff algorithms, and the measurement report of UE is to eNB (Evolved Node B, base station)Handover decisions have key effect.
According to measurement configuration information, UE can periodically or event triggering ground uploads measurement report.7 are defined in LTEKind event triggering criterion.Wherein A1-A5 event is used for the survey between heterogeneous system for the measurement in LTE system, B1, B2 eventAmount.With the trigger condition for switching closely related A3 event are as follows: service quality (RSRP (the Reference Signal of adjacent areaReceiving Power, Reference Signal Received Power)/RSRQ (Reference Signal Receiving Quality, ginsengExamine signal receiving quality)) higher than a serving cell absolute thresholding, and keep at least TTT (Time to Trigger triggers thingPart) duration.It is shown below:
Mn+Ofn+Ocn-Hys > Ms+Ofs+Ocs+Off
Wherein Mn is the measurement result of adjacent area；Ofn is adjacent area biasing relevant to carrier frequency；Ocn is that adjacent area is specificBiasing；Ms is the measurement result of serving cell；Ofs is serving cell biasing relevant to carrier frequency；Ocs is serving cell spyFixed biasing；Hys is sluggish；Off is the biasing for triggering the event.Above-mentioned expression formula can rewrite are as follows:
Mn-Ms > Hys+Off-CIO
Wherein CIO=Ocn-Ocs is the parameter between cell pair, and Hys and off are the specific parameters of serving cell.In same frequencyDeployment, there is Ofn=Ofs.The effect of TTT is in order to avoid ping-pong caused by the fluctuation as signal, so as to cause need notThe switching and signaling overheads wanted, value set have been standardized.Hys, CIO (Cell Individual Offset, cellOffset), off be important handoff parameter, if parameter setting is unreasonable, system performance can be seriously affected, deteriorate the clothes of userBusiness experience.
Fig. 1 is the handover trigger criterion schematic diagram of LTE, as shown in Figure 1, within the TTT time, the RSRP/ of adjacent cellWhen RSRQ is all larger than source cell RSRP/RSRQ some offset Hys, measurement report condition meets, and UE gives reporting measurement reportsBase station, the handover decisions of assistant base station.
LTE mainly realizes ambulant optimization by the optimal setting to above-mentioned parameter at present, but due to smallThe scrambling of area's covering and the complexity of actual channel fading environment, be difficult in some scenarios by adjusting TTT andHys parameter realizes preferably performance of handoffs.In scene as shown in Figure 2, wherein 1 is skyscraper, 2 be low-rise building, is usedFamily 3 is the user to turn round, and user 4 is the user of straight trip, is to avoid pingpang handoff for user 3, biggish TTT value should be arranged,But if TTT setting is larger, it may result in the increase of cutting off rate for user 4；If lesser TTT value is arranged, user 3Pingpang handoff can then occur.Therefore traditional handover trigger criterion cannot may provide good user in some scenarios for userExperience.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of switching method, base station and user terminals, can be avoided userPingpang handoff situation, and reduce the cutting off rate in handoff procedure.
In order to solve the above technical problems, the embodiment of the present invention offer technical solution is as follows:
On the one hand, a kind of switching method is provided, base station is applied to, the switching method includes:
Handover decisions model is obtained, the handover decisions model is previously according to history metrical information, history switching resultAnd the training of user terminal performance of handoffs obtains；
RSRP time series, RSRQ time series and the user terminal of user terminal measurement are obtained from being currently accessed base stationIt is switched to the conditional probability of adjacent base station；
The RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions model,Obtain the handover-target base station of the user terminal.
Further, the acquisition handover decisions model includes:
The input data of the training handover decisions model is obtained, the input data includes at least: user terminal is gone throughFor user terminal from the conditional probability for being currently accessed base station location and being switched to adjacent base station, historical user is whole under the conditions of history motion trackHold the RSRP time series of access base station and its adjacent base station, historical user's terminal to access base station and its adjacent base stationRSRQ time series；
Using the input data as input, using history switching result and historical user's terminal performance of handoffs as constraining itemPart, training obtain the handover decisions model.
Further, the input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further, obtaining the conditional probability includes:
Receive the user terminal path locus habit and user terminal history switching record that core net is sent；
It is general that the condition is calculated according to user terminal path locus habit and user terminal history switching recordRate.
Further, it is described will the RSRP time series, RSRQ time series and the conditional probability input described inBefore handover decisions model, the method also includes:
Judge whether to enter as defined in LTE system according to the RSRP time series of user terminal measurement and RSRQ time seriesMeasurement event；
It is described that the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions mouldType includes:
Entered according to the RSRP time series of user terminal measurement and the judgement of RSRQ time series and is measured as defined in LTE systemWhen event, the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions model, obtainedTo the handover-target base station of the user terminal.
Further, after the handover-target base station for obtaining the user terminal, the method also includes:
It is not the user terminal when being currently accessed base station in the handover-target base station, switching request is sent to instituteState handover-target base station.
The embodiment of the invention also provides a kind of switching methods, are applied to user terminal, and the switching method includes:
Handover decisions model is obtained, the handover decisions model is previously according to history metrical information, history switching resultAnd the training of user terminal performance of handoffs obtains；
RSRP time series, RSRQ time series and the user terminal of user terminal measurement are obtained from being currently accessed base stationIt is switched to the conditional probability of adjacent base station；
The RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions model,Obtain the handover-target base station of the user terminal.
Further, the acquisition handover decisions model includes:
The handover decisions model that the user terminal is currently accessed base station transmission is received, the handover decisions model isIt is instructed using input data as input using history switching result and historical user's terminal performance of handoffs as constraint condition the base stationGet, the input data includes at least: user terminal is from being currently accessed base under the conditions of the historical movement path of user terminalStation location is switched to the conditional probability of adjacent base station, the RSRP time sequence of historical user's terminal to access base station and its adjacent base stationColumn, the RSRQ time series of historical user's terminal to access base station and its adjacent base station.
Further, the input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further, the conditional probability be the base station receive core net send user terminal path locus habit andAfter user terminal history switching record, calculated according to user terminal path locus habit and user terminal history switching recordIt obtains.
Further, it is described will the RSRP time series, RSRQ time series and the conditional probability input described inBefore handover decisions model, the method also includes:
Judge whether to enter as defined in LTE system according to the RSRP time series of measurement and RSRQ time series and measures thingPart；
It is described that the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions mouldType includes:
When entering measurement event as defined in LTE system according to the RSRP time series of measurement and the judgement of RSRQ time series,The RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions model, obtained describedThe handover-target base station of user terminal.
Further, after the handover-target base station for obtaining the user terminal, the method also includes:
It is not the user terminal when being currently accessed base station in the handover-target base station, will will include the switching meshThe handover-target base station list of mark Base Station Identification, which is sent to, is currently accessed base station, determines that the user is whole to be currently accessed base stationThe target BS of switching is held, and switching request is sent to the target BS.
The embodiment of the invention also provides a kind of base stations, including processor and transceiver:
The processor is to measure to believe previously according to history for obtaining handover decisions model, the handover decisions modelBreath, history switching result and the training of user terminal performance of handoffs obtain；Obtain RSRP time series, the RSRQ of user terminal measurementTime series and user terminal are from the conditional probability for being currently accessed base station and being switched to adjacent base station；By the RSRP time sequenceColumn, RSRQ time series and the conditional probability input the handover decisions model, obtain the switching mesh of the user terminalMark base station.
Further,
The processor is specifically used for obtaining the input data of the training handover decisions model, and the input data is at leastInclude: user terminal historical movement path under the conditions of user terminal from the item for being currently accessed base station location and being switched to adjacent base stationThe RSRP time series of part probability, historical user's terminal to access base station and its adjacent base station, historical user's terminal to access baseIt stands and its RSRQ time series of adjacent base station；Using the input data as input, with history switching result and historical userTerminal performance of handoffs obtains the handover decisions model as constraint condition, training.
Further,
The input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further,
The transceiver is used to receive the user terminal path locus habit and the switching of user terminal history of core net transmissionRecord；
The processor is specifically used for according to user terminal path locus habit and user terminal history switching recordThe conditional probability is calculated.
Further,
The processor is also used to be judged whether according to the RSRP time series and RSRQ time series of user terminal measurementInto measurement event as defined in LTE system, judge according to the RSRP time series and RSRQ time series of user terminal measurementIt is into as defined in LTE system when measurement event, the RSRP time series, RSRQ time series and the conditional probability is defeatedEnter the handover decisions model, obtains the handover-target base station of the user terminal.
Further,
The transceiver is also used in the handover-target base station not be the user terminal when being currently accessed base station, willSwitching request is sent to the handover-target base station.
The embodiment of the invention also provides a kind of user terminal, including processor and transceiver,
The processor is to measure to believe previously according to history for obtaining handover decisions model, the handover decisions modelBreath, history switching result and the training of user terminal performance of handoffs obtain, and obtain RSRP time series, the RSRQ of user terminal measurementTime series and user terminal are from the conditional probability for being currently accessed base station and being switched to adjacent base station, by the RSRP time sequenceColumn, RSRQ time series and the conditional probability input the handover decisions model, obtain the switching mesh of the user terminalMark base station.
Further,
The transceiver is used to receive the handover decisions model that the user terminal is currently accessed base station transmission, describedHandover decisions model is the base station using input data as input, with history switching result and historical user's terminal performance of handoffsObtain as constraint condition training, the input data includes at least: user is whole under the conditions of the historical movement path of user terminalIt holds from the conditional probability for being currently accessed base station location and being switched to adjacent base station, historical user's terminal to access base station and its adjacent baseThe RSRQ time series of the RSRP time series stood, historical user's terminal to access base station and its adjacent base station.
Further,
The input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further,
The conditional probability is the user terminal path locus habit and user terminal that the base station receives that core net is sentAfter history switching record, it is calculated according to user terminal path locus habit and user terminal history switching record.
Further,
The processor is also used to be judged whether according to the RSRP time series and RSRQ time series of measurement to enter LTE systemMeasurement event as defined in uniting enters as defined in LTE system according to the RSRP time series of measurement and the judgement of RSRQ time seriesWhen measurement event, the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions mouldType obtains the handover-target base station of the user terminal.
Further,
The transceiver is also used in the handover-target base station not be the user terminal when being currently accessed base station, willIt include that the handover-target base station list of handover-target base station mark is sent to and is currently accessed base station, to be currently accessed baseIt stands and determines the target BS of the user terminal switching, and switching request is sent to the target BS.
The embodiment of the invention also provides a kind of base station, including memory, processor and it is stored on the memory simultaneouslyThe computer program that can be run on the processor；The processor realizes switching side as described above when executing described programMethod.
The embodiment of the invention also provides a kind of user terminal, including memory, processor and it is stored in the memoryComputer program that is upper and can running on the processor；It realizes when the processor executes described program and cuts as described aboveChange method.
The embodiment of the invention also provides a kind of computer readable storage mediums, are stored thereon with computer program, the journeyThe step in switching method as described above is realized when sequence is executed by processor.
The embodiment of the present invention has the advantages that
In above scheme, using history metrical information such as RSRP, RSRQ sequence etc. and history switching result and user are wholeEnd performance of handoffs is trained handover decisions model, is then based on real-time metrical information, makes handover decisions, determines userThe handover-target base station of terminal.The present embodiment can effectively be learnt by historical data training to specific MPS process, letterRoad fading environment and the track of user etc. make more preferably handover decisions, avoid user's pingpang handoff situation in traditional algorithm,And reduce the cutting off rate in handoff procedure.
Detailed description of the invention
Fig. 1 is the handover trigger criterion schematic diagram of LTE；
Fig. 2 is a concrete scene schematic diagram；
Fig. 3 is the flow diagram for the switching method that the embodiment of the present invention is applied to base station；
Fig. 4 is the flow diagram for the switching method that the embodiment of the present invention is applied to user terminal；
Fig. 5 is the structural schematic diagram of base station of the embodiment of the present invention；
Fig. 6 is the structural schematic diagram of user terminal of the embodiment of the present invention；
Fig. 7 is the flow diagram for the switching method that the specific embodiment of the invention is applied to base station；
Fig. 8 is the flow diagram for the switching method that the specific embodiment of the invention is applied to user terminal；
Fig. 9 is the schematic diagram of training handover decisions model of the embodiment of the present invention.
Specific embodiment
To keep the embodiment of the present invention technical problems to be solved, technical solution and advantage clearer, below in conjunction withDrawings and the specific embodiments are described in detail.
The embodiment of the present invention is directed to traditional handover trigger criterion and may cannot provide in some scenarios for user very wellUser experience the problem of, a kind of switching method, base station and user terminal are provided, can be avoided user's pingpang handoff situation, andReduce the cutting off rate in handoff procedure.
The embodiment of the present invention provides a kind of switching method, is applied to base station, as shown in figure 3, the switching method includes:
Step 101: obtaining handover decisions model, the handover decisions model is previously according to history metrical information, historySwitching result and the training of user terminal performance of handoffs obtain；
Step 102: obtaining the RSRP time series, RSRQ time series and user terminal of user terminal measurement from currentAccess base station is switched to the conditional probability of adjacent base station；
Step 103: the RSRP time series, RSRQ time series and the conditional probability input switching are determinedPlan model obtains the handover-target base station of the user terminal.
In the present embodiment, using history metrical information such as RSRP, RSRQ sequence etc. and history switching result and user are wholeEnd performance of handoffs is trained handover decisions model, is then based on real-time metrical information, makes handover decisions, determines userThe handover-target base station of terminal.The present embodiment can effectively be learnt by historical data training to specific MPS process, letterRoad fading environment and the track of user etc. make more preferably handover decisions, avoid user's pingpang handoff situation in traditional algorithm,And reduce the cutting off rate in handoff procedure.
Further, the acquisition handover decisions model includes:
The input data of the training handover decisions model is obtained, the input data includes at least: user terminal is gone throughFor user terminal from the conditional probability for being currently accessed base station location and being switched to adjacent base station, historical user is whole under the conditions of history motion trackHold the RSRP time series of access base station and its adjacent base station, historical user's terminal to access base station and its adjacent base stationRSRQ time series；
Using the input data as input, using history switching result and historical user's terminal performance of handoffs as constraining itemPart, training obtain the handover decisions model.
Further, the input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further, obtaining the conditional probability includes:
Receive the user terminal path locus habit and user terminal history switching record that core net is sent；
It is general that the condition is calculated according to user terminal path locus habit and user terminal history switching recordRate.
Further, it is described will the RSRP time series, RSRQ time series and the conditional probability input described inBefore handover decisions model, the method also includes:
Judge whether to enter as defined in LTE system according to the RSRP time series of user terminal measurement and RSRQ time seriesMeasurement event；
It is described that the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions mouldType includes:
Entered according to the RSRP time series of user terminal measurement and the judgement of RSRQ time series and is measured as defined in LTE systemWhen event, the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions model, obtainedTo the handover-target base station of the user terminal.
Further, after the handover-target base station for obtaining the user terminal, the method also includes:
It is not the user terminal when being currently accessed base station in the handover-target base station, switching request is sent to instituteState handover-target base station.
The embodiment of the invention also provides a kind of switching methods, are applied to user terminal, as shown in figure 4, the switching sideMethod includes:
Step 201: obtaining handover decisions model, the handover decisions model is previously according to history metrical information, historySwitching result and the training of user terminal performance of handoffs obtain；
Step 202: obtaining the RSRP time series, RSRQ time series and user terminal of user terminal measurement from currentAccess base station is switched to the conditional probability of adjacent base station；
Step 203: the RSRP time series, RSRQ time series and the conditional probability input switching are determinedPlan model obtains the handover-target base station of the user terminal.
In the present embodiment, using history metrical information such as RSRP, RSRQ sequence etc. and history switching result and user are wholeEnd performance of handoffs is trained handover decisions model, is then based on real-time metrical information, makes handover decisions, determines userThe handover-target base station of terminal.The present embodiment can effectively be learnt by historical data training to specific MPS process, letterRoad fading environment and the track of user etc. make more preferably handover decisions, avoid user's pingpang handoff situation in traditional algorithm,And reduce the cutting off rate in handoff procedure.
Further, the acquisition handover decisions model includes:
The handover decisions model that the user terminal is currently accessed base station transmission is received, the handover decisions model isIt is instructed using input data as input using history switching result and historical user's terminal performance of handoffs as constraint condition the base stationGet, the input data includes at least: user terminal is from being currently accessed base under the conditions of the historical movement path of user terminalStation location is switched to the conditional probability of adjacent base station, the RSRP time sequence of historical user's terminal to access base station and its adjacent base stationColumn, the RSRQ time series of historical user's terminal to access base station and its adjacent base station.
Further, the input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further, the conditional probability be the base station receive core net send user terminal path locus habit andAfter user terminal history switching record, calculated according to user terminal path locus habit and user terminal history switching recordIt obtains.
Further, it is described will the RSRP time series, RSRQ time series and the conditional probability input described inBefore handover decisions model, the method also includes:
Judge whether to enter as defined in LTE system according to the RSRP time series of measurement and RSRQ time series and measures thingPart；
It is described that the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions mouldType includes:
When entering measurement event as defined in LTE system according to the RSRP time series of measurement and the judgement of RSRQ time series,The RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisions model, obtained describedThe handover-target base station of user terminal.
Further, after the handover-target base station for obtaining the user terminal, the method also includes:
It is not the user terminal when being currently accessed base station in the handover-target base station, will will include the switching meshThe handover-target base station list of mark Base Station Identification, which is sent to, is currently accessed base station, determines that the user is whole to be currently accessed base stationThe target BS of switching is held, and switching request is sent to the target BS.
The embodiment of the invention also provides a kind of base stations, as shown in Figure 5, comprising: processor 31 and transceiver 32:
The processor 31 is to measure to believe previously according to history for obtaining handover decisions model, the handover decisions modelBreath, history switching result and the training of user terminal performance of handoffs obtain；Obtain RSRP time series, the RSRQ of user terminal measurementTime series and user terminal are from the conditional probability for being currently accessed base station and being switched to adjacent base station；By the RSRP time sequenceColumn, RSRQ time series and the conditional probability input the handover decisions model, obtain the switching mesh of the user terminalMark base station.
In the present embodiment, using history metrical information such as RSRP, RSRQ sequence etc. and history switching result and user are wholeEnd performance of handoffs is trained handover decisions model, is then based on real-time metrical information, makes handover decisions, determines userThe handover-target base station of terminal.The present embodiment can effectively be learnt by historical data training to specific MPS process, letterRoad fading environment and the track of user etc. make more preferably handover decisions, avoid user's pingpang handoff situation in traditional algorithm,And reduce the cutting off rate in handoff procedure.
Further,
The processor 31 is specifically used for obtaining the input data of the training handover decisions model, and the input data is extremelyInclude: less user terminal historical movement path under the conditions of user terminal from base station location is currently accessed be switched to adjacent base stationThe RSRP time series of conditional probability, historical user's terminal to access base station and its adjacent base station, historical user's terminal to accessThe RSRQ time series of base station and its adjacent base station；Using the input data as input, used with history switching result and historyFamily terminal performance of handoffs obtains the handover decisions model as constraint condition, training.
Further,
The input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further,
The transceiver 32 is used to receive the user terminal path locus habit of core net transmission and user terminal history is cutChange record；
The processor 31 is specifically used for according to user terminal path locus habit and user terminal history switching noteThe conditional probability is calculated in record.
Further,
The processor 31 is also used to be judged according to the RSRP time series and RSRQ time series of user terminal measurementIt is no to enter measurement event as defined in LTE system, sentence according to the RSRP time series and RSRQ time series of user terminal measurementIt is disconnected when entering measurement event as defined in LTE system, by the RSRP time series, RSRQ time series and the conditional probabilityThe handover decisions model is inputted, the handover-target base station of the user terminal is obtained.
Further,
The transceiver 32 is also used in the handover-target base station not be the user terminal when being currently accessed base station,Switching request is sent to the handover-target base station.
The embodiment of the invention also provides a kind of user terminals, as shown in fig. 6, including processor 41 and transceiver 42,
The processor 41 is to measure to believe previously according to history for obtaining handover decisions model, the handover decisions modelBreath, history switching result and the training of user terminal performance of handoffs obtain, and obtain RSRP time series, the RSRQ of user terminal measurementTime series and user terminal are from the conditional probability for being currently accessed base station and being switched to adjacent base station, by the RSRP time sequenceColumn, RSRQ time series and the conditional probability input the handover decisions model, obtain the switching mesh of the user terminalMark base station.
In the present embodiment, using history metrical information such as RSRP, RSRQ sequence etc. and history switching result and user are wholeEnd performance of handoffs is trained handover decisions model, is then based on real-time metrical information, makes handover decisions, determines userThe handover-target base station of terminal.The present embodiment can effectively be learnt by historical data training to specific MPS process, letterRoad fading environment and the track of user etc. make more preferably handover decisions, avoid user's pingpang handoff situation in traditional algorithm,And reduce the cutting off rate in handoff procedure.
Further,
The transceiver 42 is used to receive the handover decisions model that the user terminal is currently accessed base station transmission, instituteStating handover decisions model is the base station using input data as input, with history switching result and historical user's terminal switchingIt can be obtained as constraint condition training, the input data includes at least: user under the conditions of the historical movement path of user terminalTerminal is from the conditional probability for being currently accessed base station location and being switched to adjacent base station, historical user's terminal to access base station and its adjacentThe RSRQ time series of the RSRP time series of base station, historical user's terminal to access base station and its adjacent base station.
Further,
The input data further includes following at least one:
Teleservice QoS requirement；
Network resource conditions；
Load of base station.
Further,
The conditional probability is the user terminal path locus habit and user terminal that the base station receives that core net is sentAfter history switching record, it is calculated according to user terminal path locus habit and user terminal history switching record.
Further,
The processor 41 is also used to be judged whether according to the RSRP time series and RSRQ time series of measurement to enter LTEMeasurement event as defined in system is provided entering LTE system according to the judgement of the RSRP time series and RSRQ time series of measurementMeasurement event when, the RSRP time series, RSRQ time series and the conditional probability are inputted into the handover decisionsModel obtains the handover-target base station of the user terminal.
Further,
The transceiver 42 is also used in the handover-target base station not be the user terminal when being currently accessed base station,It will include that the handover-target base station list of handover-target base station mark is sent to and is currently accessed base station, to be currently accessedBase station determines the target BS of the user terminal switching, and switching request is sent to the target BS.
It describes in detail below with reference to specific embodiment to switching method of the invention:
Embodiment one
The switching method of the present embodiment is applied to base station, as shown in fig. 7, comprises following steps:
The big data processing platform collection network of core net and the data of user terminal, analysis obtain user terminal at certainThe path locus habit of (such as weeks/months), the big data processing platform can be deployed in core net or network management in the section timeMEC (the Mobile Edge Computing, movement of the heart or RAN (Radio Access Network, wireless access network) sideEdge calculations) platform；
When user terminal is initially accessed or is to switch to certain base station, user's path locus habit is sent to by core netUser terminal is currently accessed base station, i.e. source base station；
The switching of user's path locus habit and user's history that source base station is sent using core-network side records, and calculating obtainsUnder the conditions of motion track before known users terminal user terminal be switched to its adjacent base station being currently accessed base station locationConditional probability；
The big data processing platform of source base station side utilizes historical user to access base station and its RSRP time sequence of neighboring stationThe RSRQ time series of column, historical user to access base station and its neighboring station, it is known that user under the conditions of the motion track before userIn the conditional probability and history switching result and user's performance of handoffs for being currently accessed base station location and being switched to its adjacent base station, adoptHandover decisions model is trained with machine learning algorithm；As shown in figure 9, the input data of handover decisions model training includesBut it is not limited to: when the RSRQ of the RSRP time series of user to access base station and its neighboring station, user to access base station and its neighboring stationBetween sequence, user be switched to the conditional probability of other adjacent base stations, customer service QoS (service quality) being currently accessed base stationDemand, network resource conditions, load of base station, such as PRB (physical resource block, Physical Resource Block) occupancy feelingsCondition.The output data of handover decisions model is the optimal switching base judged according to history switching result and user's performance of handoffsIt stands；
The RSRP that source base station measures in real time according to uplink, RSRQ judge whether to enter measurement event as defined in LTE system；
If source base station is according to the RSRP measured in real time, RSRQ time series, and is currently accessed base station into measurement eventIt is switched to input of the probability of other adjacent base stations as trained handover decisions model, obtains switching base station decision；
If handover-target base station is not to be currently accessed base station, i.e. handover decisions model determination should switch, and be currently accessedSwitching request is sent to target BS by X2 interface by base station.
Process later is consistent with LTE mobility switching flow, comprising:
Target BS returns to Handover Request Acknowledge (ACK) message to source base station；
Source base station sends the RRC connection reconfiguration comprising mobility control information to user terminal and sets message；
User terminal leaves source cell and synchronous with the foundation of new cell；
The data transmitted in caching are transmitted to target BS by source base station；
User terminal is synchronized with target BS；
Target BS sends ascending resource assignment messages and terminal timed message to user terminal；
User terminal is set to target BS transmission RRC connection reconfiguration and completes message；
Target BS updates request to core net transmitting path convert requests and user plane；
Core net converts downstream data paths；
Core net returns to user plane to target BS and updates response and path integration request confirmation；
Target BS sends terminal contexts release message to source base station；
Source base station discharges user terminal related resource.
Embodiment two
The switching method of the present embodiment is applied to user terminal, as shown in Figure 8, comprising the following steps:
The big data processing platform collection network of core net and the data of user terminal, analysis obtain user terminal at certainThe path locus habit of (such as weeks/months), the big data processing platform can be deployed in core net or network management in the section timeMEC (the Mobile Edge Computing, movement of the heart or RAN (Radio Access Network, wireless access network) sideEdge calculations) platform；
When user terminal is initially accessed or is to switch to certain base station, user's path locus habit is sent to by core netUser terminal is currently accessed base station, i.e. source base station；
The switching of user's path locus habit and user's history that source base station is sent using core-network side records, and calculating obtainsUnder the conditions of motion track before known users terminal user terminal be switched to its adjacent base station being currently accessed base station locationConditional probability；
The big data processing platform of source base station side utilizes historical user to access base station and its RSRP time sequence of neighboring stationThe RSRQ time series of column, historical user to access base station and its neighboring station, it is known that user under the conditions of the motion track before userIn the conditional probability and history switching result and user's performance of handoffs for being currently accessed base station location and being switched to its adjacent base station, adoptHandover decisions model is trained with machine learning algorithm；As shown in figure 9, the input data of handover decisions model training includesBut it is not limited to: when the RSRQ of the RSRP time series of user to access base station and its neighboring station, user to access base station and its neighboring stationBetween sequence, user be switched to the conditional probability of other adjacent base stations, customer service QoS (service quality) being currently accessed base stationDemand, network resource conditions, load of base station, such as PRB (physical resource block, Physical Resource Block) occupancy feelingsCondition.The output data of handover decisions model is the optimal switching base judged according to history switching result and user's performance of handoffsIt stands；
Trained handover decisions model is sent to the user terminal for accessing the base station by source base station；
User terminal judges whether to enter measurement event as defined in LTE system according to the RSRP measured in real time, RSRQ；
If user terminal is according to the RSRP measured in real time, RSRQ time series, and is currently accessed base into measurement eventStation is switched to input of the probability of other adjacent base stations as trained handover decisions model, obtains switching base station decision；
If handover-target base station is not to be currently accessed base station, i.e. handover decisions model determination should switch, and UE, which is reported, to be cutIt changes target BS list and (is marked to base station, the base station IDs of handover-target base station list arranged in sequence UE expectation switching are currently accessedKnow)；
Source base station determines the target BS that UE finally switches according to network load condition etc.；
Switching request is sent to target BS by X2 interface by source base station.
Process later is consistent with LTE mobility switching flow, comprising:
Target BS returns to Handover Request Acknowledge (ACK) message to source base station；
Source base station sends the RRC connection reconfiguration comprising mobility control information to user terminal and sets message；
User terminal leaves source cell and synchronous with the foundation of new cell；
The data transmitted in caching are transmitted to target BS by source base station；
User terminal is synchronized with target BS；
Target BS sends ascending resource assignment messages and terminal timed message to user terminal；
User terminal is set to target BS transmission RRC connection reconfiguration and completes message；
Target BS updates request to core net transmitting path convert requests and user plane；
Core net converts downstream data paths；
Core net returns to user plane to target BS and updates response and path integration request confirmation；
Target BS sends terminal contexts release message to source base station；
Source base station discharges user terminal related resource.
The embodiment of the invention also provides a kind of base station, including memory, processor and it is stored on the memory simultaneouslyThe computer program that can be run on the processor；The processor realizes switching side as described above when executing described programMethod.
The embodiment of the invention also provides a kind of user terminal, including memory, processor and it is stored in the memoryComputer program that is upper and can running on the processor；It realizes when the processor executes described program and cuts as described aboveChange method.
The embodiment of the invention also provides a kind of computer readable storage mediums, are stored thereon with computer program, the journeyThe step in switching method as described above is realized when sequence is executed by processor.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any methodOr technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), movesState random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasableProgrammable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devicesOr any other non-transmission medium, can be used for storage can be accessed by a computing device information.As defined in this article, it calculatesMachine readable medium does not include temporary computer readable media (transitory media), such as the data-signal and carrier wave of modulation.
The above is a preferred embodiment of the present invention, it is noted that for those skilled in the artFor, without departing from the principles of the present invention, it can also make several improvements and retouch, these improvements and modificationsIt should be regarded as protection scope of the present invention.