WO2020098575A1 - 一种容量规划方法及装置 - Google Patents

一种容量规划方法及装置 Download PDF

Info

Publication number
WO2020098575A1
WO2020098575A1 PCT/CN2019/116670 CN2019116670W WO2020098575A1 WO 2020098575 A1 WO2020098575 A1 WO 2020098575A1 CN 2019116670 W CN2019116670 W CN 2019116670W WO 2020098575 A1 WO2020098575 A1 WO 2020098575A1
Authority
WO
WIPO (PCT)
Prior art keywords
distribution model
distribution
service
user experience
parameter value
Prior art date
Application number
PCT/CN2019/116670
Other languages
English (en)
French (fr)
Inventor
彭曦
白铂
张弓
兰宇
戚浩峰
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Priority to EP19885672.6A priority Critical patent/EP3873123B1/en
Publication of WO2020098575A1 publication Critical patent/WO2020098575A1/zh
Priority to US17/321,104 priority patent/US11736954B2/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W16/00Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
    • H04W16/18Network planning tools
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W16/00Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
    • H04W16/22Traffic simulation tools or models
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W24/00Supervisory, monitoring or testing arrangements
    • H04W24/02Arrangements for optimising operational condition
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W28/00Network traffic management; Network resource management
    • H04W28/02Traffic management, e.g. flow control or congestion control
    • H04W28/08Load balancing or load distribution
    • H04W28/09Management thereof
    • H04W28/0925Management thereof using policies
    • H04W28/095Management thereof using policies based on usage history, e.g. usage history of devices
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W28/00Network traffic management; Network resource management
    • H04W28/02Traffic management, e.g. flow control or congestion control
    • H04W28/08Load balancing or load distribution
    • H04W28/09Management thereof
    • H04W28/0958Management thereof based on metrics or performance parameters
    • H04W28/0967Quality of Service [QoS] parameters
    • H04W28/0983Quality of Service [QoS] parameters for optimizing bandwidth or throughput
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W28/00Network traffic management; Network resource management
    • H04W28/16Central resource management; Negotiation of resources or communication parameters, e.g. negotiating bandwidth or QoS [Quality of Service]
    • H04W28/18Negotiating wireless communication parameters
    • H04W28/20Negotiating bandwidth
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W24/00Supervisory, monitoring or testing arrangements
    • H04W24/10Scheduling measurement reports ; Arrangements for measurement reports

Definitions

  • This application relates to the technical field of mobile communications, and in particular to a capacity planning method and device.
  • the European Telecommunications Standards Institute (The European Telecommunications Standards Institute, ETSI) proposed Mobile Edge Computing (MEC) as a future network architecture solution, through Long Term Evolution (LTE) and Fifth Generation (5th generation, 5G) Mobile network edges (such as base stations) provide information technology (IT) service environment and computing capabilities, sinking network services to the wireless access side closer to users, thereby reducing the operating pressure of the core network , And allows network operators to improve the user experience.
  • ETSI European Telecommunications Standards Institute
  • MEC Mobile Edge Computing
  • MEC Mobility Control
  • the key step is to reasonably predict network traffic, not only to protect the user experience, but also to control operating costs. This is not only a research hotspot in academia, but also an urgent need for industry to solve problem.
  • network services are becoming more and more complex, gradually changing from voice services to data services such as video and interactive entertainment.
  • the statistics of the live network show that its business volume presents significant burst characteristics, so the “average effect” obeyed by the traditional voice network is no longer applicable, and a new network business model is urgently needed to describe it.
  • Traditional average performance indicators cannot truly reflect network performance.
  • Statistics on the live network show that average performance indicators (such as hourly averages) often mask busy resource shortages and poor user experience during busy hours. Therefore, we need to introduce new metrics to reflect directly User experience rate, and then accurately plan user capacity.
  • This application provides a method and device for capacity planning to improve the accuracy of user capacity planning.
  • the present application provides a capacity planning method, including: a capacity planning device receiving a distribution parameter value of a first distribution model and a distribution parameter value of a second distribution model from a service measurement device, the first distribution model and the second distribution model Respectively, a distribution model matched by the number of service packets in each transmission time interval and a distribution model matched by the length of the service packets acquired by the service measurement device within the set duration.
  • the capacity planning device performs bandwidth control according to the first distribution model, the second distribution model, the distribution parameter value of the first distribution model, and the distribution parameter value of the second distribution model.
  • the distribution model is matched according to the number of service packets in each transmission time interval within a set duration to obtain a matched first distribution model, and the distribution model is matched according to the length of service packets to obtain a second distribution model, and Bandwidth control is performed based on the first distribution model, the second distribution model, the distribution parameters of the first distribution model, and the distribution parameters of the second distribution model. Since the bandwidth control is performed based on the matching distribution model and the distribution parameters of the distribution model, it helps to provide users with more accurate capacity planning.
  • the capacity planning apparatus determines a user experience rate distribution model based on the first distribution model, the second distribution model, the downlink transmission rate of the base station, and the transmission time interval.
  • the capacity planning device performs bandwidth control according to the user experience rate distribution model, the distribution parameter value of the first distribution model, the distribution parameter value of the second distribution model, and the service quality requirement parameter value.
  • a user experience rate distribution model is introduced, and the user experience rate distribution model is based on the first distribution model that matches the number of service packets in each transmission time interval within a set duration, and the length of service packets matches
  • the second distribution model, the downlink transmission rate of the base station, and the length of the time interval are determined, because the distribution model of the number of service packets and the length of the service packet length are adopted in the transmission time interval granularity, and the length of the service packet is in bits In units, this solution can obtain the distribution model of the total number of bits of the service for each transmission time interval, thereby providing a finer bit-level granularity, so the determined user experience rate distribution model can more accurately characterize the user experience Speed, which can provide users with more accurate capacity planning.
  • the downlink transmission rate of the base station is the downlink transmission rate of the base station for the terminal.
  • the downlink transmission rate of the base station is the base station's target for all terminals that access the base station The sum of downlink transmission rates.
  • the service quality requirement parameter value is a preset bandwidth utilization rate
  • the capacity planning device is based on the user experience rate distribution model, the distribution parameter value of the first distribution model, and the distribution parameter of the second distribution model Value and service quality requirement parameter value, perform bandwidth control, including: capacity planning device is determined according to the user experience rate distribution model, the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model and the preset bandwidth utilization rate The average user experience rate value when busy.
  • the capacity planning device performs bandwidth control according to the value of the average user experience rate during busy hours. Based on this solution, the value of the average user experience rate during busy hours is introduced, so that based on this parameter, the user's experience rate during busy hours can be more finely characterized so as to provide more accurate capacity planning for users.
  • the first distribution model is the Zeta distribution model
  • the distribution model parameters of the first distribution model include s
  • the second distribution model is the Pareto distribution model
  • the distribution model parameters of the second distribution model include m and ⁇ .
  • the user experience rate distribution model is:
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to buffer service packets
  • is the transmission time interval
  • ⁇ () is the Riemann function
  • E [S] is a transmission time
  • p 0 is the probability that the number of arriving service packets is zero.
  • the capacity planning device determines the average user experience rate value during busy hours according to the following formula:
  • the capacity planning device performs bandwidth control according to the average value of the busy user experience rate, including: if the difference between the average value of the busy user experience rate and the downlink transmission rate of the base station is greater than the first difference threshold, Then increase the bandwidth. Or, if the difference between the busy user average experience rate value and the downlink transmission rate of the base station is less than the second difference threshold, the bandwidth is reduced.
  • the quality of service parameter value is a preset user experience rate satisfaction;
  • the capacity planning device is based on the user experience rate distribution model, the distribution parameter value of the first distribution model, and the value of the second distribution model
  • the distribution parameter value and the service quality parameter value perform bandwidth control, including: the capacity planning device according to the user experience rate distribution model, the distribution parameter value of the first distribution model, the distribution parameter value of the second distribution model and the preset user experience rate satisfaction Degree to determine the lower limit of user experience rate.
  • the capacity planning device performs bandwidth control according to the lower limit of the user experience rate. Based on this solution, the lower limit of the user experience rate is introduced, so that based on this parameter, the user experience rate can be more finely characterized, so as to provide more accurate capacity planning for the user.
  • the first distribution model is the Zeta distribution model
  • the distribution model parameters of the first distribution model include s
  • the second distribution model is the Pareto distribution model
  • the distribution model parameters of the second distribution model include m and ⁇ .
  • the user experience rate distribution model is:
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to buffer service packets
  • is the transmission time interval
  • ⁇ () is the Riemann function
  • E [S] is a transmission time
  • p 0 is the probability that the number of arriving service packets is zero.
  • the capacity planning device determines the lower limit of the user experience rate according to the following formula:
  • R min is the lower limit of the user experience rate
  • is the preset user experience rate satisfaction.
  • the capacity planning apparatus performs bandwidth control according to the lower limit value of the user experience rate, including: if the difference between the lower limit value of the user experience rate and the downlink transmission rate of the base station is greater than the third difference threshold, increase bandwidth. Or, if the difference between the busy user average experience rate value and the downlink transmission rate of the base station is less than the fourth difference threshold, the bandwidth is reduced.
  • the capacity planning device further receives the identification information of the first distribution model and the identification information of the second distribution model from the service measurement device.
  • the identification information of the first distribution model is used
  • the identification information of the second distribution model is used to identify the selected second distribution model.
  • the capacity planning device further receives identification information of a service arrival model from the service measurement device, and the identification information of the service arrival model is used to identify the selected first distribution model and The service arrival model corresponding to the second distribution model.
  • the capacity planning device determines the selected first distribution model and the second distribution model according to the identification information of the service arrival model.
  • the present application provides a method for capacity planning.
  • the method includes: a service measurement device acquiring the number of service packets and the length of service packets in each transmission time interval within a set duration.
  • the service measurement device determines the distribution parameter value of the first distribution model that matches the number of service packages, and determines the distribution parameter value of the second distribution model that matches the length of the service packages.
  • the service measurement device sends the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model to the capacity planning device.
  • the first distribution model, the second distribution model, the downlink transmission rate of the base station and the transmission time interval are used to determine the user Experience rate distribution model.
  • the user experience rate distribution model, the parameter value of the first distribution model, the parameter value of the second distribution model, and the quality of service requirement parameter value can be used to perform bandwidth control.
  • This solution introduces a user experience rate distribution model, and the user experience rate distribution model is based on the first distribution model that matches the number of service packets in each transmission time interval within a set duration, and the first Two distribution model The base station downlink transmission rate and the length of the time interval are determined, because the distribution model of the number of service packets with the transmission time interval granularity and the distribution model of the length of the service packets are used, and the length of the service packets is in bits ,
  • This solution can obtain the distribution model of the total number of bits of the service for each transmission time interval, thereby providing a finer bit-level granularity, so the determined user experience rate distribution model can more accurately characterize the user experience rate, and then Can provide users with more accurate capacity planning.
  • the service measurement device determining the distribution parameter value of the first distribution model matching the number of service packages includes: the service measurement device uses the number of service packages to fit at least Two distribution models to obtain the fitting degree of each distribution model and the distribution parameter value of the distribution model; the service measurement device determines that the distribution model with the highest degree of fit is the first distribution model matching the number of the service packages And determining that the distribution parameter value of the distribution model with the highest degree of fit is the distribution parameter value of the first distribution model.
  • the at least two distribution models include one or more of the following distribution models: Poisson distribution model and Zeta distribution model.
  • the service measurement device uses the number of service packages to fit the Poisson distribution model to obtain the first fit and the distribution parameter value of the Poisson distribution model.
  • the service measurement device uses the number of service packages to fit the Zeta distribution model to obtain the second fitting degree and the distribution parameter value of the Zeta distribution model. If the first fitting degree is greater than the second fitting degree, the service measurement device determines that the distribution model parameter of the Poisson distribution model is the distribution parameter value of the first distribution model matching the number of service packages, and the first distribution model is the Poisson distribution model.
  • the service measurement device determines that the distribution parameter value of the Zeta distribution model is the distribution parameter value of the first distribution model matching the number of service packages, and the first distribution model is Zeta Distribution model.
  • the Zeta distribution model is a heavy-tailed distribution model, which can accurately reflect the user's burst business traffic.
  • the service measurement device determining the distribution parameter value of the second distribution model that matches the length of the service package includes the service measurement device using the length of the service package to fit at least Two distribution models to obtain the fitting degree of each distribution model and the distribution parameter value of the distribution model; the service measurement device determines that the distribution model with the highest degree of fit is the second distribution model whose length of the service package matches And determining that the distribution parameter value of the distribution model with the highest fitting degree is the distribution parameter value of the second distribution model.
  • the at least two distribution models include one or more of the following distribution models: exponential distribution model and Pareto distribution model.
  • the service measurement device uses the length of the service package to fit the exponential distribution model to obtain the third fitting degree and the distribution parameter value of the exponential distribution model.
  • the service measurement device uses the length of the service package to fit the Pareto distribution model to obtain the fourth fit and the distribution parameter value of the Pareto distribution model. If the third degree of fitting is greater than the fourth degree of fitting, the service measurement device determines that the distribution parameter value of the exponential distribution model is the value of the distribution parameter of the second distribution model that matches the length of the service packet, and the second distribution model is the exponential distribution model.
  • the service measurement device determines that the distribution model parameter of the Pareto distribution model is the distribution model parameter of the second distribution model matching the length of the service package, and the second distribution model is Pareto Distribution model.
  • the Pareto distribution model is a heavy-tailed distribution model, which can accurately reflect the user's burst business traffic.
  • the service measurement device determining the distribution parameter value of the first distribution model matching the number of the service packages includes: the service measurement device fitting the first preset distribution model, Obtaining the distribution parameter value of the first preset distribution model; the service measuring device determines that the first preset distribution model is the first distribution model whose number of service packages match, and determines the first preset The distribution parameter value of the distribution model is the distribution parameter value of the first distribution model.
  • the service measurement device determining the distribution parameter value of the second distribution model matching the length of the service package includes: the service measurement device fits the second preset distribution model to obtain the second preset distribution model Distribution parameter value; the service measurement device determines that the second preset distribution model is a second distribution model that matches the length of the service packet, and determines that the distribution parameter value of the second preset distribution model is the first The distribution parameter value of the second distribution model.
  • the first preset distribution model is a Poisson distribution model, or a Zeta distribution model
  • the second preset distribution model is an exponential distribution model, or a Pareto distribution model.
  • the service measurement device sending the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model to the capacity planning device includes: the service measurement device sending a first report message to the capacity planning device, The first report message includes the identifier of the first distribution model and the identifier of the second distribution model. The service measurement device receives the first response message for the first report message from the capacity planning device. The service measurement device sends a second report message to the capacity planning device. The second report message includes the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model.
  • the service measurement device receives the first notification message from the capacity planning device, and the first notification message includes a bandwidth control strategy.
  • the service measurement device sends a first confirmation message for the first notification message to the capacity planning device.
  • the service measurement device receives the second notification message from the capacity planning device, and the second notification message includes the bandwidth value.
  • the service measurement device performs bandwidth control based on the bandwidth value.
  • the service measurement device is deployed on the terminal, and the capacity planning device is deployed on the base station.
  • the service measuring device acquires the number of service packets and the length of the service packets in each transmission time interval within the set duration, including: the service measuring device acquires the number of service packets of the terminal within each transmission time interval within the set duration And the length of the service package of the terminal.
  • the service measurement device is deployed on the base station, and the capacity planning device is deployed on the mobile edge computing server.
  • the service measurement device acquires the number of service packets and the length of the service packet in each transmission time interval within the set duration, including: the service measurement device acquires each access to the base station within each transmission time interval within the set duration The number of service packets of the terminal and the length of the service packets of each terminal.
  • the service measurement device acquires the number of service packets of the base station and the length of the service packets of the base station within each transmission time interval within a set time duration.
  • the service measurement device acquiring the number of service packets and the length of the service packet in each transmission time interval within a set duration includes: the service measurement device periodically acquiring The number of service packets and the length of the service packets in each transmission time interval within the set duration; or, the service measurement device periodically acquires each transmission that satisfies the preset busy hour condition within the set duration The number of service packets in the time interval and the length of the service packets.
  • the present application provides a communication device having a function of implementing a service measurement device or a capacity planning device in the above method embodiment.
  • This function can be realized by hardware, and can also be realized by hardware executing corresponding software.
  • the hardware or software includes one or more units or modules corresponding to the above functions.
  • the communication device includes: a processor, a memory, a bus, and a communication interface; the memory stores computer-executed instructions, the processor and the memory are connected through the bus, and when the communication device is running, the The processor executes the computer-executed instructions stored in the memory, so that the communication device executes the capacity planning method in the first aspect described above, or in any implementation manner of the first aspect described above.
  • the communication device may be a capacity planning device.
  • the communication device may also be a chip, such as a chip for a service measurement device, or a chip for a capacity planning device, the chip includes a processing unit, and optionally, a storage unit, The chip may be used to execute the capacity planning method in the capacity planning method in the first aspect described above, or in any implementation manner of the first aspect described above.
  • the present application provides a communication device, which has the function of implementing a service measurement device or a capacity planning device in the above method embodiment.
  • This function can be realized by hardware, and can also be realized by hardware executing corresponding software.
  • the hardware or software includes one or more units or modules corresponding to the above functions.
  • the communication device includes: a processor, a memory, a bus, and a communication interface; the memory stores computer-executed instructions, the processor and the memory are connected through the bus, and when the communication device is running, the The processor executes the computer-executed instructions stored in the memory, so that the communication device executes the capacity planning method in the second aspect described above, or in any implementation manner of the second aspect described above.
  • the communication device may be a traffic measurement device.
  • the communication device may also be a chip, such as a chip for a service measurement device, or a chip for a capacity planning device, the chip includes a processing unit, and optionally, a storage unit, The chip may be used to perform the capacity planning method in the capacity planning method in the second aspect described above or in any implementation manner of the second aspect described above.
  • the present application provides a computer storage medium that stores computer software instructions for the above-mentioned terminal, which includes a program designed to execute the above-mentioned first aspect, or any implementation manner of the first aspect.
  • the present application provides a computer storage medium that stores computer software instructions for the above terminal, which includes a program designed to execute the above first aspect, or any implementation manner of the first aspect.
  • the present application provides a computer program product.
  • the computer program product includes computer software instructions, and the computer software instructions can be loaded by a processor to implement the flow in the capacity planning method of the first aspect or any one of the above aspects.
  • this application provides a computer program product.
  • the computer program product includes computer software instructions, which can be loaded by a processor to implement the process in the capacity planning method of the second aspect or any one of the second aspect.
  • the present application provides a system including the service measurement device according to any one of the above aspects and the capacity planning device according to any one of the above aspects.
  • FIG. 1 is a schematic diagram of a possible network architecture provided by this application.
  • FIG. 3 is a schematic diagram of a device provided by this application.
  • FIG. 5 is a schematic diagram of another device provided by the present application.
  • the network architecture is the MEC network architecture.
  • each network edge node ie, base station
  • several base stations are connected to the sink node on the base station side
  • the sink node is connected to the MEC server.
  • the MEC server is located between the wireless access point and the core network, and has storage and calculation capabilities. By superimposing the MEC server on the base station side, localized services can be provided to users, thereby effectively saving system resources on the core network side and significantly shortening the corresponding response time.
  • the capacity planning method proposed in this application is jointly completed by the service measurement device and the capacity planning device.
  • the service measurement device is responsible for measuring data services and learning service feature parameters, and the capacity planning device is responsible for overall capacity planning based on the service feature parameters reported by each service measurement device.
  • the service measurement device is deployed on the base station, and the capacity planning device is deployed on the MEC server.
  • the service measurement device is deployed on the terminal, and the capacity planning device is deployed on the base station.
  • the terminal is a device with wireless transceiver function.
  • the terminal can be deployed on land, including indoor or outdoor, handheld, or vehicle-mounted; it can also be deployed on the water (such as ships, etc.); it can also be deployed in the air (such as Airplanes, balloons and satellites etc.).
  • the terminal may be a mobile phone, a tablet computer, a computer with wireless transceiver function, a virtual reality (VR) terminal, an augmented reality (AR) terminal, an industrial control (industrial control) Wireless terminal in self-driving, wireless terminal in self-driving, wireless terminal in remote medical, wireless terminal in smart grid, wireless terminal in transportation safety, Wireless terminals in smart cities (smart cities), wireless terminals in smart homes (smart homes), and may also include user equipment (UE).
  • VR virtual reality
  • AR augmented reality
  • UE user equipment
  • Terminals can also be cellular phones, cordless phones, session initiation protocol (SIP) phones, wireless local loop (WLL) stations, personal digital assistants (personal digital assistants, PDAs), and wireless communication functions Handheld devices, computing devices, or other processing devices connected to wireless modems, in-vehicle devices, wearable devices, terminal devices in future 5G networks or terminals in future public land mobile communications networks (PLMN) Equipment etc.
  • Terminals are sometimes referred to as terminal equipment, user equipment (UE), access terminal equipment, vehicle-mounted terminals, industrial control terminals, UE units, UE stations, mobile stations, mobile stations, remote stations, remote terminal equipment, mobile Equipment, UE terminal equipment, terminal equipment, wireless communication equipment, UE agents or UE devices, etc.
  • the terminal may also be fixed or mobile. The embodiments of the present application are not limited thereto.
  • a base station is a device that provides wireless communication functions for terminals.
  • Base stations include, but are not limited to, 5G next-generation base stations (gNodeB, gNB), evolved node B (evolved node B, eNB), radio network controller (radio network controller, RNC), and node B (node B, NB), base station controller (BSC), base transceiver station (BTS), home base station (for example, home evolved node B, or home node B, HNB), baseband unit (BBU) , Transmission point (transmitting and receiving point, TRP), transmitting point (transmitting point, TP), mobile switching center, etc.
  • gNodeB 5G next-generation base stations
  • eNode B evolved node B
  • RNC radio network controller
  • RNC radio network controller
  • node B node B
  • BSC base station controller
  • BTS base transceiver station
  • home base station for example, home evolved node B, or home node B, HNB
  • the service packet is used as the minimum queuing element, so as to calculate the performance indicators such as the average queue length, waiting delay and transmission rate of the network.
  • the performance indicators such as the average queue length, waiting delay and transmission rate of the network.
  • Existing technologies widely use classic Poisson, exponential and other models to model the arrival of network services.
  • the service arrival is subject to the Poisson distribution model
  • the duration of the service is subject to the exponential distribution model.
  • These models are all light-tail models, which are easier to analyze and derive their mathematical properties, and have been shown to have good approximations to traditional discourse services and low-speed data services.
  • the above-mentioned classic light-tail service model has an "average effect", that is, the average service volume is highly representative, and the deviation between the random arrival service volume and the average service volume is small, that is, the tail probability attenuation rate is not slower than the exponential decay.
  • this distribution model is often quite different from the phenomena observed on the live network. Through the data measurement and analysis of multiple live network scenarios, it is found that the current traffic arriving in the current data network often presents serious bursts, and the tailing attenuation of the traffic probability distribution model is slower than the exponential distribution model, presenting a "heavy tail", making The average business volume is not representative. Therefore, existing business models cannot characterize high-speed and complex data services, and have application limitations.
  • the server of the base station of the LTE system and the 5G system uses a single transmission time interval (Transmission Time Interval, TTI) as the smallest scheduling unit in the time domain, and a service packet may be between adjacent TTIs Was cut.
  • TTI Transmission Time Interval
  • the existing capacity planning method assumes that the business model follows the light-tailed distribution model, but the actual measured data on the live network indicates that the arriving traffic often has burst characteristics, and the average network performance is often not representative, which is different from the busy-time network performance Significantly, the existing light tail model is no longer applicable. For this reason, a new model has been established in this application.
  • the existing capacity planning method is based on the classic queuing theory in units of packets, which is inconsistent with the scheduling method adopted by the servers of LTE and 5G base stations. Therefore, this application calculates in bits.
  • This application proposes performance indicators that can truly reflect user experience based on the data traffic characteristics of existing networks, gives network traffic forecasts, and solves network capacity planning problems.
  • FIG. 2 it is a capacity planning method provided by the present application.
  • the method includes the following steps:
  • Step 201 The service measurement device obtains the number of service packets and the length of the service packets in each TTI within a set duration.
  • the unit of the length of the service packet is bits.
  • the service packages in TTI1 include service package 1, service package 2 and service package 3, and the lengths are 100 bits, 200 bits and 300 bits respectively, then the number of service packages in TTI1 is 3 and the lengths of the service packages are respectively 100 bits, 200 bits, 300 bits.
  • each service measurement device acquires the number of service packets of the terminal and the service of the terminal within each TTI within a set duration The length of the package. For example, the service measurement device on terminal 1 acquires the number of service packets of terminal 1 in each TTI and the length of the service packet of terminal 1 in the TTI within a set duration, and the service measurement device on terminal 2 acquires the device The number of service packets of the terminal 2 in each TTI within the timing length and the length of the service packets of the terminal 2 in the TTI, and so on.
  • the service measurement device may obtain the number and length of service packets at the terminal granularity, Specifically, each service measurement device acquires the number of service packets of each terminal that accesses the base station and the length of the service packets of each terminal within each TTI within a set duration.
  • the service measurement device on base station 1 acquires the number of service packets of terminal 1 in each TTI within the set duration and the The length of the service packet of the terminal 1, the number of service packets of the terminal 2 in each TTI within the set duration and the length of the service packet of the terminal 2 within the TTI, and each TTI within the set duration The number of service packets of terminal 3 and the length of service packets of terminal 3 in the TTI.
  • the service measurement device may also obtain the number and length of service packets at the granularity of the base station.
  • each service measurement device obtains the number of service packets of the base station in each TTI within a set duration
  • the length of the service packet of the base station refers to the sum of the number of service packets of all terminals in each TTI that access to the base station, in each TTI here
  • the length of the service packet of the base station refers to the sum of the lengths of the service packets of all terminals accessing the base station in each TTI.
  • the service measurement device on the base station 1 acquires the number of service packets of the base station 1 in each TTI and the base station 1 in the TTI within the set duration The length of the service packet.
  • the number of service packets of base station 1 in each TTI here refers to the total number of service packets of terminal 1 and terminal 2 in each TTI.
  • the service of base station 1 in each TTI The length of the packet refers to the total length of the service packets of terminal 1 and terminal 2 in each TTI.
  • the service measurement device determines the distribution parameter value of the first distribution model that matches the number of service packages, and determines the distribution parameter value of the second distribution model that matches the length of the service packages.
  • the distribution model may also be called a distribution function, or simply a distribution.
  • the service measurement device may determine the distribution parameter value of the first distribution model that matches the number of service packages according to the following method: the service measurement device uses the number of service packages to fit at least two distribution models to obtain the fitting degree of each distribution model and The distribution parameter value of the distribution model, and then the service measurement device determines that the distribution model with the highest degree of fit is the first distribution model matching the number of service packages, and determines that the distribution parameter value of the distribution model with the highest degree of fit is the first The distribution parameter value of a distribution model.
  • the at least two distribution models include one or more of the following distribution models: Poisson distribution model and Zeta distribution model.
  • the service measurement device uses the acquired number of service packages in each TTI within the set duration to fit the Poisson distribution model to obtain the first fitting degree and the distribution parameter value of the Poisson distribution model.
  • the service measurement device uses the number of service packages to fit the Zeta distribution model to obtain the second fitting degree and the distribution parameter value of the Zeta distribution model. If the first fitting degree is greater than the second fitting degree, the service measurement device determines that the distribution model parameter of the Poisson distribution model is the distribution parameter value of the first distribution model matching the number of service packages, and the first distribution model is the Poisson distribution model.
  • the service measurement device determines that the distribution parameter value of the Zeta distribution model is the distribution parameter value of the first distribution model matching the number of service packages, and the first distribution model is Zeta Distribution model.
  • the Possion distribution model takes ⁇ > 0 as a parameter and its form is:
  • the Zeta distribution model takes s> 0 as a parameter, ⁇ (s) represents the Riemann Zeta function, p 0 represents the probability that the number of arriving service packets is zero, and its form is:
  • the determined distribution parameter value of the first distribution model is the value of ⁇ . If the service measurement device determines that the first distribution model whose number of service packages match is the Zeta distribution model, the distribution parameter value of the determined first distribution model is the value of s.
  • the above method helps to improve the accuracy of capacity planning by selecting a distribution model with the best fit to the number of service packages from the multiple distribution models as the first distribution model to use.
  • the service measurement device may determine the distribution parameter value of the second distribution model that matches the length of the service package according to the following method: the service measurement device uses the length of the service package to fit at least two distribution models to obtain a fit for each distribution model Distribution parameter values of the degree and distribution model; then, the service measurement device determines the distribution model with the highest degree of fit as the second distribution model matching the length of the service package, and determines the distribution parameter value of the distribution model with the highest degree of fit as the second The distribution parameter value of the distribution model.
  • the at least two distribution models include one or more of the following distribution models: exponential distribution model and Pareto distribution model.
  • the service measurement device uses the length of the service package to fit the exponential distribution model to obtain the third fitting degree and the distribution parameter value of the exponential distribution model.
  • the service measurement device uses the length of the service package to fit the Pareto distribution model to obtain the fourth fit and the distribution parameter value of the Pareto distribution model. If the third degree of fitting is greater than the fourth degree of fitting, the service measurement device determines that the distribution parameter value of the exponential distribution model is the value of the distribution parameter of the second distribution model that matches the length of the service packet, and the second distribution model is the exponential distribution model.
  • the service measurement device determines that the distribution model parameter of the Pareto distribution model is the distribution model parameter of the second distribution model matching the length of the service package, and the second distribution model is Pareto Distribution model.
  • the exponential distribution model takes ⁇ > 0 as the parameter, and its form is:
  • the Pareto distribution model takes m> 0 and ⁇ > 0 as parameters, and its form is:
  • the distribution parameter value of the determined exponential distribution model is the value of ⁇ . If the service measurement device determines that the second distribution model whose length of the service packet matches is the Pareto distribution model, the determined distribution parameter values of the second distribution model are the values of m and ⁇ .
  • the above method helps to improve the accuracy of capacity planning by selecting a distribution model with the best fit to the length of the service package as the second distribution model to use from multiple distribution models.
  • the service measurement device may also determine the distribution parameter value of the first distribution model matching the number of service packages and the distribution parameter value of the first distribution model matching the length of the service package according to the following methods :
  • the service measurement device fits the first preset distribution model to obtain the distribution parameter value of the first preset distribution model; the service measurement device determines that the first preset distribution model is the first distribution model matching the number of the service packages, and determines The distribution parameter value of the first preset distribution model is the distribution parameter value of the first distribution model.
  • the service measurement device fits the second preset distribution model to obtain the distribution parameter value of the second preset distribution model; the service measurement device determines that the second preset distribution model is the second distribution model whose length of the service package matches, And determining that the distribution parameter value of the second preset distribution model is the distribution parameter value of the second distribution model.
  • the first preset distribution model is a Poisson distribution model, or a Zeta distribution model
  • the second preset distribution model is an exponential distribution model, or a Pareto distribution model.
  • the service measurement device uses the first preset distribution model as the first distribution model for matching the number of service packages, and uses the second preset distribution model as the second distribution model for matching the length of the service packages. Since there is no need to select a distribution model from multiple distribution models, the efficiency of capacity planning can be improved.
  • the above Zeta distribution model and Pareto distribution model represent discrete and continuous power law (heavy tail) distribution models, respectively.
  • the distribution model in which the number of service packages is matched and the distribution model in which the length of service packages are matched can constitute the four service arrival models shown in Table 1 below.
  • Table 1 Four business arrival models and typical application scenarios
  • the service can be considered to conform to the PP model, and the general scenarios are: macrocells, remote areas.
  • the general scenario is: when the business district is idle.
  • the business can be considered to conform to the PP model, and the general scenario is: when the business district is busy.
  • the method of “fitting” used is not limited, for example, it may be fitting using least square method.
  • Step 203 The service measurement device sends the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model to the capacity planning device.
  • the service measurement device sends the identification information of the first distribution model, the identification information of the second distribution model, the distribution parameter values of the first distribution model and the distribution parameter values of the second distribution model to the capacity planning device.
  • the identification information of the first distribution model is used to identify the selected first distribution model
  • the identification information of the second distribution model is used to identify the selected second distribution model.
  • the capacity planning device can learn the first distribution model and the second distribution model selected by the service measurement device, as well as the distribution parameter values of the first distribution model and the second distribution model.
  • the service measurement device sends identification information of the service arrival model to the capacity planning device, and the identification information of the service arrival model is used to identify the selected first distribution model and the second distribution model.
  • the capacity planning device may determine the selected first distribution model and the second distribution model according to the identification information of the service arrival model.
  • the service measurement device may determine that the selected first distribution model and second distribution model are Poisson The distribution model and the exponential distribution model; for another example, the service measurement device sends the identification information of the service arrival model to the capacity planning device to indicate the PP model in Table 1 above, then the service measurement device may determine the selected first distribution model and the The second distribution models are Poisson distribution model and Pareto distribution model, etc.
  • the service measurement device may send a first report message to the capacity planning device, where the first report message includes the identifier of the first distribution model and the identifier of the second distribution model. Therefore, the capacity planning device can learn the type of the matching distribution model determined by the service measurement device, so as to reserve corresponding resources.
  • the capacity planning device sends a first response message to the first report message to the service measurement device.
  • the service measurement device sends a second report message to the capacity planning device, where the second report message includes the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model.
  • the capacity planning apparatus may also send a second response message for the second report message to the service measurement apparatus. Based on this method, it is possible to send the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model to the capacity planning device.
  • the first report message includes the identifier of the first distribution model and the identifier of the second distribution model, which may be replaced by that the first report message includes the model identifier, such as "00" indicating the PE model and "01” indicating For the PP model, "10” indicates the ZE model and "11” indicates the ZP model, so that the types of the first distribution model and the second distribution model can also be sent to the capacity planning device.
  • Step 204 The capacity planning device performs bandwidth control according to the first distribution model, the second distribution model, the distribution parameters of the first distribution model, and the distribution parameters of the second distribution model.
  • the distribution model is matched according to the number of service packets in each transmission time interval within a set duration to obtain a matched first distribution model, and the distribution model is matched according to the length of service packets, A second distribution model is obtained, and bandwidth control is performed according to the first distribution model, the second distribution model, the distribution parameters of the first distribution model, and the distribution parameters of the second distribution model. Since the bandwidth control is performed according to the matching distribution model and the distribution parameters of the distribution model, it helps to provide users with more accurate capacity planning.
  • step 204 a specific implementation manner of the above step 204 is given below.
  • the capacity planning device performs bandwidth control according to the first distribution model, the second distribution model, the distribution parameters of the first distribution model, and the distribution parameters of the second distribution model, specifically including:
  • Step A The capacity planning device determines a user experience rate distribution model based on the first distribution model, the second distribution model, the base station downlink transmission rate, and the transmission time interval.
  • Step B The capacity planning device performs bandwidth control according to the user experience rate distribution model, the distribution parameter value of the first distribution model, the distribution parameter value of the second distribution model, and the service quality requirement parameter value.
  • the above step A an example will be described below.
  • the user experience rate here may also be called a user rate, or a user actual rate, or a user usage rate.
  • the user here refers to the terminal accessing the base station.
  • the user experience rate refers to the experience rate of the terminal corresponding to the service measurement device.
  • the downlink transmission rate of the base station is the downlink transmission rate of the base station for the terminal.
  • the service measurement device when the service measurement device is deployed on the base station and the capacity planning device is deployed on the MEC server, if the service measurement device measures the number and length of service packets of each terminal accessing the base station in each TTI with the terminal granularity, then The user experience rate refers to the experience rate of each terminal; if the service measurement device measures the number and length of service packets of all terminals accessing the base station in each TTI at the base station granularity, the user experience rate refers to Is the sum of the experience rates of all terminals.
  • the downlink transmission rate of the base station is the sum of the downlink transmission rates of the base station for all the terminals that access the base station.
  • the obtained user experience rate distribution model specifically includes the following four types:
  • User experience rate distribution model 1 corresponds to the PE model (the first distribution model is a Poisson distribution model, and the second distribution model is an exponential distribution model)
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to cache service packets
  • is TTI
  • is the parameter of the Poisson (Possion) distribution model
  • is the exponential distribution model parameter.
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to buffer service packets
  • is the TTI
  • E [S] is the expected value of the number of bits arriving in a TTI
  • is The parameters of the Possion distribution model
  • m and ⁇ are the parameters of the Pareto distribution model.
  • User experience rate distribution model 3 corresponds to the ZE model (the first distribution model is the Zeta distribution model and the second distribution model is the exponential distribution model)
  • p F q (a 1 , ..., a p ; b 1 , ..., b q ; x) is a generalized hypergeometric function, and its series expansion is:
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to buffer service packets
  • is TTI
  • E [S] is the expected value of the number of bits arriving in a TTI
  • ⁇ () Is the Riemann function
  • p 0 is the probability that the number of arriving service packets is zero
  • s is the parameter of the Zeta distribution model
  • is the parameter of the exponential distribution model.
  • User experience rate distribution model 4 corresponds to the ZP model (the first distribution model is the Zeta distribution model and the second distribution model is the Pareto distribution model)
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to buffer service packets
  • is TTI
  • E [S] is the expected value of the number of bits arriving in a TTI
  • ⁇ () Is the Riemann function
  • p 0 is the probability that the number of arriving service packets is zero
  • s is the parameter of the Zeta distribution model
  • m and ⁇ are the parameters of the Pareto distribution model.
  • step B an example will be described below.
  • the service quality requirement parameter value is the value of the service quality requirement parameter.
  • This application takes the service quality requirement parameter as a preset bandwidth utilization rate or a preset user experience rate satisfaction degree as an example for illustration.
  • Solution 1 The capacity planning device determines the average user experience rate value during busy hours based on the user experience rate distribution model, the distribution parameter values of the first distribution model, the distribution parameter values of the second distribution model, and the preset bandwidth utilization rate. The capacity planning device performs bandwidth control according to the value of the average user experience rate during busy hours.
  • Formula 1 Average user experience rate during busy hours: Corresponds to the user experience rate distribution model 1, that is, the average experience rate formula 1 for busy users can be obtained according to the user experience rate distribution model 1.
  • R is the downlink transmission rate of the base station
  • is TTI
  • is the parameter of the Poisson (Possion) distribution model
  • is the parameter of the exponential distribution model.
  • Formula 2 for average user experience rate during busy hours Corresponding to user experience rate distribution model 2, that is, formula 2 for average user experience rate during busy hours can be obtained according to user experience rate distribution model 2.
  • R is the downlink transmission rate of the base station
  • is TTI
  • is the parameter of the Poisson (Possion) distribution model
  • m and ⁇ are the parameters of the Pareto distribution model.
  • Formula 3 Average user experience rate during busy hours: Corresponding to the user experience rate distribution model 3, that is, according to the user experience rate distribution model 3, the average experience rate formula 3 for busy users can be obtained.
  • H (z; ⁇ , s) e- ⁇ z s F s (a 1 , ..., a s ; b 1 , ..., b s ; ⁇ z).
  • Is the average user experience rate value during busy hours
  • R is the downlink transmission rate of the base station
  • is TTI
  • s is the parameter of the Zeta distribution model
  • is the parameter of the exponential distribution model.
  • Formula 4 for the average user experience rate during busy hours corresponding to the user experience rate distribution model 4, that is, the formula 4 for the average user experience rate during busy hours can be obtained according to the user experience rate distribution model 4.
  • Is the average user experience rate during busy hours
  • R is the downlink transmission rate of the base station
  • is the TTI
  • s is the parameter of the Zeta distribution model
  • is the preset bandwidth utilization rate and 0 ⁇ 1
  • m and ⁇ are the Pareto distribution model Parameters.
  • the capacity planning device After calculating the busy user average experience rate value, the capacity planning device performs bandwidth control based on the busy user average experience rate value, specifically including: if the difference between the busy user average experience rate value and the base station downlink transmission rate is greater than the first The difference threshold increases the bandwidth. Or, if the difference between the busy user average experience rate value and the downlink transmission rate of the base station is less than the second difference threshold, the bandwidth is reduced.
  • the first difference threshold is 20M
  • the second difference threshold is -10M
  • the downlink transmission rate of the base station is 50M.
  • the fixed bandwidth value can also be increased based on the difference between the busy user average experience rate value and the downlink transmission rate of the base station.
  • the bandwidth needs to be reduced.
  • the reduction method may be to reduce the fixed bandwidth value, or it may be reduced according to the difference between the average value of the busy user experience rate and the downlink transmission rate of the base station.
  • the hourly traffic forecast value can also be obtained based on the average user experience rate value during busy hours
  • the unit is GB.
  • the bandwidth allocation method of the base station is determined, thereby performing bandwidth control. For example, when the difference between T prediction and ⁇ th is greater than the preset first flow difference threshold, the bandwidth is increased, and when the difference between T prediction and ⁇ th is less than the preset second flow difference threshold, the bandwidth is reduced.
  • the service measurement device in step 201 above obtains the number of service packets and the length of the service packets in each transmission time interval within a set duration, for example, for:
  • Method 1 The service measurement device periodically obtains the number of service packets and the length of the service packets in each transmission time interval within a set duration.
  • the service measuring device can measure and record the number of service packages and the length of service packages arriving in each TTI in real time, and then periodically fit the number of service packages and the length of service packages respectively to obtain a matching distribution model. Distribution parameters. Therefore, before fitting, it is necessary to obtain the number of recorded service packets and the length of the service packets in the TTI. Based on this method 1, the service measurement device periodically obtains the number of recorded service packages and the length of the service package in the TTI, and periodically fits the obtained number of service packages and the length of the service package.
  • the service measurement device acquires the number of service packets and the length of the service packets in each transmission time interval every 1 hour, And the distribution model is fitted according to the number of acquired service packages and the length of the service packages, to obtain the distribution parameters of the first distribution model and the distribution parameters of the second distribution model, respectively, and then report the first distribution model and the second distribution to the capacity planning device The distribution parameters of the model, the first distribution model, and the distribution parameters of the second distribution model.
  • Method 2 The service measurement device periodically obtains the number of service packets and the length of the service packets in each transmission time interval that meets the preset busy hour condition within a set duration.
  • the service measuring device can measure and record the number of service packages and the length of service packages arriving in each TTI in real time, and then periodically fit the number of service packages and the length of service packages respectively to obtain a matching distribution model. Distribution parameters. Therefore, before fitting, it is necessary to obtain the number of recorded service packets and the length of the service packets in the TTI that satisfy the busy-time condition. Based on this method 2, the service measurement device periodically obtains the number of recorded service packages and the length of the service package in the TTI that meets the busy hour condition, and periodically performs the number of service packages and the length of the service package acquired Fitting.
  • the service measurement device obtains the number of service packets and the number of service packets in each transmission time interval that meets the busy hour condition every 1 hour
  • the length of the service package, and the distribution model is fitted according to the number of obtained service packages and the length of the service package to obtain the distribution parameters of the first distribution model and the distribution parameters of the second distribution model, respectively, and then report the first distribution to the capacity planning device
  • the number of service packets and the length of the service packet in each transmission time interval that satisfies the busy hour condition within that hour are used, and according to the obtained
  • the number of service packages and the length of service packages fit the distribution model to obtain the distribution parameters of the first distribution model and the distribution parameters of the second distribution model, respectively, and then report the first distribution model, the second distribution model, and the first to the capacity planning device
  • the distribution parameters of the distribution model and the distribution parameters of the second distribution model are used, and according to the obtained.
  • the distribution model is fitted according to the number of acquired service packages and the length of the service packages, to obtain the distribution parameters of the first distribution model and the distribution parameters of the second distribution model respectively, and then report the first distribution model and the second distribution to the capacity planning device.
  • the 30 minutes and 25 minutes satisfying the busy hour condition in the above example may be continuous time or discontinuous accumulated time.
  • satisfying the busy hour condition may be, for example, if the length of the queue on the base station at time t (that is, the t-th transmission time interval) is greater than a preset queue threshold, it is determined that the busy hour condition is satisfied.
  • Solution 2 The capacity planning device determines the lower limit value of the user experience rate according to the user experience rate distribution model, the distribution parameter values of the first distribution model, the distribution parameter values of the second distribution model, and the preset user experience rate satisfaction.
  • the capacity planning device performs bandwidth control according to the lower limit of the user experience rate.
  • User experience rate lower limit formula 1 corresponds to user experience rate distribution model 1, that is, user experience rate lower limit formula 1 can be obtained according to user experience rate distribution model 1.
  • is the parameter of Poisson (Possion) distribution model
  • is the parameter of exponential distribution model
  • User experience rate lower limit formula 2 corresponds to user experience rate distribution model 2, that is, user experience rate lower limit formula 2 can be obtained according to user experience rate distribution model 2.
  • R is the downlink transmission rate of the base station
  • is TTI
  • E [S] is the expected value of the number of bits arriving in a TTI
  • is the parameter of the Poisson (Possion) distribution model
  • m and ⁇ are the parameters of the Pareto distribution model.
  • User experience rate lower limit formula 3 corresponds to user experience rate distribution model 3, that is, user experience rate lower limit formula 3 can be obtained according to user experience rate distribution model 3.
  • is TTI
  • E [S] is the expected value of the number of bits arriving in a TTI
  • ⁇ () is the Riemann function
  • R is the downlink transmission rate of the base station
  • p 0 is the probability that the number of arriving service packets is zero
  • s is the parameter of the Zeta distribution model
  • is the parameter of the exponential distribution model.
  • User experience rate lower limit formula 4 corresponds to the user experience rate distribution model 4, that is, the user experience rate lower limit formula 4 can be obtained according to the user experience rate distribution model 4.
  • R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to buffer service packets
  • is TTI
  • E [S] is the expected value of the number of bits arriving in a TTI
  • ⁇ () Is the Riemann function
  • p 0 is the probability that the number of arriving service packets is zero
  • s is the parameter of the Zeta distribution model
  • m and ⁇ are the parameters of the Pareto distribution model.
  • the preset user experience rate satisfaction degree ⁇ 95%, combined with the reported service packet length and the fitting parameters of the packet number, and the system parameters such as the base station downlink transmission rate and TTI duration, using the above four formulas, we can determine Solve the user experience rate lower limit R min .
  • the capacity planning apparatus performs bandwidth control according to the lower limit value of the user experience rate, including: if the difference between the lower limit value of the user experience rate and the downlink transmission rate of the base station is greater than the third difference threshold, increasing the bandwidth. Or, if the difference between the busy user average experience rate value and the downlink transmission rate of the base station is less than the fourth difference threshold, the bandwidth is reduced.
  • the first difference threshold is 20M
  • the second difference threshold is -10M
  • the downlink transmission rate of the base station is 50M.
  • the bandwidth value can also be increased according to the difference between the lower limit of the user experience rate and the downlink transmission rate of the base station.
  • the bandwidth needs to be reduced.
  • the reduction method may be to reduce the fixed bandwidth value, or it may be reduced according to the difference between the lower limit value of the user experience rate and the downlink transmission rate of the base station.
  • the hourly traffic prediction value can also be obtained according to the lower limit of the user experience rate 4 units are GB.
  • the bandwidth allocation method of the base station is determined, thereby performing bandwidth control. For example, when the difference between T prediction and ⁇ th is greater than the preset third flow difference threshold, the bandwidth is increased, and when the difference between T prediction and ⁇ th is less than the preset fourth flow difference threshold, the bandwidth is reduced.
  • the service measurement device in step 201 above obtains the number of service packets and the length of the service packets in each transmission time interval within a set duration, for example, for:
  • the service measuring device periodically acquires the number of service packets and the length of the service packets in each transmission time interval within a set time period.
  • the service measurement device acquires the number of service packets and the length of the service packets in each transmission time interval every 1 hour, And the distribution model is fitted according to the number of acquired service packages and the length of the service packages, to obtain the distribution parameters of the first distribution model and the distribution parameters of the second distribution model, respectively, and then report the first distribution model and the second distribution to the capacity planning device The distribution parameters of the model, the first distribution model, and the distribution parameters of the second distribution model.
  • the capacity planning device performs bandwidth control, for example, it can be implemented by the following method:
  • the measurement device sends a first notification message.
  • the first notification message includes a bandwidth control strategy, where the bandwidth control strategy may be to increase the bandwidth or decrease the bandwidth.
  • the service measurement device sends a first confirmation message for the first notification message to the capacity planning device.
  • the capacity planning device sends a second notification message to the service measurement device, where the second notification message includes the bandwidth value. Therefore, the service measurement device can perform bandwidth control according to the bandwidth value.
  • the base station may perform bandwidth control on the corresponding terminal according to the bandwidth value.
  • the bandwidth value here may be an increased or decreased relative bandwidth value, or a bandwidth value after the service measurement device controls the bandwidth of the terminal or the base station.
  • the service measurement device may also send a second response message to the second notification message to the capacity planning device.
  • the capacity planning device when the service measurement device is deployed on the terminal and the capacity planning device is deployed on the base station, in the above step B, the capacity planning device performs bandwidth control, for example, the following method may be implemented: the capacity planning device notifies the base station The bandwidth value of each terminal, where the bandwidth value may be a relative bandwidth value that is increased or decreased, or a bandwidth value after the service measurement device controls the bandwidth of the terminal or the base station. Then, the base station performs bandwidth control on the corresponding terminal according to the received bandwidth value.
  • the total network bandwidth can be evenly distributed among the base stations, and the network bandwidth of the base station can be evenly distributed among the terminals in a base station.
  • the above capacity planning method of this application can Performing bandwidth control on each terminal / base station makes it possible to perform corresponding bandwidth control according to the service busyness of each terminal, which helps to improve system efficiency and resource utilization efficiency.
  • the above solution of the present application has refined a simple and practical model of business arrival and business package length, so as to realize rapid learning of business characteristic parameters, and the complexity is lower than that of the existing technology.
  • the capacity planning device can directly reflect the user experience quality in the network based on the user experience rate distribution model, which provides a rigorous basis for the operator to improve the network service quality. Therefore, the capacity planning method of this application can effectively predict network traffic and plan bandwidth allocation reasonably.
  • FIG. 3 shows a possible exemplary block diagram of the device involved in the embodiment of the present invention.
  • the device 300 may exist in the form of software or hardware.
  • the device 300 may include an acquisition unit 301, a determination unit 302, and a communication unit 303.
  • the device 300 may further include a control unit 304.
  • the communication unit 303 may include a receiving unit and a sending unit.
  • the acquisition unit 301, the determination unit 302, and the control unit 304 may be integrated into a processing unit, which is used to control and manage the actions of the device 300.
  • the communication unit 303 is used to support communication between the device 300 and other network entities.
  • the processing unit may be a processor or a controller, for example, may be a general-purpose central processing unit (central processing unit, CPU).
  • CPU central processing unit
  • DSP digital signal processing
  • ASIC application specific integrated circuits
  • FPGA field programmable gate array
  • the processor may also be a combination of computing functions, for example, including one or more microprocessor combinations, DSP and microprocessor combinations, and so on.
  • the communication unit 304 may be a communication interface, a transceiver or a transceiver circuit, etc., wherein the communication interface is collectively referred to, and in a specific implementation, the communication interface may include multiple interfaces.
  • the device 300 may be the service measurement device in any of the foregoing embodiments, or may be a chip that can be used in the service measurement device.
  • the processing unit may be, for example, a processor
  • the communication unit 303 may be, for example, a transceiver, which includes RF circuit.
  • the processing unit may be, for example, a processor
  • the communication unit 303 may be, for example, an input / Output interface, pin or circuit, etc.
  • the obtaining unit 301 is configured to obtain the number of service packets and the length of the service packets in each transmission time interval within a set duration.
  • the determining unit 302 is configured to determine the distribution parameter value of the first distribution model that matches the number of service packages, and determine the distribution parameter value of the second distribution model that matches the length of the service packages.
  • the communication unit 303 is configured to send the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model to the capacity planning device.
  • the determining unit 302 is specifically configured to: fit at least two distribution models using the number of the service packages to obtain the fitting degree of each distribution model and the distribution parameter value of the distribution model Determining that the distribution model with the highest degree of fit is the first distribution model matching the number of the service packages, and determining that the distribution parameter value of the distribution model with the highest degree of fit is the distribution parameter value of the first distribution model.
  • the at least two distribution models include one or more of the following distribution models: Poisson distribution model and Zeta distribution model.
  • the determining unit 302 is specifically configured to: fit at least two distribution models using the length of the service package to obtain the fitting degree of each distribution model and the distribution parameter value of the distribution model Determining that the distribution model with the highest degree of fit is the second distribution model with a matching length of the service package, and determining that the distribution parameter value of the distribution model with the highest degree of fit is the distribution parameter value of the second distribution model.
  • the at least two distribution models include one or more of the following distribution models: exponential distribution model and Pareto distribution model.
  • the determining unit 302 is specifically used to:
  • the first preset distribution model is a Poisson distribution model, or a Zeta distribution model
  • the second preset distribution model is an exponential distribution model, or a Pareto distribution model.
  • the determining unit 302 is specifically configured to fit the Poisson distribution model using the number of service packages to obtain the first fitting degree and the distribution parameter value of the Poisson distribution model. Use the number of business packages to fit the Zeta distribution model to obtain the second fit and the distribution parameter values of the Zeta distribution model. If the first fitting degree is greater than the second fitting degree, the distribution model parameter of the Poisson distribution model is determined to be the distribution parameter value of the first distribution model whose number of service packages match, and the first distribution model is the Poisson distribution model.
  • the distribution parameter value of the Zeta distribution model is determined to be the value of the distribution parameter of the first distribution model that matches the number of service packages, and the first distribution model is the Zeta distribution model.
  • the determining unit 302 is specifically used to:
  • the distribution parameter value of the exponential distribution model is the distribution parameter value of the second distribution model whose length of the service packet matches, and the second distribution model is the exponential distribution model.
  • the distribution model parameters of the Pareto distribution model are the distribution model parameters of the second distribution model matching the length of the service package, and the second distribution model is the Pareto distribution model.
  • the communication unit 303 is specifically used to:
  • the communication unit 303 is further configured to: receive a first notification message from the capacity planning device, where the first notification message includes a bandwidth control strategy. Send a first confirmation message for the first notification message to the capacity planning device. A second notification message is received from the capacity planning device, and the second notification message includes the bandwidth value.
  • the control unit 304 is configured to perform bandwidth control according to the bandwidth value.
  • the obtaining unit 301 is specifically configured to:
  • the number of service packets and the length of the service packets in each transmission time interval satisfying the preset busy time condition within a set time period are periodically obtained.
  • the service measurement device is deployed on the terminal, and the capacity planning device is deployed on the base station.
  • the obtaining unit 301 is specifically configured to obtain the number of service packets of the terminal and the length of the service packets of the terminal within each transmission time interval within a set duration.
  • the service measurement device is deployed on the base station, and the capacity planning device is deployed on the mobile edge computing server.
  • the obtaining unit 301 is specifically used for:
  • the device shown in FIG. 3 is a service measurement device, for the specific beneficial effects of the capacity planning method used for execution, reference may be made to the related description in the foregoing method embodiments, and details are not repeated here. It can be understood that, the units in the embodiments of the present application may also be referred to as modules. The above units or modules can exist independently or can be integrated together.
  • FIG. 4 shows a possible exemplary block diagram of the device involved in the embodiment of the present invention.
  • the device 400 may exist in the form of software or hardware.
  • the device 400 may include a communication unit 401 and a control unit 403.
  • the apparatus 400 may further include a determining unit 402.
  • the communication unit 401 may include a receiving unit and a sending unit.
  • the determination unit 402 and the control unit 403 may be integrated into one processing unit, which is used to control and manage the actions of the device 400.
  • the communication unit 401 is used to support communication between the device 400 and other network entities.
  • the processing unit may be a processor or a controller, such as a CPU, general-purpose processor, DSP, ASIC, FPGA or other programmable logic devices, transistors Logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules and circuits described in conjunction with the disclosure of the present invention.
  • the processor may also be a combination of computing functions, for example, including one or more microprocessor combinations, DSP and microprocessor combinations, and so on.
  • the communication unit 401 may be a communication interface, a transceiver, or a transceiver circuit, etc., where the communication interface is collectively referred to, and in a specific implementation, the communication interface may include multiple interfaces.
  • the device 400 may be the capacity planning device in any of the foregoing embodiments, or may be a chip that can be used in the capacity planning device.
  • the processing unit may be, for example, a processor
  • the communication unit 401 may be, for example, a transceiver including a radio frequency circuit.
  • the processing unit may be, for example, a processor
  • the communication unit 401 may be, for example, an input / output interface, a tube Feet or circuits, etc.
  • the communication unit 401 is configured to receive the distribution parameter value of the first distribution model and the distribution parameter value of the second distribution model from the service measurement device, where the first distribution model and the second distribution model are respectively within the set duration acquired by the service measurement device
  • the distribution model matching the number of service packets in each transmission time interval and the distribution model matching the length of service packets.
  • the control unit 403 is configured to perform bandwidth control according to the first distribution model, the second distribution model, the distribution parameter value of the first distribution model, and the distribution parameter value of the second distribution model.
  • the determining unit 402 is configured to determine a user experience rate distribution model based on the first distribution model, the second distribution model, the base station downlink transmission rate, and the transmission time interval;
  • the control unit 403 is specifically configured to perform bandwidth control according to the user experience rate distribution model, the distribution parameter value of the first distribution model, the distribution parameter value of the second distribution model, and the service quality requirement parameter value.
  • the service quality requirement parameter value is a preset bandwidth utilization rate
  • the control unit 403 is specifically configured to: according to the user experience rate distribution model, the distribution parameter value of the first distribution model, and the second The distribution parameter value of the distribution model and the preset bandwidth utilization rate determine the average user experience rate value during busy hours. Perform bandwidth control based on the value of the average user experience rate during busy hours.
  • the first distribution model is the Zeta distribution model
  • the distribution model parameters of the first distribution model include s
  • the second distribution model is the Pareto distribution model
  • the distribution model parameters of the second distribution model include m and ⁇ .
  • the user experience rate distribution model is:
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the queue length on the base station at time t
  • the queue is used to buffer service packets
  • is the transmission time interval
  • ⁇ () is the Riemann function
  • E [S] is a transmission time
  • p 0 is the probability that the number of arriving service packets is zero.
  • the control unit 403 is used to determine the average value of the busy user experience rate according to the following formula:
  • control unit 403 is specifically configured to increase the bandwidth if the difference between the busy user average experience rate value and the downlink transmission rate of the base station is greater than the first difference threshold. Or, if the difference between the busy user average experience rate value and the downlink transmission rate of the base station is less than the second difference threshold, the bandwidth is reduced.
  • the service quality parameter value is a preset user experience rate satisfaction
  • the control unit 403 is specifically configured to: according to the user experience rate distribution model, the distribution parameter value of the first distribution model and the first The distribution parameter value of the two distribution model and the preset user experience rate satisfaction degree determine the lower limit value of the user experience rate. Perform bandwidth control based on the user experience rate lower limit.
  • the first distribution model is the Zeta distribution model
  • the distribution model parameters of the first distribution model include s
  • the second distribution model is the Pareto distribution model
  • the distribution model parameters of the second distribution model include m and ⁇ .
  • the user experience rate distribution model is:
  • Pr () is the user experience rate distribution model
  • x is the independent variable of the user experience rate
  • R U is the user experience rate
  • the user experience rate at time t is R is the downlink transmission rate of the base station
  • Q (t) is the length of the queue on the base station at time t
  • the queue is used to buffer service packets
  • is the transmission time interval
  • ⁇ () is the Riemann function
  • E [S] is a transmission time
  • p 0 is the probability that the number of arriving service packets is zero.
  • the control unit 403 is used to determine the lower limit value of the user experience rate according to the following formula:
  • R min is the lower limit of the user experience rate
  • is the preset user experience rate satisfaction.
  • control unit 403 is specifically configured to increase the bandwidth if the difference between the lower limit of the user experience rate and the downlink transmission rate of the base station is greater than the third difference threshold. Or, if the difference between the busy user average experience rate value and the downlink transmission rate of the base station is less than the fourth difference threshold, the bandwidth is reduced.
  • the communication unit 401 is further configured to: receive the identification information of the first distribution model and the identification information of the second distribution model from the service measurement device, the first distribution model The identification information of is used to identify the selected first distribution model, and the identification information of the second distribution model is used to identify the selected second distribution model.
  • the communication unit 401 is further configured to: receive identification information of a service arrival model from the service measurement device, and the identification information of the service arrival model is used to identify the selected first A distribution model and a service arrival model corresponding to the second distribution model.
  • the determining unit 402 is further configured to determine the selected first distribution model and the second distribution model according to the identification information of the service arrival model.
  • the device shown in FIG. 4 is a capacity planning device, for specific beneficial effects of the capacity planning method used for execution, reference may be made to the related description in the foregoing method embodiments, and details are not repeated here. It can be understood that, the units in the embodiments of the present application may also be referred to as modules. The above units or modules can exist independently or can be integrated together.
  • the device 500 may be a service measurement device or a capacity planning device in the embodiments of the present application, or may be a component that can be used for a service measurement device or a capacity planning device .
  • the device 500 includes a processor 502, a communication interface 503, and a memory 501.
  • the device 500 may further include a bus 504.
  • the communication interface 503, the processor 502, and the memory 501 may be connected to each other through a communication line 504;
  • the communication line 504 may be a peripheral component interconnection (PCI) bus or an extended industry standard architecture (extended industry standard architecture) , Referred to as EISA) bus.
  • PCI peripheral component interconnection
  • EISA extended industry standard architecture
  • the communication line 504 can be divided into an address bus, a data bus, and a control bus. For ease of representation, only a thick line is used in FIG. 5, but it does not mean that there is only one bus or one type of bus.
  • the processor 502 may be a CPU, a microprocessor, an ASIC, or one or more integrated circuits for controlling the execution of the program programs of the present application.
  • the communication interface 503 can be a device that uses any transceiver to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN) ), Wired access network, etc.
  • RAN radio access network
  • WLAN wireless local area networks
  • Wired access network etc.
  • the memory 501 may be read-only memory (ROM) or other types of static storage devices that can store static information and instructions, random access memory (random access memory, RAM), or other types that can store information and instructions
  • the dynamic storage device can also be an electrically erasable programmable read-only memory (electrically programmable server read-only memory (EEPROM), compact disc-read memory (CD-ROM) or other optical disk storage, Disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or can be used to carry or store desired program code in the form of instructions or data structures and Any other media accessed by the computer, but not limited to this.
  • the memory may exist independently and be connected to the processor through the communication line 504. The memory can also be integrated with the processor.
  • the memory 501 is used to store computer execution instructions for executing the solution of the present application, and the processor 502 controls execution.
  • the processor 502 is used to execute computer-executed instructions stored in the memory 501, so as to implement the capacity planning method provided by the foregoing embodiments of the present application.
  • the computer execution instructions in the embodiments of the present application may also be called application program codes, which are not specifically limited in the embodiments of the present application.
  • the computer program product includes one or more computer instructions.
  • the computer may be a general-purpose computer, a dedicated computer, a computer network, or other programmable devices.
  • the computer instructions may be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be from a website site, computer, server or data center Transmit to another website, computer, server or data center via wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.).
  • the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device including one or more available media integrated servers, data centers, and the like.
  • the usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, Solid State Disk (SSD)), or the like.
  • a magnetic medium for example, a floppy disk, a hard disk, a magnetic tape
  • an optical medium for example, a DVD
  • a semiconductor medium for example, Solid State Disk (SSD)
  • the various illustrative logic units and circuits described in the embodiments of the present application may be implemented by a general-purpose processor, a digital signal processor, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or other programmable logic devices. Discrete gate or transistor logic, discrete hardware components, or any combination of the above are designed to implement or operate the described functions.
  • the general-purpose processor may be a microprocessor, and optionally, the general-purpose processor may also be any conventional processor, controller, microcontroller, or state machine.
  • the processor may also be implemented by a combination of computing devices, such as a digital signal processor and a microprocessor, multiple microprocessors, one or more microprocessors combined with a digital signal processor core, or any other similar configuration achieve.
  • the steps of the method or algorithm described in the embodiments of the present application may be directly embedded in hardware, a software unit executed by a processor, or a combination of both.
  • the software unit may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium in the art.
  • the storage medium may be connected to the processor, so that the processor can read information from the storage medium and can write information to the storage medium.
  • the storage medium may also be integrated into the processor.
  • the processor and the storage medium may be provided in the ASIC, and the ASIC may be provided in the terminal.
  • the processor and the storage medium may also be provided in different components in the terminal.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operating steps are performed on the computer or other programmable device to produce computer-implemented processing, which is executed on the computer or other programmable device
  • the instructions provide steps for implementing the functions specified in one block or multiple blocks of the flowchart one flow or multiple flows and / or block diagrams.

Landscapes

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

Abstract

本申请提供一种容量规划方法及装置。该方法包括:根据设定时长内的每个传输时间间隔内的业务包的数量匹配分布模型,得到匹配的第一分布模型,以及根据业务包的长度匹配分布模型,得到第二分布模型,并根据第一分布模型、第二分布模型、第一分布模型的分布参数以及第二分布模型的分布参数,执行带宽控制。由于是根据匹配的分布模型及分布模型的分布参数执行的带宽控制,因而有助于实现为用户提供更为准确的容量规划。

Description

一种容量规划方法及装置
本申请要求在2018年11月16日提交中华人民共和国知识产权局、申请号为201811367720.6、发明名称为“一种容量规划方法及装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及移动通信技术领域,尤其涉及一种容量规划方法及装置。
背景技术
2014年,欧洲电信标准协会(The European Telecommunications Standards Institute,ETSI)提出移动边缘计算(Mobile Edge Computing,MEC)作为一种未来网络架构方案,通过在长期演进(Long Term Evolution,LTE)和第五代(5th generation,5G)移动网边缘(例如基站)提供信息技术(Information Technology,IT)服务环境和计算能力,将网络业务下沉到更接近用户的无线接入侧,从而减轻核心网的营运压力,并使得网络运营商可以改善用户体验。
MEC的实现建立在合理的网络容量规划基础之上,关键步骤是合理预测网络流量,既要保障用户体验,又要控制营运成本,这不但是学术界的研究热点,也是工业界迫切需要解决的问题。随着智能设备的普及和各种网络应用的发展,网络业务日趋复杂,从话音业务为主逐渐转变为视频、交互式娱乐等数据业务为主。现网统计数据显示其业务量呈现显著的突发特性,因此传统话音网络所服从的“平均效应”不再适用,急需新的网络业务模型来刻画。传统的平均性能指标不能真实反映网络性能,现网统计数据表明平均性能指标(例如小时级平均)经常会掩盖忙时资源不足和忙时用户体验差,因此需要我们引入新的衡量指标来直接反映用户体验速率,进而准确地进行用户的容量规划。
发明内容
本申请提供一种容量规划方法及装置,用以提升用户的容量规划准确率。
第一方面,本申请提供一种容量规划方法,包括:容量规划装置从业务测量装置接收第一分布模型的分布参数值和第二分布模型的分布参数值,第一分布模型和第二分布模型分别为业务测量装置获取到的设定时长内的每个传输时间间隔内的业务包的数量所匹配的分布模型和业务包的长度所匹配的分布模型。容量规划装置根据所述第一分布模型、所述第二分布模型、所述第一分布模型的分布参数值和所述第二分布模型的分布参数值,执行带宽控制。基于该方案,根据设定时长内的每个传输时间间隔内的业务包的数量匹配分布模型,得到匹配的第一分布模型,以及根据业务包的长度匹配分布模型,得到第二分布模型,并根据第一分布模型、第二分布模型、第一分布模型的分布参数以及第二分布模型的分布参数,执行带宽控制。由于是根据匹配的分布模型及分布模型的分布参数执行的带宽控制,因而有助于实现为用户提供更为准确 的容量规划。
在一种可能的实现方法中,容量规划装置根据所述第一分布模型、所述第二分布模型、基站下行传输速率和所述传输时间间隔,确定用户体验速率分布模型。容量规划装置根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制。基于该方案,引入了用户体验速率分布模型,且该用户体验速率分布模型式是根据设定时长内的每个传输时间间隔内的业务包的数量匹配的第一分布模型、业务包的长度匹配的第二分布模型、基站下行传输速率以及时间间隔的长度确定的,由于采用了传输时间间隔粒度的业务包的数量的分布模型和业务包的长度的分布模型,且业务包的长度是以比特为单位的,本方案可以得到每个传输时间间隔的业务总比特数的分布模型,从而提供了更为精细的比特级粒度,因此确定的用户体验速率分布模型可以更为精确地刻画用户的体验速率,进而可以实现为用户提供更为准确的容量规划。
在一种可能的实现方式中,当业务测量装置部署于终端上,容量规划装置部署于基站上,则上述基站下行传输速率为该基站针对该终端的下行传输速率。
在一种可能的实现方式中,当业务测量装置部署于基站上,容量规划装置部署于移动边缘计算服务器上,则上述基站下行传输速率为该基站针对该接入到该基站的所有的终端的下行传输速率的总和。
在一种可能的实现方法中,所述服务质量要求参数值为预设的带宽利用率;容量规划装置根据用户体验速率分布模型、第一分布模型的分布参数值、第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制,包括:容量规划装置根据用户体验速率分布模型、第一分布模型的分布参数值和第二分布模型的分布参数值和预设的带宽利用率,确定忙时用户平均体验速率值。容量规划装置根据忙时用户平均体验速率值,执行带宽控制。基于该方案,引入了忙时用户平均体验速率值,从而可以基于该参数,更为精细地刻画用户在忙时的体验速率,以便于更为准确地为用户提供容量规划。
在一种可能的实现方法中,第一分布模型为Zeta分布模型,第一分布模型的分布模型参数包括s,第二分布模型为Pareto分布模型,第二分布模型的分布模型参数包括m和α。
用户体验速率分布模型为:
Figure PCTCN2019116670-appb-000001
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量,R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000002
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为传输时间间隔,ζ()为黎曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数 量为零的概率。
容量规划装置根据下列公式确定忙时用户平均体验速率值:
Figure PCTCN2019116670-appb-000003
其中,
Figure PCTCN2019116670-appb-000004
为忙时用户平均体验速率值,ε为预设的带宽利用率且0≤ε≤1。
在一种可能的实现方法中,容量规划装置根据忙时用户平均体验速率值,执行带宽控制,包括:若忙时用户平均体验速率值与基站下行传输速率的差值大于第一差值阈值,则增加带宽。或者,若忙时用户平均体验速率值与基站下行传输速率的差值小于第二差值阈值,则减少带宽。
在又一种可能的实现方法中,所述服务质量参数值为预设的用户体验速率满足度;容量规划装置根据用户体验速率分布模型、第一分布模型的分布参数值、第二分布模型的分布参数值和服务质量参数值,执行带宽控制,包括:容量规划装置根据用户体验速率分布模型、第一分布模型的分布参数值、第二分布模型的分布参数值和预设的用户体验速率满足度,确定用户体验速率下限值。容量规划装置根据用户体验速率下限值,执行带宽控制。基于该方案,引入了用户体验速率下限值,从而可以基于该参数,更为精细地刻画用户的体验速率,以便于更为准确地为用户提供容量规划。
在一种可能的实现方法中,第一分布模型为Zeta分布模型,第一分布模型的分布模型参数包括s,第二分布模型为Pareto分布模型,第二分布模型的分布模型参数包括m和α。
用户体验速率分布模型为:
Figure PCTCN2019116670-appb-000005
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量,R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000006
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为传输时间间隔,ζ()为黎曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数量为零的概率。
容量规划装置根据下列公式确定用户体验速率下限值:
Figure PCTCN2019116670-appb-000007
其中,R min为用户体验速率下限值,η为预设的用户体验速率满足度。
在一种可能的实现方法中,容量规划装置根据用户体验速率下限值,执行带宽控 制,包括:若用户体验速率下限值与基站下行传输速率的差值大于第三差值阈值,则增加带宽。或者,若忙时用户平均体验速率值与基站下行传输速率的差值小于第四差值阈值,则减少带宽。
在一种可能的实现方法中,所述容量规划装置还从所述业务测量装置接收所述第一分布模型的标识信息和第二分布模型的标识信息,所述第一分布模型的标识信息用于标识选择的所述第一分布模型,所述第二分布模型的标识信息用于标识选择的所述第二分布模型。
在又一种可能的实现方法中,所述容量规划装置还从所述业务测量装置接收业务到达模型的标识信息,所述业务到达模型的标识信息用于标识选择的所述第一分布模型和所述第二分布模型所对应的业务到达模型。所述容量规划装置根据所述业务到达模型的标识信息,确定选择的所述第一分布模型和所述第二分布模型。
第二方面,本申请提供一种容量规划方法,该方法包括:业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和业务包的长度。业务测量装置确定与业务包的数量匹配的第一分布模型的分布参数值,以及确定与业务包的长度匹配的第二分布模型的分布参数值。业务测量装置向容量规划装置发送第一分布模型的分布参数值和第二分布模型的分布参数值,第一分布模型、第二分布模型、基站下行传输速率和所述传输时间间隔用于确定用户体验速率分布模型,进一步的,用户体验速率分布模型、第一分布模型的参数值、第二分布模型的参数值和服务质量要求参数值可用于执行带宽控制。该方案引入了用户体验速率分布模型,且该用户体验速率分布模型式是根据设定时长内的每个传输时间间隔内的业务包的数量匹配的第一分布模型、业务包的长度匹配的第二分布模型基站下行传输速率以及时间间隔的长度确定的,由于采用了传输时间间隔粒度的业务包的数量的分布模型和业务包的长度的分布模型,且业务包的长度是以比特为单位的,本方案可以得到每个传输时间间隔的业务总比特数的分布模型,从而提供了更为精细的比特级粒度,因此确定的用户体验速率分布模型可以更为精确地刻画用户的体验速率,进而可以实现为用户提供更为准确的容量规划。
在一种可能的实现方法中,所述业务测量装置确定与所述业务包的数量匹配的第一分布模型的分布参数值,包括:所述业务测量装置使用所述业务包的数量拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;所述业务测量装置确定拟合度最高的分布模型为所述业务包的数量匹配的第一分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第一分布模型的分布参数值。
可选的,所述至少两个分布模型包括以下分布模型中的一种或多种:泊松分布模型、Zeta分布模型。
例如,在具体实现中,业务测量装置使用业务包的数量拟合泊松分布模型,得到第一拟合度和泊松分布模型的分布参数值。业务测量装置使用业务包的数量拟合Zeta分布模型,得到第二拟合度和Zeta分布模型的分布参数值。若第一拟合度大于第二拟合度,则业务测量装置确定泊松分布模型的分布模型参数为业务包的数量匹配的第一分布模型的分布参数值,第一分布模型为泊松分布模型。或者,若第一拟合度不大于第二拟合度,则业务测量装置确定Zeta分布模型的分布参数值为业务包的数量匹配的第一分布模型的分布参数值,第一分布模型为Zeta分布模型。其中,Zeta分布模型是重尾分布模型,可以准确地体现用户的突发业务流量。
在一种可能的实现方法中,所述业务测量装置确定与所述业务包的长度匹配的第二分布模型的分布参数值,包括:所述业务测量装置使用所述业务包的长度拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;所述业务测量装置确定拟合度最高的分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第二分布模型的分布参数值。
可选的,所述至少两个分布模型包括以下分布模型中的一种或多种:指数分布模型、Pareto分布模型。
例如,在具体实现中,业务测量装置使用业务包的长度拟合指数分布模型,得到第三拟合度和指数分布模型的分布参数值。业务测量装置使用业务包的长度拟合Pareto分布模型,得到第四拟合度和Pareto分布模型的分布参数值。若第三拟合度大于第四拟合度,则业务测量装置确定指数分布模型的分布参数值为业务包的长度匹配的第二分布模型的分布参数值,第二分布模型为指数分布模型。或者,若第三拟合度不大于第四拟合度,则业务测量装置确定Pareto分布模型的分布模型参数为业务包的长度匹配的第二分布模型的分布模型参数,第二分布模型为Pareto分布模型。其中,Pareto分布模型是重尾分布模型,可以准确地体现用户的突发业务流量。
在又一种可能的实现方法中,所述业务测量装置确定与所述业务包的数量匹配的第一分布模型的分布参数值,包括:所述业务测量装置拟合第一预设分布模型,得到所述第一预设分布模型的分布参数值;所述业务测量装置确定所述第一预设分布模型为所述业务包的数量匹配的第一分布模型,以及确定所述第一预设分布模型的分布参数值为所述第一分布模型的分布参数值。所述业务测量装置确定与所述业务包的长度匹配的第二分布模型的分布参数值,包括:所述业务测量装置拟合第二预设分布模型,得到所述第二预设分布模型的分布参数值;所述业务测量装置确定所述第二预设分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述第二预设分布模型的分布参数值为所述第二分布模型的分布参数值。
可选的,所述第一预设分布模型为泊松分布模型、或Zeta分布模型,所述第二预设分布模型为指数分布模型、或Pareto分布模型。
在一种可能的实现方法中,业务测量装置向容量规划装置发送第一分布模型的分布参数值和第二分布模型的分布参数值,包括:业务测量装置向容量规划装置发送第一上报消息,第一上报消息包括第一分布模型的标识和第二分布模型的标识。业务测量装置从容量规划装置接收针对第一上报消息的第一响应消息。业务测量装置向容量规划装置发送第二上报消息,第二上报消息包括第一分布模型的分布参数值和第二分布模型的分布参数值。
在一种可能的实现方法中,业务测量装置从容量规划装置接收第一通知消息,第一通知消息包括带宽的控制策略。业务测量装置向容量规划装置发送针对第一通知消息的第一确认消息。业务测量装置从容量规划装置接收第二通知消息,第二通知消息包括带宽值。业务测量装置根据带宽值,执行带宽控制。
在一种可能的实现方法中,业务测量装置部署于终端上,容量规划装置部署于基站上。业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和业务包的长度,包括:业务测量装置获取设定时长内的每个传输时间间隔内的终端的业务包的数量和终端的业务包的长度。
在又一种可能的实现方法中,业务测量装置部署于基站上,容量规划装置部署于移动边缘计算服务器上。业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和业务包的长度,包括:业务测量装置获取设定时长内的每个传输时间间隔内的接入到基站的各个终端的业务包的数量和各个终端的业务包的长度。或者,业务测量装置获取设定时长内的每个传输时间间隔内的基站的业务包的数量和基站的业务包的长度。
在一种可能的实现方法中,所述业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度,包括:所述业务测量装置周期性地获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度;或者,所述业务测量装置周期性地获取设定时长内满足预设的忙时条件的每个传输时间间隔内的业务包的数量和所述业务包的长度。
第三方面,本申请提供一种通信装置,该通信装置具有实现上述方法实施例中业务测量装置或容量规划装置的功能。该功能可以通过硬件实现,也可以通过硬件执行相应的软件实现。该硬件或软件包括一个或多个与上述功能相对应的单元或者模块。
在一种可能的设计中,该通信装置包括:处理器、存储器、总线和通信接口;该存储器存储有计算机执行指令,该处理器与该存储器通过该总线连接,当该通信装置运行时,该处理器执行该存储器存储的该计算机执行指令,以使该通信装置执行如上述第一方面、或执行上述第一方面的任一实现方式中的容量规划方法。例如,该通信装置可以是容量规划装置。
在另一种可能的设计中,该通信装置还可以是芯片,如用于业务测量装置的芯片、若用于容量规划装置的芯片,该芯片包括处理单元,可选地,还包括存储单元,该芯片可用于执行如上述第一方面、或执行上述第一方面的任一实现方式中的容量规划方法中的容量规划方法。
第四方面,本申请提供一种通信装置,该通信装置具有实现上述方法实施例中业务测量装置或容量规划装置的功能。该功能可以通过硬件实现,也可以通过硬件执行相应的软件实现。该硬件或软件包括一个或多个与上述功能相对应的单元或者模块。
在一种可能的设计中,该通信装置包括:处理器、存储器、总线和通信接口;该存储器存储有计算机执行指令,该处理器与该存储器通过该总线连接,当该通信装置运行时,该处理器执行该存储器存储的该计算机执行指令,以使该通信装置执行如上述第二方面、或执行上述第二方面的任一实现方式中的容量规划方法。例如,该通信装置可以是业务测量装置。
在另一种可能的设计中,该通信装置还可以是芯片,如用于业务测量装置的芯片、若用于容量规划装置的芯片,该芯片包括处理单元,可选地,还包括存储单元,该芯片可用于执行如上述第二方面、或执行上述第二方面的任一实现方式中的容量规划方法中的容量规划方法。
第五方面,本申请提供了一种计算机存储介质,储存有为上述终端所用的计算机软件指令,其包含用于为执行上述第一方面、或第一方面的任一实现方式所设计的程序。
第六方面,本申请提供了一种计算机存储介质,储存有为上述终端所用的计算机软件指令,其包含用于为执行上述第一方面、或第一方面的任一实现方式所设计的程 序。
第七方面,本申请提供了一种计算机程序产品。该计算机程序产品包括计算机软件指令,该计算机软件指令可通过处理器进行加载来实现上述第一方面或第一方面中任意一项的容量规划方法中的流程。
第八方面,本申请提供了一种计算机程序产品。该计算机程序产品包括计算机软件指令,该计算机软件指令可通过处理器进行加载来实现上述第二方面或第二方面中任意一项的容量规划方法中的流程。
第九方面,本申请提供一种系统,包括上述任一方面所述的业务测量装置和上述任一方面所述的容量规划装置。
附图说明
图1为本申请提供的一种可能的网络架构示意图;
图2为本申请提供的一种容量规划方法流程图;
图3为本申请提供的一种装置示意图;
图4为本申请提供的又一种装置示意图;
图5为本申请提供的又一种装置示意图。
具体实施方式
为了使本申请的目的、技术方案和优点更加清楚,下面将结合附图对本申请作进一步地详细描述。方法实施例中的具体操作方法也可以应用于装置实施例或系统实施例中。其中,在本申请的描述中,除非另有说明,“多个”的含义是两个或两个以上。
如图1所示,为本申请所适用的网络架构示意图。该网络架构为MEC网络架构。在MEC架构中,每个网络边缘节点(即基站)为其服务范围内的用户(即终端)提供数据服务,若干个基站连接至基站侧的汇聚节点,汇聚节点连接至MEC服务器。MEC服务器位于无线接入点和核心网之间,有存储和计算的能力。通过在基站侧叠加MEC服务器,可以为用户提供本地化的服务,从而有效节省核心网侧的系统资源,同时显著缩短相应的响应时间。
本申请所提出的容量规划方法由业务测量装置和容量规划装置共同完成。业务测量装置负责测量数据业务并学习业务特征参数,容量规划装置负责根据各个业务测量装置上报的业务特征参数,进行整体容量规划。在一种实现方式中,业务测量装置部署于基站上,容量规划装置部署于MEC服务器上。在又一种实现方式中,业务测量装置部署于终端上,容量规划装置部署于基站上。
本申请中,终端是一种具有无线收发功能的设备,终端可以部署在陆地上,包括室内或室外、手持或车载;也可以部署在水面上(如轮船等);还可以部署在空中(例如飞机、气球和卫星上等)。所述终端可以是手机(mobile phone)、平板电脑(pad)、带无线收发功能的电脑、虚拟现实(virtual reality,VR)终端、增强现实(augmented reality,AR)终端、工业控制(industrial control)中的无线终端、无人驾驶(self driving)中的无线终端、远程医疗(remote medical)中的无线终端、智能电网(smart grid)中的无线终端、运输安全(transportation safety)中的无线终端、智慧城市(smart city) 中的无线终端、智慧家庭(smart home)中的无线终端,以及还可以包括用户设备(user equipment,UE)等。终端还可以是蜂窝电话、无绳电话、会话启动协议(session initiation protocol,SIP)电话、无线本地环路(wireless local loop,WLL)站、个人数字助理(personal digital assistant,PDA)、具有无线通信功能的手持设备、计算设备或连接到无线调制解调器的其它处理设备、车载设备、可穿戴设备,未来5G网络中的终端设备或者未来演进的公用陆地移动通信网络(public land mobile network,PLMN)中的终端设备等。终端有时也可以称为终端设备、用户设备(user equipment,UE)、接入终端设备、车载终端、工业控制终端、UE单元、UE站、移动站、移动台、远方站、远程终端设备、移动设备、UE终端设备、终端设备、无线通信设备、UE代理或UE装置等。终端也可以是固定的或者移动的。本申请实施例对此并不限定。
基站,是一种为终端提供无线通信功能的设备。基站例如包括但不限于:5G中的下一代基站(g nodeB,gNB)、演进型节点B(evolved node B,eNB)、无线网络控制器(radio network controller,RNC)、节点B(node B,NB)、基站控制器(base station controller,BSC)、基站收发台(base transceiver station,BTS)、家庭基站(例如,home evolved nodeB,或home node B,HNB)、基带单元(baseBand unit,BBU)、传输点(transmitting and receiving point,TRP)、发射点(transmitting point,TP)、移动交换中心等。
在现有的网络性能分析方法中,主要是基于传统排队论和它的一些变式,将业务包作为最小排队元素,从而计算网络的平均队列长度、等待时延和传输速率等性能指标,从而提供容量评估和规划。现有技术广泛采用经典的泊松、指数等模型来建模网络业务到达,例如业务达到服从泊松分布模型,业务持续时长服从指数分布模型。这些模型均属于轻尾模型,较容易分析推导其数学性质,并且被证明对传统话语业务和低速数据业务有良好的近似。
上述经典的轻尾业务模型具有“平均效应”,即平均业务量具有很强代表性,随机到达业务量与平均业务量的值偏差小,即尾概率衰减速率不慢于指数衰减。但是这种分布模型模式与现网中观察到的现象经常相差较大。通过多处现网场景的数据测量与分析,发现当前的数据网络中到达的业务量经常呈现严重突发,业务量概率分布模型的拖尾衰减慢于指数分布模型,呈现“重尾”,使得平均业务量不具代表性。因此现有业务模型无法刻画高速复杂的数据业务,存在应用局限。另一方面,在传输数据时,LTE系统和5G系统的基站的服务器以单个传输时间间隔(Transmission Time Interval,TTI)为时域上最小的调度单元,一个业务包可能在相邻的TTI之间被切分。这与以包为最小单位的经典排队论的应用条件不相符,因此传统排队论无法适用于LTE和5G网络,需要从更为精细的比特级进行分析。
综上所述,现有的容量规划方法假定业务模型服从轻尾分布模型,但是现网实测数据表明到达的业务量经常存在突发特性,平均网络性能经常不具代表性,与忙时网络性能差别显著,导致现有轻尾模型不再适用,为此,本申请建立了新的模型。同时,现有容量规划方法基于以包为单位的经典排队论,与LTE和5G的基站的服务器采用的调度方式不相符,为此,本申请以比特为单位进行计算。
本申请针对现有的网络的数据业务量特征,提出能够真实反映用户体验的性能指标,给出网络流量预测,解决网络容量规划问题。
如图2所示,为本申请提供的一种容量规划方法,该方法包括以下步骤:
步骤201,业务测量装置获取设定时长内的每个TTI内的业务包的数量和业务包的长度。
其中,业务包的长度的单位为比特。比如,TTI1内的业务包包括业务包1、业务包2和业务包3,且长度分别为100比特,200比特和300比特,则TTI1内的业务包的数量为3,业务包的长度分别为100比特,200比特,300比特。
在一种实现方式中,若业务测量装置部署于终端上,容量规划装置部署于基站上,每个业务测量装置获取设定时长内的每个TTI内的终端的业务包的数量和终端的业务包的长度。例如,终端1上的业务测量装置获取到设定时长内每个TTI内的终端1的业务包的数量和该TTI内的终端1的业务包的长度,终端2上的业务测量装置获取到设定时长内每个TTI内的终端2的业务包的数量和该TTI内的终端2的业务包的长度,等等。
在又一种实现方式中,若业务测量装置部署于基站上,容量规划装置部署于MEC服务器上,则在一种实现方式中,业务测量装置可以是以终端粒度获取业务包的数量和长度,具体的,每个业务测量装置获取设定时长内的每个TTI内的接入到基站的各个终端的业务包的数量和各个终端的业务包的长度。例如,接入到基站1的终端包括终端1、终端2和终端3,则基站1上的业务测量装置获取到设定时长内每个TTI内的终端1的业务包的数量和该TTI内的终端1的业务包的长度,获取到设定时长内每个TTI内的终端2的业务包的数量和该TTI内的终端2的业务包的长度,以及获取到设定时长内每个TTI内的终端3的业务包的数量和该TTI内的终端3的业务包的长度。在又一种实现方式中,业务测量装置还可以是以基站粒度获取业务包的数量和长度,具体的,每个业务测量装置获取设定时长内的每个TTI内的基站的业务包的数量和基站的业务包的长度,这里的每个TTI内的基站的业务包的数量指的是每个TTI内的接入到该基站的所有终端的业务包的数量总和,这里的每个TTI内的基站的业务包的长度指的是每个TTI内的接入到该基站的所有终端的业务包的长度总和。例如,接入到基站1的终端包括终端1和终端2,则基站1上的业务测量装置获取到设定时长内每个TTI内的基站1的业务包的数量和该TTI内的基站1的业务包的长度,这里的每个TTI内的基站1的业务包的数量指的是每个TTI内的终端1和终端2的业务包的数量总和,这里的每个TTI内的基站1的业务包的长度指的是每个TTI内的终端1和终端2的业务包的长度总和。
步骤202,业务测量装置确定与业务包的数量匹配的第一分布模型的分布参数值,以及确定与业务包的长度匹配的第二分布模型的分布参数值。
本申请中,分布模型也可以称为分布函数、或者简称为分布。
业务测量装置可以根据以下方法确定与业务包的数量匹配的第一分布模型的分布参数值:业务测量装置使用业务包的数量拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值,然后业务测量装置确定拟合度最高的分布模型为业务包的数量匹配的第一分布模型,以及确定拟合度最高的分布模型的分布参数值为所述第一分布模型的分布参数值。可选的,所述至少两个分布模型包括以下分布模型中的一种或多种:泊松分布模型、Zeta分布模型。
下面,以所述至少两个分布模型包括泊松分布模型、Zeta分布模型为例,说明业 务测量装置确定与业务包的数量匹配的第一分布模型的分布参数值的具体实现方法。
比如,业务测量装置使用获取到的设定时长内的各个TTI内的业务包的数量拟合泊松分布模型,得到第一拟合度和泊松分布模型的分布参数值。业务测量装置使用业务包的数量拟合Zeta分布模型,得到第二拟合度和Zeta分布模型的分布参数值。若第一拟合度大于第二拟合度,则业务测量装置确定泊松分布模型的分布模型参数为业务包的数量匹配的第一分布模型的分布参数值,第一分布模型为泊松分布模型。或者,若第一拟合度不大于第二拟合度,则业务测量装置确定Zeta分布模型的分布参数值为业务包的数量匹配的第一分布模型的分布参数值,第一分布模型为Zeta分布模型。
泊松(Possion)分布模型以λ>0为参数,其形式为:
Figure PCTCN2019116670-appb-000008
Zeta分布模型以s>0为参数,ζ(s)表示黎曼Zeta函数,p 0表示到达的业务包的数量为零的概率,其形式为:
Figure PCTCN2019116670-appb-000009
其中,若业务测量装置确定业务包的数量匹配的第一分布模型为泊松分布模型,则确定的第一分布模型的分布参数值为λ的值。若业务测量装置确定业务包的数量匹配的第一分布模型为Zeta分布模型,则确定的第一分布模型的分布参数值为s的值。
上述方法,通过从多个分布模型中,选择一个与业务包的数量拟合度最高的分布模型作为使用的第一分布模型,有助于提升容量规划的准确率。
业务测量装置可以根据以下方法确定与业务包的长度匹配的第二分布模型的分布参数值:业务测量装置使用所述业务包的长度拟合至少两个分布模型,得到每个分布模型的拟合度和分布模型的分布参数值;然后,业务测量装置确定拟合度最高的分布模型为业务包的长度匹配的第二分布模型,以及确定拟合度最高的分布模型的分布参数值为第二分布模型的分布参数值。可选的,所述至少两个分布模型包括以下分布模型中的一种或多种:指数分布模型、Pareto分布模型。
下面,以所述至少两个分布模型包括指数分布模型、Pareto分布模型为例,说明业务测量装置确定与业务包的数量匹配的第一分布模型的分布参数值的具体实现方法。
比如,业务测量装置使用业务包的长度拟合指数分布模型,得到第三拟合度和指数分布模型的分布参数值。业务测量装置使用业务包的长度拟合Pareto分布模型,得到第四拟合度和Pareto分布模型的分布参数值。若第三拟合度大于第四拟合度,则业务测量装置确定指数分布模型的分布参数值为业务包的长度匹配的第二分布模型的分布参数值,第二分布模型为指数分布模型。或者,若第三拟合度不大于第四拟合度,则业务测量装置确定Pareto分布模型的分布模型参数为业务包的长度匹配的第二分布模型的分布模型参数,第二分布模型为Pareto分布模型。
指数(Exponential)分布模型以θ>0为参数,其形式为:
Figure PCTCN2019116670-appb-000010
Pareto分布模型以m>0和α>0为参数,其形式为:
Figure PCTCN2019116670-appb-000011
其中,若业务测量装置确定业务包的长度匹配的第二分布模型为指数分布模型,则确定的指数分布模型的分布参数值为θ的值。若业务测量装置确定业务包的长度匹配的第二分布模型为Pareto分布模型,则确定的第二分布模型的分布参数值为m和α的值。
上述方法,通过从多个分布模型中,选择一个与业务包的长度拟合度最高的分布模型作为使用的第二分布模型,有助于提升容量规划的准确率。
在又一种实现方式中,业务测量装置还可以根据以下方法确定与业务包的数量匹配的第一分布模型的分布参数值,以及确定与业务包的长度匹配的第一分布模型的分布参数值:
业务测量装置拟合第一预设分布模型,得到第一预设分布模型的分布参数值;业务测量装置确定第一预设分布模型为所述业务包的数量匹配的第一分布模型,以及确定第一预设分布模型的分布参数值为所述第一分布模型的分布参数值。
业务测量装置拟合第二预设分布模型,得到第二预设分布模型的分布参数值;业务测量装置确定所述第二预设分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述第二预设分布模型的分布参数值为所述第二分布模型的分布参数值。
可选的,所述第一预设分布模型为泊松分布模型、或Zeta分布模型,所述第二预设分布模型为指数分布模型、或Pareto分布模型。
该实现方式中,业务测量装置使用第一预设分布模型作为业务包的数量匹配的第一分布模型,以及使用第二预设分布模型作为业务包的长度匹配的第二分布模型。由于不需要从多个分布模型中选择分布模型,因而可以提升容量规划的效率。
上述Zeta分布模型和Pareto分布模型分别代表离散和连续的幂律(重尾)分布模型。
业务包的数量匹配的分布模型和业务包的长度匹配的分布模型可以构成如下表1所示的四种业务到达模型。
表1 四种业务到达模型与典型应用场景
Figure PCTCN2019116670-appb-000012
通过上述表1可以看出,当业务包的数量匹配泊松分布模型、业务包的长度匹配 指数分布模型时,则可以认为业务符合PE模型,一般针对的场景是:单话音。
当业务包的数量匹配泊松分布模型、业务包的长度匹配Pareto分布模型时,则可以认为业务符合PP模型,一般针对的场景是:宏蜂窝,偏僻地区。
当业务包的数量匹配Zeta分布模型、业务包的长度匹配指数分布模型时,则可以认为业务符合ZE模型,一般针对的场景是:商业区闲时。
当业务包的数量匹配Zeta分布模型、业务包的长度匹配Pareto分布模型时,则可以认为业务符合PP模型,一般针对的场景是:商业区忙时。
需要说明的是,其中,本申请对于使用的“拟合”的方法不限,例如可以是使用最小二乘法进行拟合等。
步骤203,业务测量装置向容量规划装置发送第一分布模型的分布参数值和第二分布模型的分布参数值。
在又一种实现中,业务测量装置向容量规划装置发送第一分布模型的标识信息、第二分布模型的标识信息、第一分布模型的分布参数值和第二分布模型的分布参数值。其中,所述第一分布模型的标识信息用于标识选择的所述第一分布模型,所述第二分布模型的标识信息用于标识选择的所述第二分布模型。基于该实现方法,容量规划装置可以获知业务测量装置选择的第一分布模型和第二分布模型,以及获知第一分布模型的分布参数值和第二分布模型的分布参数值。
在又一种实现中,业务测量装置向容量规划装置发送业务到达模型的标识信息,所述业务到达模型的标识信息用于标识选择的所述第一分布模型和所述第二分布模型所对应的业务到达模型。进而容量规划装置可以根据所述业务到达模型的标识信息,确定选择的所述第一分布模型和所述第二分布模型。比如,业务测量装置向容量规划装置发送业务到达模型的标识信息指示为上述表1的PE模型,则业务测量装置可以确定选择的所述第一分布模型和所述第二分布模型分别为泊松分布模型和指数分布模型;再比如,业务测量装置向容量规划装置发送业务到达模型的标识信息指示为上述表1的PP模型,则业务测量装置可以确定选择的所述第一分布模型和所述第二分布模型分别为泊松分布模型和Pareto分布模型,等等。
在一种实现方式中,业务测量装置可以向容量规划装置发送第一上报消息,第一上报消息包括第一分布模型的标识和第二分布模型的标识。从而,容量规划装置可以获知业务测量装置确定的匹配的分布模型的类型,以便于预留相应的资源。接着,容量规划装置向业务测量装置发送针对第一上报消息的第一响应消息。然后,业务测量装置向容量规划装置发送第二上报消息,第二上报消息包括第一分布模型的分布参数值和第二分布模型的分布参数值。可选的,容量规划装置还可以向业务测量装置发送针对第二上报消息的第二响应消息。基于该方法,可以实现向容量规划装置发送第一分布模型的分布参数值和第二分布模型的分布参数值。
作为一种实现方式,上述第一上报消息包括第一分布模型的标识和第二分布模型的标识,可以替换为,第一上报消息包括模型标识,如“00”指示PE模型,“01”指示PP模型,“10”指示ZE模型,“11”指示ZP模型,从而也可以实现向容量规划装置发送第一分布模型和第二分布模型的类型。
步骤204,容量规划装置根据第一分布模型、第二分布模型、第一分布模型的分布参数和第二分布模型的分布参数,执行带宽控制。
基于图1所示的实施例的方案,根据设定时长内的每个传输时间间隔内的业务包的数量匹配分布模型,得到匹配的第一分布模型,以及根据业务包的长度匹配分布模型,得到第二分布模型,并根据第一分布模型、第二分布模型、第一分布模型的分布参数以及第二分布模型的分布参数,执行带宽控制。由于是根据匹配的分布模型及分布模型的分布参数执行的带宽控制,因而有助于实现为用户提供更为准确的容量规划。
作为示例,下面给出上述步骤204的一种具体实现方式。
容量规划装置根据第一分布模型、第二分布模型、第一分布模型的分布参数和第二分布模型的分布参数,执行带宽控制,具体包括:
步骤A,容量规划装置根据所述第一分布模型、所述第二分布模型、基站下行传输速率和所述传输时间间隔,确定用户体验速率分布模型。
步骤B,容量规划装置根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制。针对上述步骤A,下面进行示例说明。
这里的用户体验速率也可以称为用户速率、或称为用户实际速率、或称为用户使用速率。这里的用户指的是接入到基站的终端。
比如,当业务测量装置部署于终端上,容量规划装置部署于基站上,则该用户体验速率指的是该业务测量装置对应的终端的体验速率。并且上述基站下行传输速率为该基站针对该终端的下行传输速率。
再比如,当业务测量装置部署于基站上,容量规划装置部署于MEC服务器上,若业务测量装置是以终端粒度测量每个TTI内接入到基站的各个终端的业务包的数量及长度,则该用户体验速率指的是该各个终端的体验速率;若业务测量装置是以基站粒度测量每个TTI内的接入到该基站的所有终端的业务包的数量及长度,则该用户体验速率指的是该所有终端的体验速率的总和。并且,上述基站下行传输速率为该基站针对该接入到该基站的所有的终端的下行传输速率的总和。
其中,根据第一分布模型、第二分布模型的不同,得到的用户体验速率分布模型具体包括以下四种:
(1)用户体验速率分布模型1:与PE模型(第一分布模型为泊松分布模型、第二分布模型为指数分布模型)对应
Figure PCTCN2019116670-appb-000013
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000014
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为TTI,λ为泊松(Possion)分布模型的参数,θ为指数分布模型的参数。
(2)用户体验速率分布模型2:与PP模型(第一分布模型为泊松分布模型、第 二分布模型为Pareto分布模型)对应
Figure PCTCN2019116670-appb-000015
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000016
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为TTI,E[S]为一个TTI内到达的比特数的期望值,λ为泊松(Possion)分布模型的参数,m和α为Pareto分布模型的参数。
(3)用户体验速率分布模型3:与ZE模型(第一分布模型为Zeta分布模型、第二分布模型为指数分布模型)对应
Figure PCTCN2019116670-appb-000017
其中,a 1=…=a s=1,b 1=…=b s=2。
注: pF q(a 1,...,a p;b 1,...,b q;x)为广义超几何函数,它的级数展开为:
Figure PCTCN2019116670-appb-000018
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000019
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为TTI,E[S]为一个TTI内到达的比特数的期望值,ζ()为黎曼函数,p 0为到达的业务包的数量为零的概率, s为Zeta分布模型的参数,θ为指数分布模型的参数。
(4)用户体验速率分布模型4:与ZP模型(第一分布模型为Zeta分布模型、第二分布模型为Pareto分布模型)对应
Figure PCTCN2019116670-appb-000020
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000021
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为TTI,E[S]为一个TTI内到达的比特数的期望值,ζ()为黎曼函数,p 0为到达的业务包的数量为零的概率,s为Zeta分布模型的参数,m和α为Pareto分布模型的参数。
针对上述步骤B,下面进行示例说明。
其中,服务质量要求参数值是服务质量要求参数的取值。服务质量要求参数在具体实现中,可以根据实际需要有多种实现方式,本申请以服务质量要求参数为预设的带宽利用率、或者为预设的用户体验速率满足度为例进行说明。
下面给出执行带宽控制的两种方案。
方案一,容量规划装置根据用户体验速率分布模型、第一分布模型的分布参数值和第二分布模型的分布参数值和预设的带宽利用率,确定忙时用户平均体验速率值。容量规划装置根据忙时用户平均体验速率值,执行带宽控制。
为反映忙时的用户体验,本申请定义一个新的衡量指标:忙时用户体验速率R busy,其物理意义解释如下:当给定队列门限Q ε,若Q(t)>Q ε,则定义该时刻为网络忙时,此时的用户体验速率定义为忙时用户体验速率,即{R busy}={R U(t)|Q(t)>Q ε},从而忙时用户平均体验速率为
Figure PCTCN2019116670-appb-000022
针对上述四种用户体验速率分布模型,分别对应不同的忙时用户平均体验速率公式,如下:
(1)忙时用户平均体验速率公式1:与用户体验速率分布模型1对应,即根据用户体验速率分布模型1可以得到忙时用户平均体验速率公式1。
Figure PCTCN2019116670-appb-000023
其中,
Figure PCTCN2019116670-appb-000024
为忙时用户平均体验速率值,R为基站下行传输速率,τ为TTI,λ为泊松(Possion)分布模型的参数,θ为指数分布模型的参数。
(2)忙时用户平均体验速率公式2:与用户体验速率分布模型2对应,即根据用户体验速率分布模型2可以得到忙时用户平均体验速率公式2。
Figure PCTCN2019116670-appb-000025
其中,
Figure PCTCN2019116670-appb-000026
为忙时用户平均体验速率值,R为基站下行传输速率,τ为TTI,λ为泊松(Possion)分布模型的参数,m和α为Pareto分布模型的参数。
(3)忙时用户平均体验速率公式3:与用户体验速率分布模型3对应,即根据用户体验速率分布模型3可以得到忙时用户平均体验速率公式3。
Figure PCTCN2019116670-appb-000027
其中,H(z;θ,s)=e -θz sF s(a 1,…,a s;b 1,…,b s;θz)。
其中,
Figure PCTCN2019116670-appb-000028
为忙时用户平均体验速率值,R为基站下行传输速率,τ为TTI,s为Zeta分布模型的参数,θ为指数分布模型的参数。
(4)忙时用户平均体验速率公式4:与用户体验速率分布模型4对应,即根据用户体验速率分布模型4可以得到忙时用户平均体验速率公式4。
Figure PCTCN2019116670-appb-000029
其中,
Figure PCTCN2019116670-appb-000030
为忙时用户平均体验速率值,R为基站下行传输速率,τ为TTI,s为Zeta分布模型的参数,ε为预设的带宽利用率且0≤ε≤1,m和α为Pareto分布模型的参数。
在计算得到忙时用户平均体验速率值之后,容量规划装置根据忙时用户平均体验速率值,执行带宽控制,具体包括:若忙时用户平均体验速率值与基站下行传输速率的差值大于第一差值阈值,则增加带宽。或者,若忙时用户平均体验速率值与基站下行传输速率的差值小于第二差值阈值,则减少带宽。比如,第一差值阈值为20M,第二差值阈值为-10M,基站下行传输速率为50M,则当忙时用户平均体验速率值超过70M时,则需要增加带宽,增加方法例如可以是增加固定带宽值,也可以是根据忙时用户平均体验速率值与基站下行传输速率的差值进行增加。当忙时用户平均体验速率值低于40M时,则需要减少带宽,减少方法例如可以是减少固定带宽值,也可以是根据忙时用户平均体验速率值与基站下行传输速率的差值进行减少。
可选的,还可以根据忙时用户平均体验速率值得到小时级流量预测值
Figure PCTCN2019116670-appb-000031
单位为GB。通过将基站的流量门限θ th与T prediction进行比较,确定基站的带宽分配方法,从而执行带宽控制。比如,当T prediction与θ th的差值大于预设的第一流量差值阈值,则增加带宽,当T prediction与θ th的差值小于预设的第二流量差值阈值,则减少带宽。
在一种可能的实现方法中,针对该方案一,则上述步骤201中的业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度,例如可以为:
方法1:业务测量装置周期性地获取设定时长内的每个传输时间间隔内的业务包的数量和业务包的长度。
业务测量装置可以是实时地测量和记录每个TTI内到达业务包的数量和业务包的长度,然后分别周期性地对业务包的数量和业务包的长度进行拟合,获取匹配的分布模型的分布参数。因此,在拟合之前,需要获取记录的TTI内到达业务包的数量和业务包的长度。基于该方法1,业务测量装置是周期性地获取记录的TTI内到达业务包的数量和业务包的长度,并周期性地对获取的业务包的数量和业务包的长度进行拟合。
比如,设定时长为设定的某1天,周期设定为1小时,则业务测量装置每隔1小时获取1次每个传输时间间隔内的业务包的数量和所述业务包的长度,并且根据获取的业务包的数量和业务包的长度拟合分布模型,分别得到第一分布模型的分布参数和第二分布模型的分布参数,进而向容量规划装置上报第一分布模型、第二分布模型、第一分布模型的分布参数和第二分布模型的分布参数。
方法2:业务测量装置周期性地获取设定时长内满足预设的忙时条件的每个传输时间间隔内的业务包的数量和所述业务包的长度。
业务测量装置可以是实时地测量和记录每个TTI内到达业务包的数量和业务包的长度,然后分别周期性地对业务包的数量和业务包的长度进行拟合,获取匹配的分布模型的分布参数。因此,在拟合之前,需要获取记录的满足忙时条件的TTI内到达业务包的数量和业务包的长度。基于该方法2,业务测量装置是周期性地获取记录的满足忙时条件的TTI内到达业务包的数量和业务包的长度,并周期性地对获取的业务包的数量和业务包的长度进行拟合。
比如,设定时长为设定的某1天,周期设定为1小时,则业务测量装置每隔1小时获取1次满足忙时条件的每个传输时间间隔内的业务包的数量和所述业务包的长度,并且根据获取的业务包的数量和业务包的长度拟合分布模型,分别得到第一分布模型的分布参数和第二分布模型的分布参数,进而向容量规划装置上报第一分布模型、第二分布模型、第一分布模型的分布参数和第二分布模型的分布参数。
比如,第1个小时中有30分钟是满足忙时条件,则使用该小时内的满足忙时条件的每个传输时间间隔内的业务包的数量和所述业务包的长度,并且根据获取的业务包的数量和业务包的长度拟合分布模型,分别得到第一分布模型的分布参数和第二分布模型的分布参数,进而向容量规划装置上报第一分布模型、第二分布模型、第一分布模型的分布参数和第二分布模型的分布参数。
再比如,第2个小时中有25分钟是满足忙时条件,则使用该小时内的满足忙时条件的25分钟的每个传输时间间隔内的业务包的数量和所述业务包的长度,并且根据获取的业务包的数量和业务包的长度拟合分布模型,分别得到第一分布模型的分布参数和第二分布模型的分布参数,进而向容量规划装置上报第一分布模型、第二分布模型、第一分布模型的分布参数和第二分布模型的分布参数。
需要说明的是,上述示例中的满足忙时条件的30分钟、25分钟可以是连续时间,也可以是非连续的累计时间。
在一种实现方式中,满足忙时条件,例如可以是:若t时刻(即第t个传输时间间隔)基站上的队列的长度大于预设的队列门限值,则确定满足忙时条件。其中,基站上的队列用于缓存业务包,预设的队列门限值例如可以定义为:Q ε=εRτ。其中,Q ε为队列门限值,ε为预设的带宽利用率,R为基站下行传输速率,τ为传输时间间隔的长度。
方案二,容量规划装置根据用户体验速率分布模型、第一分布模型的分布参数值和第二分布模型的分布参数值和预设的用户体验速率满足度,确定用户体验速率下限值。容量规划装置根据用户体验速率下限值,执行带宽控制。
给定预设的用户体验速率满足度η,为保证用户服务质量,我们需要找到能保证用户体验速率满足度要求的最小速率R 0,使得用户体验速率超过R 0的概率不小于预设的用户体验速率满足度η,即
Pr(R U>R 0)≥η
取Pr(R U>R 0)=η,可以得出用户体验速率下限值R 0=R min。本申请利用用户体验速率分布模型公式,通过解方程Pr(R U>R min)=η,计算出用户体验速率下限R min
其中,用户体验速率下限公式如下:
(1)用户体验速率下限公式1:与用户体验速率分布模型1对应,即根据用户体验速率分布模型1可以得到用户体验速率下限公式1。
Figure PCTCN2019116670-appb-000032
其中,λ为泊松(Possion)分布模型的参数,θ为指数分布模型的参数。
(2)用户体验速率下限公式2:与用户体验速率分布模型2对应,即根据用户体验速率分布模型2可以得到用户体验速率下限公式2。
Figure PCTCN2019116670-appb-000033
其中,R为基站下行传输速率,τ为TTI,E[S]为一个TTI内到达的比特数的期望值,λ为泊松(Possion)分布模型的参数,m和α为Pareto分布模型的参数。
(3)用户体验速率下限公式3:与用户体验速率分布模型3对应,即根据用户体验速率分布模型3可以得到用户体验速率下限公式3。
Figure PCTCN2019116670-appb-000034
其中,τ为TTI,E[S]为一个TTI内到达的比特数的期望值,ζ()为黎曼函数,R为基站下行传输速率,p 0为到达的业务包的数量为零的概率,s为Zeta分布模型的参数,θ为指数分布模型的参数。
(4)用户体验速率下限公式4:与用户体验速率分布模型4对应,即根据用户体验速率分布模型4可以得到用户体验速率下限公式4。
Figure PCTCN2019116670-appb-000035
其中,R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓 存业务包,τ为TTI,E[S]为一个TTI内到达的比特数的期望值,ζ()为黎曼函数,p 0为到达的业务包的数量为零的概率,s为Zeta分布模型的参数,m和α为Pareto分布模型的参数。
例如,预设的用户体验速率满足度η=95%,结合上报的业务包长、包数的拟合参数,以及基站下行传输速率、TTI时长等系统参数,利用上述四式,我们可以确定的求解用户体验速率下限R min
在一种实现方式中,容量规划装置根据用户体验速率下限值,执行带宽控制,包括:若用户体验速率下限值与基站下行传输速率的差值大于第三差值阈值,则增加带宽。或者,若忙时用户平均体验速率值与基站下行传输速率的差值小于第四差值阈值,则减少带宽。比如,第一差值阈值为20M,第二差值阈值为-10M,基站下行传输速率为50M,则当用户体验速率下限值超过70M时,则需要增加带宽,增加方法例如可以是增加固定带宽值,也可以是根据用户体验速率下限值与基站下行传输速率的差值进行增加。当用户体验速率下限值低于40M时,则需要减少带宽,减少方法例如可以是减少固定带宽值,也可以是根据用户体验速率下限值与基站下行传输速率的差值进行减少。
可选的,还可以根据用户体验速率下限值得到小时级流量预测值
Figure PCTCN2019116670-appb-000036
4单位为GB。通过将基站的流量下限门限θ th与T prediction进行比较,确定基站的带宽分配方法,从而执行带宽控制。比如,当T prediction与θ th的差值大于预设的第三流量差值阈值,则增加带宽,当T prediction与θ th的差值小于预设的第四流量差值阈值,则减少带宽。
在一种可能的实现方法中,针对该方案二,则上述步骤201中的业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度,例如可以为:
业务测量装置周期性地获取设定时长内的每个传输时间间隔内的业务包的数量和业务包的长度。
比如,设定时长为设定的某1天,周期设定为1小时,则业务测量装置每隔1小时获取1次每个传输时间间隔内的业务包的数量和所述业务包的长度,并且根据获取的业务包的数量和业务包的长度拟合分布模型,分别得到第一分布模型的分布参数和第二分布模型的分布参数,进而向容量规划装置上报第一分布模型、第二分布模型、第一分布模型的分布参数和第二分布模型的分布参数。
作为一种实现方式,当业务测量装置部署于基站上,容量规划装置部署于MEC服务器上,则上述步骤B中,容量规划装置执行带宽控制,例如可以是通过以下方法实现:容量规划装置向业务测量装置发送第一通知消息,第一通知消息包括带宽的控制策略,这里的带宽控制策略可以是增加带宽或减少带宽。然后,业务测量装置向容 量规划装置发送针对第一通知消息的第一确认消息。接着,容量规划装置向业务测量装置发送第二通知消息,该第二通知消息包括带宽值。从而业务测量装置可以根据带宽值,执行带宽控制。具体的,基站接收到带宽值后,可以根据带宽值,执行对相应的终端的带宽控制。这里的带宽值可以是增加或减少的相对带宽值,也可以是业务测量装置对终端或基站的带宽进行控制后的带宽值。可选的,业务测量装置还可以向容量规划装置发送针对第二通知消息的第二响应消息。
作为又一种实现方式,当业务测量装置部署于终端上,容量规划装置部署于基站上,则上述步骤B中,容量规划装置执行带宽控制,例如可以是通过以下方法实现:容量规划装置通知基站各个终端的带宽值,这里的带宽值可以是增加或减少的相对带宽值,也可以是业务测量装置对终端或基站的带宽进行控制后的带宽值。然后,基站根据接收到的带宽值,执行对相应的终端的带宽控制。
需要说明的是,在系统初始化时,各个基站之间可以平均分配网络总带宽,一个基站内的各个终端之间也可以平均分配该基站的网络带宽。在业务进行过程中,由于有的终端/基站业务较忙,对带宽的需求增加,而有的终端/基站业务较闲,对带宽的需求减少,因此,通过本申请的上述容量规划方法,可以执行对各个终端/基站的带宽控制,使得可以根据各个终端的业务繁忙程度,执行相应的带宽控制,有助于提升系统效率和提升资源利用效率。
本申请的上述方案,提炼出了简洁实用的业务到达和业务包长的模型,从而实现快速的业务特征参数学习,复杂度低于现有技术。并且,容量规划装置可以基于用户体验速率分布模型,直接反映网络中用户体验质量,为运营商改善网络服务质量提供了严谨的依据。因此,本申请的容量规划方法能够有效预测网络流量,合理规划带宽分配。
图3示出了本发明实施例中所涉及的装置的可能的示例性框图,该装置300可以以软件或者硬件的形式存在。装置300可以包括:获取单元301、确定单元302和通信单元303。可选的,该装置300还可以包括控制单元304。作为一种实现方式,该通信单元303可以包括接收单元和发送单元。作为一种实现方式,获取单元301、确定单元302和控制单元304可以集成于一个处理单元,该处理单元用于对装置300的动作进行控制管理。通信单元303用于支持装置300与其他网络实体的通信。
其中,当获取单元301、确定单元302和控制单元304可以集成于一个处理单元时,该处理单元可以是处理器或控制器,例如可以是通用中央处理器(central processing unit,CPU),通用处理器,数字信号处理(digital signal processing,DSP),专用集成电路(application specific integrated circuits,ASIC),现场可编程门阵列(field programmable gate array,FPGA)或者其他可编程逻辑器件、晶体管逻辑器件、硬件部件或者其任意组合。其可以实现或执行结合本发明公开内容所描述的各种示例性的逻辑方框,模块和电路。所述处理器也可以是实现计算功能的组合,例如包括一个或多个微处理器组合,DSP和微处理器的组合等等。通信单元304可以是通信接口、收发器或收发电路等,其中,该通信接口是统称,在具体实现中,该通信接口可以包括多个接口。
该装置300可以为上述任一实施例中的业务测量装置,还可以为可用于业务测量装置的芯片。例如,当装置300为业务测量装置,获取单元301、确定单元302和控 制单元304集成于一个处理单元时,该处理单元例如可以是处理器,通信单元303例如可以是收发器,该收发器包括射频电路。例如,当装置300为可用于业务测量装置的芯片,获取单元301、确定单元302和控制单元304集成于一个处理单元时,该处理单元例如可以是处理器,该通信单元303例如可以是输入/输出接口、管脚或电路等。
获取单元301,用于获取设定时长内的每个传输时间间隔内的业务包的数量和业务包的长度。
确定单元302,用于确定与业务包的数量匹配的第一分布模型的分布参数值,以及确定与业务包的长度匹配的第二分布模型的分布参数值。
通信单元303,用于向容量规划装置发送第一分布模型的分布参数值和第二分布模型的分布参数值。
在一种可能的实现方法中,确定单元302,具体用于:使用所述业务包的数量拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;确定拟合度最高的分布模型为所述业务包的数量匹配的第一分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第一分布模型的分布参数值。
在一种可能的实现方法中,所述至少两个分布模型包括以下分布模型中的一种或多种:泊松分布模型、Zeta分布模型。
在一种可能的实现方法中,确定单元302,具体用于:使用所述业务包的长度拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;确定拟合度最高的分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第二分布模型的分布参数值。
在一种可能的实现方法中,所述至少两个分布模型包括以下分布模型中的一种或多种:指数分布模型、Pareto分布模型。
在一种可能的实现方法中,确定单元302,具体用于:
拟合第一预设分布模型,得到所述第一预设分布模型的分布参数值;并确定所述第一预设分布模型为所述业务包的数量匹配的第一分布模型,以及确定所述第一预设分布模型的分布参数值为所述第一分布模型的分布参数值;
拟合第二预设分布模型,得到所述第二预设分布模型的分布参数值;并确定所述第二预设分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述第二预设分布模型的分布参数值为所述第二分布模型的分布参数值。
在一种可能的实现方法中,第一预设分布模型为泊松分布模型、或Zeta分布模型,所述第二预设分布模型为指数分布模型、或Pareto分布模型。
在一种可能的实现方法中,确定单元302,具体用于:使用业务包的数量拟合泊松分布模型,得到第一拟合度和泊松分布模型的分布参数值。使用业务包的数量拟合Zeta分布模型,得到第二拟合度和Zeta分布模型的分布参数值。若第一拟合度大于第二拟合度,则确定泊松分布模型的分布模型参数为业务包的数量匹配的第一分布模型的分布参数值,第一分布模型为泊松分布模型。或者,若第一拟合度不大于第二拟合度,则确定Zeta分布模型的分布参数值为业务包的数量匹配的第一分布模型的分布参数值,第一分布模型为Zeta分布模型。
在一种可能的实现方法中,确定单元302,具体用于:
使用业务包的长度拟合指数分布模型,得到第三拟合度和指数分布模型的分布参 数值。
使用业务包的长度拟合Pareto分布模型,得到第四拟合度和Pareto分布模型的分布参数值。
若第三拟合度大于第四拟合度,则确定指数分布模型的分布参数值为业务包的长度匹配的第二分布模型的分布参数值,第二分布模型为指数分布模型。或者,
若第三拟合度不大于第四拟合度,则确定Pareto分布模型的分布模型参数为业务包的长度匹配的第二分布模型的分布模型参数,第二分布模型为Pareto分布模型。
在一种可能的实现方法中,通信单元303,具体用于:
向容量规划装置发送第一上报消息,第一上报消息包括第一分布模型的标识和第二分布模型的标识。
从容量规划装置接收针对第一上报消息的第一响应消息。
向容量规划装置发送第二上报消息,第二上报消息包括第一分布模型的分布参数值和第二分布模型的分布参数值。
在一种可能的实现方法中,通信单元303,还用于:从容量规划装置接收第一通知消息,第一通知消息包括带宽的控制策略。向容量规划装置发送针对第一通知消息的第一确认消息。从容量规划装置接收第二通知消息,第二通知消息包括带宽值。
控制单元304,用于根据带宽值,执行带宽控制。
在一种可能的实现方法中,所述获取单元301,具体用于:
周期性地获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度;或者,
周期性地获取设定时长内满足预设的忙时条件的每个传输时间间隔内的业务包的数量和所述业务包的长度。
在一种可能的实现方法中,业务测量装置部署于终端上,容量规划装置部署于基站上。获取单元301,具体用于获取设定时长内的每个传输时间间隔内的终端的业务包的数量和终端的业务包的长度。
在一种可能的实现方法中,业务测量装置部署于基站上,容量规划装置部署于移动边缘计算服务器上。获取单元301,具体用于:
获取设定时长内的每个传输时间间隔内的接入到基站的各个终端的业务包的数量和各个终端的业务包的长度。或者,
获取设定时长内的每个传输时间间隔内的基站的业务包的数量和所述基站的业务包的长度。
图3所示的装置为业务测量装置时,所用于执行的容量规划方法的具体有益效果,可参考前述方法实施例中的相关描述,这里不再赘述。可以理解的是,本申请实施例中的单元也可以称为模块。上述单元或者模块可以独立存在,也可以集成在一起。
图4示出了本发明实施例中所涉及的装置的可能的示例性框图,该装置400可以以软件或者硬件的形式存在。装置400可以包括:通信单元401、控制单元403。可选的,该装置400还可以包括确定单元402。作为一种实现方式,该通信单元401可以包括接收单元和发送单元。作为一种实现方式,确定单元402和控制单元403可以集成于一个处理单元,该处理单元用于对装置400的动作进行控制管理。通信单元401用于支持装置400与其他网络实体的通信。
其中,当确定单元402和控制单元403集成于一个处理单元时,该处理单元可以是处理器或控制器,例如可以是CPU,通用处理器,DSP,ASIC,FPGA或者其他可编程逻辑器件、晶体管逻辑器件、硬件部件或者其任意组合。其可以实现或执行结合本发明公开内容所描述的各种示例性的逻辑方框,模块和电路。所述处理器也可以是实现计算功能的组合,例如包括一个或多个微处理器组合,DSP和微处理器的组合等等。通信单元401可以是通信接口、收发器或收发电路等,其中,该通信接口是统称,在具体实现中,该通信接口可以包括多个接口。
该装置400可以为上述任一实施例中的容量规划装置,还可以为可用于容量规划装置的芯片。例如,当装置300为容量规划装置,确定单元402和控制单元403集成于一个处理单元时,该处理单元例如可以是处理器,通信单元401例如可以是收发器,该收发器包括射频电路。例如,当装置400为可用于容量规划装置的芯片,确定单元402和控制单元403集成于一个处理单元时,该处理单元例如可以是处理器,该通信单元401例如可以是输入/输出接口、管脚或电路等。
通信单元401,用于从业务测量装置接收第一分布模型的分布参数值和第二分布模型的分布参数值,第一分布模型和第二分布模型分别为业务测量装置获取到的设定时长内的每个传输时间间隔内的业务包的数量所匹配的分布模型和业务包的长度所匹配的分布模型。控制单元403,用于根据第一分布模型、第二分布模型、第一分布模型的分布参数值和第二分布模型的分布参数值,执行带宽控制。
在一种可能的实现方法中,确定单元402,用于根据所述第一分布模型、所述第二分布模型、基站下行传输速率和所述传输时间间隔,确定用户体验速率分布模型;所述控制单元403,具体用于根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制。
在一种可能的实现方法中,所述服务质量要求参数值为预设的带宽利用率;控制单元403,具体用于:根据用户体验速率分布模型、第一分布模型的分布参数值和第二分布模型的分布参数值和预设的带宽利用率,确定忙时用户平均体验速率值。根据忙时用户平均体验速率值,执行带宽控制。
在一种可能的实现方法中,第一分布模型为Zeta分布模型,第一分布模型的分布模型参数包括s,第二分布模型为Pareto分布模型,第二分布模型的分布模型参数包括m和α。
用户体验速率分布模型为:
Figure PCTCN2019116670-appb-000037
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000038
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为传输时间间隔,ζ()为黎 曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数量为零的概率。
控制单元403,用于根据下列公式确定忙时用户平均体验速率值:
Figure PCTCN2019116670-appb-000039
其中,
Figure PCTCN2019116670-appb-000040
为忙时用户平均体验速率值,ε为预设的带宽利用率且0≤ε≤1。
在一种可能的实现方法中,控制单元403,具体用于:若忙时用户平均体验速率值与基站下行传输速率的差值大于第一差值阈值,则增加带宽。或者,若忙时用户平均体验速率值与基站下行传输速率的差值小于第二差值阈值,则减少带宽。
在一种可能的实现方法中,所述服务质量参数值为预设的用户体验速率满足度;控制单元403,具体用于:根据用户体验速率分布模型、第一分布模型的分布参数值和第二分布模型的分布参数值和预设的用户体验速率满足度,确定用户体验速率下限值。根据用户体验速率下限值,执行带宽控制。
在一种可能的实现方法中,第一分布模型为Zeta分布模型,第一分布模型的分布模型参数包括s,第二分布模型为Pareto分布模型,第二分布模型的分布模型参数包括m和α。
用户体验速率分布模型为:
Figure PCTCN2019116670-appb-000041
其中,Pr()为用户体验速率分布模型,x是用户体验速率的自变量,R U为用户体验速率,且t时刻的用户体验速率为
Figure PCTCN2019116670-appb-000042
R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,队列用于缓存业务包,τ为传输时间间隔,ζ()为黎曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数量为零的概率。
控制单元403,用于根据下列公式确定用户体验速率下限值:
Figure PCTCN2019116670-appb-000043
其中,R min为用户体验速率下限值,η为预设的用户体验速率满足度。
在一种可能的实现方法中,控制单元403,具体用于:若用户体验速率下限值与基站下行传输速率的差值大于第三差值阈值,则增加带宽。或者,若忙时用户平均体验速率值与基站下行传输速率的差值小于第四差值阈值,则减少带宽。
在一种可能的实现方法中,所述通信单元401,还用于:从所述业务测量装置接收所述第一分布模型的标识信息和第二分布模型的标识信息,所述第一分布模型的标识信息用于标识选择的所述第一分布模型,所述第二分布模型的标识信息用于标识选择的所述第二分布模型。
在又一种可能的实现方法中,所述通信单元401,还用于:从所述业务测量装置接收业务到达模型的标识信息,所述业务到达模型的标识信息用于标识选择的所述第一分布模型和所述第二分布模型所对应的业务到达模型。所述确定单元402,还用于根据所述业务到达模型的标识信息,确定选择的所述第一分布模型和所述第二分布模型。
图4所示的装置为容量规划装置时,所用于执行的容量规划方法的具体有益效果,可参考前述方法实施例中的相关描述,这里不再赘述。可以理解的是,本申请实施例中的单元也可以称为模块。上述单元或者模块可以独立存在,也可以集成在一起。
参阅图5所示,为本申请提供的一种装置示意图,该装置可以是本申请实施例中的业务测量装置、或容量规划装置,也可以是可用于业务测量装置、或容量规划装置的部件。该装置500包括:处理器502、通信接口503、存储器501。可选的,装置500还可以包括总线504。其中,通信接口503、处理器502以及存储器501可以通过通信线路504相互连接;通信线路504可以是外设部件互连标准(peripheral component interconnect,简称PCI)总线或扩展工业标准结构(extended industry standard architecture,简称EISA)总线等。所述通信线路504可以分为地址总线、数据总线、控制总线等。为便于表示,图5中仅用一条粗线表示,但并不表示仅有一根总线或一种类型的总线。
处理器502可以是一个CPU,微处理器,ASIC,或一个或多个用于控制本申请方案程序执行的集成电路。
通信接口503,可以是使用任何收发器一类的装置,用于与其他设备或通信网络通信,如以太网,无线接入网(radio access network,RAN),无线局域网(wireless local area networks,WLAN),有线接入网等。
存储器501可以是只读存储器(read-only memory,ROM)或可存储静态信息和指令的其他类型的静态存储设备,随机存取存储器(random access memory,RAM)或者可存储信息和指令的其他类型的动态存储设备,也可以是电可擦可编程只读存储器(electrically er服务器able programmable read-only memory,EEPROM)、只读光盘(compact disc read-only memory,CD-ROM)或其他光盘存储、光碟存储(包括压缩光碟、激光碟、光碟、数字通用光碟、蓝光光碟等)、磁盘存储介质或者其他磁存储设备、或者能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质,但不限于此。存储器可以是独立存在,通过通信线路504与处理器相连接。存储器也可以和处理器集成在一起。
其中,存储器501用于存储执行本申请方案的计算机执行指令,并由处理器502来控制执行。处理器502用于执行存储器501中存储的计算机执行指令,从而实现本申请上述实施例提供的容量规划方法。
可选的,本申请实施例中的计算机执行指令也可以称之为应用程序代码,本申请实施例对此不作具体限定。
在上述实施例中,可以全部或部分地通过软件、硬件、固件或者其任意组合来实 现。当使用软件实现时,可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时,全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中,或者从一个计算机可读存储介质向另一个计算机可读存储介质传输,例如,所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包括一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质,(例如,软盘、硬盘、磁带)、光介质(例如,DVD)、或者半导体介质(例如固态硬盘(Solid State Disk,SSD))等。
本申请实施例中所描述的各种说明性的逻辑单元和电路可以通过通用处理器,数字信号处理器,专用集成电路(ASIC),现场可编程门阵列(FPGA)或其它可编程逻辑装置,离散门或晶体管逻辑,离散硬件部件,或上述任何组合的设计来实现或操作所描述的功能。通用处理器可以为微处理器,可选地,该通用处理器也可以为任何传统的处理器、控制器、微控制器或状态机。处理器也可以通过计算装置的组合来实现,例如数字信号处理器和微处理器,多个微处理器,一个或多个微处理器联合一个数字信号处理器核,或任何其它类似的配置来实现。
本申请实施例中所描述的方法或算法的步骤可以直接嵌入硬件、处理器执行的软件单元、或者这两者的结合。软件单元可以存储于RAM存储器、闪存、ROM存储器、EPROM存储器、EEPROM存储器、寄存器、硬盘、可移动磁盘、CD-ROM或本领域中其它任意形式的存储媒介中。示例性地,存储媒介可以与处理器连接,以使得处理器可以从存储媒介中读取信息,并可以向存储媒介存写信息。可选地,存储媒介还可以集成到处理器中。处理器和存储媒介可以设置于ASIC中,ASIC可以设置于终端中。可选地,处理器和存储媒介也可以设置于终端中的不同的部件中。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上,使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理,从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
尽管结合具体特征及其实施例对本申请进行了描述,显而易见的,在不脱离本申请的精神和范围的情况下,可对其进行各种修改和组合。相应地,本说明书和附图仅仅是所附权利要求所界定的本申请的示例性说明,且视为已覆盖本申请范围内的任意和所有修改、变化、组合或等同物。显然,本领域的技术人员可以对本申请进行各种改动和变型而不脱离本申请的精神和范围。这样,倘若本申请的这些修改和变型属于本申请权利要求及其等同技术的范围之内,则本申请也意图包括这些改动和变型在内。

Claims (30)

  1. 一种容量规划方法,其特征在于,包括:
    容量规划装置从业务测量装置接收第一分布模型的分布参数值和第二分布模型的分布参数值,所述第一分布模型和所述第二分布模型分别为所述业务测量装置获取到的设定时长内的每个传输时间间隔内的业务包的数量所匹配的分布模型和所述业务包的长度所匹配的分布模型;
    所述容量规划装置根据所述第一分布模型、所述第二分布模型、所述第一分布模型的分布参数值和所述第二分布模型的分布参数值,执行带宽控制。
  2. 如权利要求1所述的方法,其特征在于,所述容量规划装置根据所述第一分布模型、所述第二分布模型、所述第一分布模型的分布参数值和所述第二分布模型的分布参数值,执行带宽控制,包括:
    所述容量规划装置根据所述第一分布模型、所述第二分布模型、基站下行传输速率和所述传输时间间隔,确定用户体验速率分布模型;
    所述容量规划装置根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制。
  3. 如权利要求2所述的方法,其特征在于,所述服务质量要求参数值为预设的带宽利用率;
    所述容量规划装置根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制,包括:
    所述容量规划装置根据所述用户体验速率分布模型、所述第一分布模型的分布参数值和所述第二分布模型的分布参数值和预设的带宽利用率,确定忙时用户平均体验速率值;
    所述容量规划装置根据所述忙时用户平均体验速率值,执行带宽控制。
  4. 如权利要求3所述的方法,其特征在于,所述第一分布模型为Zeta分布模型,所述第一分布模型的分布模型参数包括s,所述第二分布模型为Pareto分布模型,所述第二分布模型的分布模型参数包括m和α;
    所述用户体验速率分布模型为:
    Figure PCTCN2019116670-appb-100001
    其中,Pr()为所述用户体验速率分布模型,R U为用户体验速率,且t时刻的用户体验速率为
    Figure PCTCN2019116670-appb-100002
    R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,所述队列用于缓存业务包,τ为所述传输时间间隔,ζ()为黎曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数量为零的概率;
    所述容量规划装置根据下列公式确定所述忙时用户平均体验速率值:
    Figure PCTCN2019116670-appb-100003
    其中,
    Figure PCTCN2019116670-appb-100004
    为所述忙时用户平均体验速率值,ε为预设的带宽利用率且0≤ε≤1。
  5. 如权利要求3或4所述的方法,其特征在于,所述容量规划装置根据所述忙时用户平均体验速率值,执行带宽控制,包括:
    若所述忙时用户平均体验速率值与基站下行传输速率的差值大于第一差值阈值,则增加带宽;或者,
    若所述忙时用户平均体验速率值与基站下行传输速率的差值小于第二差值阈值,则减少带宽。
  6. 如权利要求2所述的方法,其特征在于,所述服务质量参数值为预设的用户体验速率满足度;
    所述容量规划装置根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制,包括:
    所述容量规划装置根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和预设的用户体验速率满足度,确定用户体验速率下限值;
    所述容量规划装置根据所述用户体验速率下限值,执行带宽控制。
  7. 如权利要求6所述的方法,其特征在于,所述第一分布模型为Zeta分布模型,所述第一分布模型的分布模型参数包括s,所述第二分布模型为Pareto分布模型,所述第二分布模型的分布模型参数包括m和α;
    所述用户体验速率分布模型为:
    Figure PCTCN2019116670-appb-100005
    其中,Pr()为所述用户体验速率分布模型,R U为用户体验速率,且t时刻的用户体验速率为
    Figure PCTCN2019116670-appb-100006
    R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,所述队列用于缓存业务包,τ为所述传输时间间隔,ζ()为黎曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数量为零的概率;
    所述容量规划装置根据下列公式确定所述用户体验速率下限值:
    Figure PCTCN2019116670-appb-100007
    其中,R min为用户体验速率下限值,η为预设的用户体验速率满足度。
  8. 如权利要求6或7所述的方法,其特征在于,所述容量规划装置根据所述用户体验速率下限值,执行带宽控制,包括:
    若所述用户体验速率下限值与基站下行传输速率的差值大于第三差值阈值,则增加带宽;或者,
    若所述忙时用户平均体验速率值与基站下行传输速率的差值小于第四差值阈值,则减少带宽。
  9. 如权利要求1-8任一项所述的方法,其特征在于,所述方法还包括:
    所述容量规划装置从所述业务测量装置接收所述第一分布模型的标识信息和第二分布模型的标识信息,所述第一分布模型的标识信息用于标识选择的所述第一分布模型,所述第二分布模型的标识信息用于标识选择的所述第二分布模型。
  10. 如权利要求1-8任一项所述的方法,其特征在于,所述方法还包括:
    所述容量规划装置从所述业务测量装置接收业务到达模型的标识信息,所述业务到达模型的标识信息用于标识选择的所述第一分布模型和所述第二分布模型所对应的业务到达模型;
    所述容量规划装置根据所述业务到达模型的标识信息,确定选择的所述第一分布模型和所述第二分布模型。
  11. 一种容量规划方法,其特征在于,包括:
    业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度;
    所述业务测量装置确定与所述业务包的数量匹配的第一分布模型的分布参数值,以及确定与所述业务包的长度匹配的第二分布模型的分布参数值;
    所述业务测量装置向容量规划装置发送所述第一分布模型的分布参数值和所述第二分布模型的分布参数值。
  12. 如权利要求11所述的方法,其特征在于,所述业务测量装置确定与所述业务包的数量匹配的第一分布模型的分布参数值,包括:
    所述业务测量装置使用所述业务包的数量拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;
    所述业务测量装置确定拟合度最高的分布模型为所述业务包的数量匹配的第一分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第一分布模型的分布参数值。
  13. 如权利要求12所述的方法,其特征在于,所述至少两个分布模型包括以下分布模型中的一种或多种:
    泊松分布模型、Zeta分布模型。
  14. 如权利要求11-13任一项所述的方法,其特征在于,所述业务测量装置确定与所述业务包的长度匹配的第二分布模型的分布参数值,包括:
    所述业务测量装置使用所述业务包的长度拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;
    所述业务测量装置确定拟合度最高的分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第二分布模型的分布参数值。
  15. 如权利要求14所述的方法,其特征在于,所述至少两个分布模型包括以下分布模型中的一种或多种:
    指数分布模型、Pareto分布模型。
  16. 如权利要求11所述的方法,其特征在于,所述业务测量装置确定与所述业务包的数量匹配的第一分布模型的分布参数值,包括:
    所述业务测量装置拟合第一预设分布模型,得到所述第一预设分布模型的分布参数值;
    所述业务测量装置确定所述第一预设分布模型为所述业务包的数量匹配的第一分布模型,以及确定所述第一预设分布模型的分布参数值为所述第一分布模型的分布参数值;
    所述业务测量装置确定与所述业务包的长度匹配的第二分布模型的分布参数值,包括:
    所述业务测量装置拟合第二预设分布模型,得到所述第二预设分布模型的分布参数值;
    所述业务测量装置确定所述第二预设分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述第二预设分布模型的分布参数值为所述第二分布模型的分布参数值。
  17. 如权利要16所述的方法,其特征在于,所述第一预设分布模型为泊松分布模型、或Zeta分布模型,所述第二预设分布模型为指数分布模型、或Pareto分布模型。
  18. 如权利要11-17任一项所述的方法,其特征在于,所述业务测量装置获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度,包括:
    所述业务测量装置周期性地获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度;或者,
    所述业务测量装置周期性地获取设定时长内满足预设的忙时条件的每个传输时间间隔内的业务包的数量和所述业务包的长度。
  19. 一种容量规划装置,其特征在于,包括:
    通信单元,用于从业务测量装置接收第一分布模型的分布参数值和第二分布模型的分布参数值,所述第一分布模型和所述第二分布模型分别为所述业务测量装置获取到的设定时长内的每个传输时间间隔内的业务包的数量所匹配的分布模型和所述业务包的长度所匹配的分布模型;
    控制单元,用于根据所述第一分布模型、所述第二分布模型、所述第一分布模型的分布参数值和所述第二分布模型的分布参数值,执行带宽控制。
  20. 如权利要求19所述的装置,其特征在于,所述装置还包括确定单元,用于根据所述第一分布模型、所述第二分布模型、基站下行传输速率和所述传输时间间隔,确定用户体验速率分布模型;
    所述控制单元,具体用于根据所述用户体验速率分布模型、所述第一分布模型的分布参数值、所述第二分布模型的分布参数值和服务质量要求参数值,执行带宽控制。
  21. 如权利要求20所述的装置,其特征在于,所述服务质量要求参数值为预设的带宽利用率;
    所述控制单元,具体用于:
    根据所述用户体验速率分布模型、所述第一分布模型的分布参数值和所述第二分布模型的分布参数值和预设的带宽利用率,确定忙时用户平均体验速率值;
    根据所述忙时用户平均体验速率值,执行带宽控制。
  22. 如权利要求21所述的装置,其特征在于,所述第一分布模型为Zeta分布模型,所述第一分布模型的分布模型参数包括s,所述第二分布模型为Pareto分布模型,所述第二分布模型的分布模型参数包括m和α;
    所述用户体验速率分布模型为:
    Figure PCTCN2019116670-appb-100008
    其中,Pr()为所述用户体验速率分布模型,R U为用户体验速率,且t时刻的用户体验速率为
    Figure PCTCN2019116670-appb-100009
    R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,所述队列用于缓存业务包,τ为所述传输时间间隔,ζ()为黎曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数量为零的概率;
    所述控制单元,用于根据下列公式确定所述忙时用户平均体验速率值:
    Figure PCTCN2019116670-appb-100010
    其中,
    Figure PCTCN2019116670-appb-100011
    为所述忙时用户平均体验速率值,ε为预设的带宽利用率且0≤ε≤1。
  23. 如权利要求21或22所述的装置,其特征在于,所述控制单元,具体用于:
    若所述忙时用户平均体验速率值与基站下行传输速率的差值大于第一差值阈值,则增加带宽;或者,
    若所述忙时用户平均体验速率值与基站下行传输速率的差值小于第二差值阈值,则减少带宽。
  24. 如权利要求20所述的装置,其特征在于,所述服务质量参数值为预设的用户体验速率满足度;
    所述控制单元,具体用于:
    根据所述用户体验速率分布模型、所述第一分布模型的分布参数值和所述第二分布模型的分布参数值和预设的用户体验速率满足度,确定用户体验速率下限值;
    根据所述用户体验速率下限值,执行带宽控制。
  25. 如权利要求24所述的装置,其特征在于,所述第一分布模型为Zeta分布模型,所述第一分布模型的分布模型参数包括s,所述第二分布模型为Pareto分布模型,所述第二分布模型的分布模型参数包括m和α;
    所述用户体验速率分布模型为:
    Figure PCTCN2019116670-appb-100012
    其中,Pr()为所述用户体验速率分布模型,R U为用户体验速率,且t时刻的用户体验速率为
    Figure PCTCN2019116670-appb-100013
    R为基站下行传输速率,Q(t)为t时刻的基站上的队列长度,所述队列用于缓存业务包,τ为所述传输时间间隔,ζ()为黎曼函数,E[S]为一个传输时间间隔内到达的比特数的期望值,p 0为到达的业务包的数量为零的概率;
    所述控制单元,用于根据下列公式确定所述用户体验速率下限值:
    Figure PCTCN2019116670-appb-100014
    其中,R min为用户体验速率下限值,η为预设的用户体验速率满足度。
  26. 如权利要求24或25所述的装置,其特征在于,所述控制单元,具体用于:
    若所述用户体验速率下限值与基站下行传输速率的差值大于第三差值阈值,则增加带宽;或者,
    若所述忙时用户平均体验速率值与基站下行传输速率的差值小于第四差值阈值,则减少带宽。
  27. 一种业务测量装置,其特征在于,包括:
    获取单元,用于获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的长度;
    确定单元,用于确定与所述业务包的数量匹配的第一分布模型的分布参数值,以及确定与所述业务包的长度匹配的第二分布模型的分布参数值;
    通信单元,用于向容量规划装置发送所述第一分布模型的分布参数值和所述第二分布模型的分布参数值。
  28. 如权利要求27所述的装置,其特征在于,所述确定单元,具体用于:
    使用所述业务包的数量拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;
    确定拟合度最高的分布模型为所述业务包的数量匹配的第一分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第一分布模型的分布参数值。
  29. 如权利要求27或28所述的装置,其特征在于,所述确定单元,具体用于:
    使用所述业务包的长度拟合至少两个分布模型,得到每个分布模型的拟合度和所述分布模型的分布参数值;
    确定拟合度最高的分布模型为所述业务包的长度匹配的第二分布模型,以及确定所述拟合度最高的分布模型的分布参数值为所述第二分布模型的分布参数值。
  30. 如权利要27-29任一项所述的装置,其特征在于,所述获取单元,具体用于:
    周期性地获取设定时长内的每个传输时间间隔内的业务包的数量和所述业务包的 长度;或者,
    周期性地获取设定时长内满足预设的忙时条件的每个传输时间间隔内的业务包的数量和所述业务包的长度。
PCT/CN2019/116670 2018-11-16 2019-11-08 一种容量规划方法及装置 WO2020098575A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP19885672.6A EP3873123B1 (en) 2018-11-16 2019-11-08 Capacity planning method and device
US17/321,104 US11736954B2 (en) 2018-11-16 2021-05-14 Capacity planning method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811367720.6A CN111200821B (zh) 2018-11-16 2018-11-16 一种容量规划方法及装置
CN201811367720.6 2018-11-16

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US17/321,104 Continuation US11736954B2 (en) 2018-11-16 2021-05-14 Capacity planning method and apparatus

Publications (1)

Publication Number Publication Date
WO2020098575A1 true WO2020098575A1 (zh) 2020-05-22

Family

ID=70731444

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/116670 WO2020098575A1 (zh) 2018-11-16 2019-11-08 一种容量规划方法及装置

Country Status (4)

Country Link
US (1) US11736954B2 (zh)
EP (1) EP3873123B1 (zh)
CN (1) CN111200821B (zh)
WO (1) WO2020098575A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113873532A (zh) * 2021-09-02 2021-12-31 中通服咨询设计研究院有限公司 一种智慧园区5g网络规划方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114338428A (zh) * 2022-01-04 2022-04-12 国网山东省电力公司经济技术研究院 配电物联网业务流量预测方法及装置
CN116582891B (zh) * 2023-07-13 2023-10-17 上海铂联通信技术有限公司 5g无线网络系统的负载优化方法、装置、介质及终端

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101286897A (zh) * 2008-05-16 2008-10-15 华中科技大学 一种基于超统计理论的网络流量异常检测方法
CN101753381A (zh) * 2009-12-25 2010-06-23 华中科技大学 一种检测网络攻击行为的方法
CN103152749A (zh) * 2013-02-17 2013-06-12 无锡儒安科技有限公司 基于无线传感器网络流量统计的节点动态节能方法
CN103632209A (zh) * 2013-11-28 2014-03-12 国家电网公司 一种基于排队论的智能配用电业务数据传输带宽预测方法
US20160050571A1 (en) * 2013-05-02 2016-02-18 Huawei Technologies Co., Ltd. Network optimization method, and network optimization device

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1461907B1 (en) * 2001-11-28 2013-01-16 Millennial Net Network protocol for an ad hoc wireless network
WO2007032733A2 (en) * 2005-09-16 2007-03-22 Telefonaktiebolaget Lm Ericsson (Publ) Improved dimensioning methods for hsdpa traffic
CN102098135B (zh) * 2011-03-04 2013-06-05 北京邮电大学 用于异构网络融合下的业务流分发的系统及方法
CN102711129B (zh) * 2012-06-13 2018-08-03 南京中兴新软件有限责任公司 网络规划参数的确定方法及装置
US9674718B2 (en) * 2012-08-09 2017-06-06 Telefonaktiebolaget Lm Ericsson (Publ) Microwave link control
CN104160735B (zh) * 2013-02-21 2018-07-13 华为技术有限公司 发报文处理方法、转发器、报文处理设备、报文处理系统
US9800509B2 (en) * 2014-02-20 2017-10-24 Uvic Industry Partnerships Inc. System and method for efficient transport of large data files
CN108093411B (zh) * 2018-01-10 2021-11-16 重庆邮电大学 D2d通信网络中基于信道签名的资源调度优化方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101286897A (zh) * 2008-05-16 2008-10-15 华中科技大学 一种基于超统计理论的网络流量异常检测方法
CN101753381A (zh) * 2009-12-25 2010-06-23 华中科技大学 一种检测网络攻击行为的方法
CN103152749A (zh) * 2013-02-17 2013-06-12 无锡儒安科技有限公司 基于无线传感器网络流量统计的节点动态节能方法
US20160050571A1 (en) * 2013-05-02 2016-02-18 Huawei Technologies Co., Ltd. Network optimization method, and network optimization device
CN103632209A (zh) * 2013-11-28 2014-03-12 国家电网公司 一种基于排队论的智能配用电业务数据传输带宽预测方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3873123A4

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113873532A (zh) * 2021-09-02 2021-12-31 中通服咨询设计研究院有限公司 一种智慧园区5g网络规划方法
CN113873532B (zh) * 2021-09-02 2024-04-19 中通服咨询设计研究院有限公司 一种智慧园区5g网络规划方法

Also Published As

Publication number Publication date
CN111200821B (zh) 2021-12-03
EP3873123B1 (en) 2024-01-03
US11736954B2 (en) 2023-08-22
EP3873123A4 (en) 2022-01-26
US20210274359A1 (en) 2021-09-02
CN111200821A (zh) 2020-05-26
EP3873123A1 (en) 2021-09-01

Similar Documents

Publication Publication Date Title
CN109195170B (zh) 小区扩容方法、装置及存储介质
WO2019158102A1 (zh) 一种确定QoS描述信息的方法和装置
WO2020098575A1 (zh) 一种容量规划方法及装置
US10264549B2 (en) Method for indicating cell coverage enhancement mode and base station
US8675580B2 (en) Method and apparatus for facilitating packet scheduling for a hybrid communication network
US9565570B2 (en) Capacity planning method and device for wireless broadband network
WO2019191985A1 (zh) 一种信息发送、接收方法及装置
WO2020192363A1 (zh) 一种通信方法及设备
CN108834216B (zh) 一种资源调度方法及装置
TW201815189A (zh) 通訊方法和通訊裝置
RU2010136702A (ru) Способы и устройство внутрипользовательского планирования качества обслуживания восходящей линии связи
KR101749200B1 (ko) 무선 통신 네트워크의 허용 제어를 위한 방법 및 장치
US10887774B2 (en) Mobile communications system expansion method and device, storage medium, and program product
CN103313311A (zh) 一种无线资源管理方法、用户设备和网络侧
WO2013000240A1 (zh) 一种多节点联合的频谱感知方法和系统
WO2020215282A1 (en) Method and apparatus for evaluate data traffic depressed by radio issues
WO2020107155A1 (zh) 信道状态信息的获取方法、装置及计算机存储介质
EP2930617A1 (en) Resource management method and device
CN105532031A (zh) 资源优化的方法和装置
WO2017080251A1 (zh) 一种无线通信系统中用户终端速度估计的方法和装置
JP6142195B2 (ja) 無線通信システム、無線基地局、無線通信方法、及びプログラム
WO2019185065A1 (zh) 一种确定资源的方法及装置
WO2021259174A1 (zh) 随机接入资源的确定方法和自回传iab节点
CN115334525B (zh) 网络利用率计算方法和装置
WO2024114187A1 (zh) 接入网优化配置的方法、设备、装置及存储介质

Legal Events

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

Ref document number: 19885672

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2019885672

Country of ref document: EP

Effective date: 20210527