CN104468257A - Cloud application availability prediction method and system based on mobile user time-space behaviors - Google Patents

Abstract

The invention discloses a cloud application availability prediction method and system based on mobile user time-space behavior. The method includes the first step that a mobile intelligent device terminal sends a to-be-predicted cloud application selected by a user and use time t of the to-be-predicted cloud application, namely, a user use mode, to a mobile user time-space behavior data collector of the mobile intelligent device terminal; the second step that the mobile user time-space behavior data collector sends received user use mode information, electric quantity consumption statistical information of the mobile intelligent device terminal and network state mode statistical information to a could end; the third step that the cloud end calculates the availability of the to-be-predicted cloud application in a set time period according to the data sent by the mobile user time-space behavior data collector and sends a calculation result to the mobile intelligent device terminal. The cloud application availability prediction method and system based on the mobile user time-space behaviors can help a user to predict the availability of the cloud application on a mobile device of the user in a future time period, brings convenience to user application, and meanwhile helps a mobile cloud application provider or service provider to optimize service.

Description

Based on cloud application availability Forecasting Methodology and the system of mobile subscriber's time-space behavior

Technical field

The invention belongs to mobile internet technical field, be specifically related to a kind of Intelligent mobile equipment cloud application availability Forecasting Methodology based on the pattern analysis of user's time-space behavior and system, usability analyses and the prediction of cloud application on the mobile terminal devices such as smart mobile phone can be realized, mobile subscriber can be helped to predict the availability of cloud application on its mobile device of future time section, provide convenience according to the more reasonable comfortable use application of the custom of oneself for making user; Meanwhile, mobile cloud application provider or service provider is also helped to analyze its cloud application availability on the mobile apparatus, with Optimized Service according to user's time-space behavior custom.

Background technology

Cloud computing mode with the ease for use of its height, payable at sight and consumption mode and abundant infrastructure resources, just to change the mode that people use information service.Meanwhile, take smart mobile phone as the application software of mobile terminal by feature richness of representative, be deep into rapidly the every aspect in people's life, replacing conventional P C gradually becomes the mode using Internet service the most widely.Add up according to Chinese Network Information Centre, within 2013, China cell phone netizen has reached 500,000,000 people, accounts for Internet user's 81%.Along with the universal fast of mobile intelligent terminal and the covering on a large scale of mobile network, the Mobile solution sustainable growth that video download, instant messaging, mobile phone games and shopping at network etc. are multi-field.On the one hand, Internet service, the diversified media application enriched constantly in mobile terminal need to carry out the computation-intensive tasks such as natural language processing, speech recognition, image recognition, and this proposes requirements at the higher level to mobile device performance; On the other hand, there is again the restrictions such as calculate and storage resources is limited, battery life is limited, communication capacity is limited in mobile terminal.Mobile terminal ability to be combined with cloud computing mode and the mobile cloud computing proposed alleviates the contradiction of these two aspects to a certain extent.The resource process calculation task of mobile terminal cloud applications exploiting cloud computing platform, and communicate with mobile terminal or synchronous so that and user interactions, both make use of the advantages such as mobile terminal is ubiquitous, easy-to-use, the problem of resource requirement can have been solved again.According to market report, by 2016, the global expense being used for mobile cloud computing service will reach 64.7 hundred million dollars, and mobile cloud application extreme enrichment, for daily life brings huge facility.

But, compared with applying with conventional cloud, the application of mobile cloud has the features such as mobility, resource-constrained (as battery), unstable networks (as unavailable in wifi network, 3G signal blind spot etc.), its service availability is more fragile, and computational resource provides regular inefficacy or inaccessible with business service.And as the first element that QoS of customer is experienced, the service availability of cloud application directly reflects the service quality of cloud application, be the problem that user is concerned about the most, be also the emphasis that mobile cloud application provider optimizes its application product and service simultaneously.

Mobile terminal cloud application has following characteristics: 1) position dynamic: the dynamic change of position, mobile terminal directly affects the type of its network connection, as 3G network or different wifi connect, and different network connection types also brings cloud and applies different Quality of Service Experience, can ensure as stable wifi connects the high availability that media class cloud is applied; 2) temporal dynamic property: the service condition of mobile terminal cloud application and environment dynamic change in time, decisive factor is the behavior pattern of user, as user's conventional music class application before sleeping, then for music class be applied in sleep before the time period require high availability; 3) resource-constrained: the resource-constrained of mobile terminal is mainly reflected in the continuous electricity time, and the continuous electricity time is supposing that the custom that connected to the network under air time same case and user uses cloud to apply exists incidence relation; Therefore, mobile terminal cloud application availability is mainly subject to the impact of following factor: 1) network connects (relevant with geographical position), with the transfer in user geographical position, residing for user, network state shifts thereupon, and the change of network state directly affects mobile terminal cloud application availability and remote-effects mobile terminal mobile phone power consumption.2) user distributes (relevant with application using forestland) service time, and user uses the application of mobile cloud to have certain metastable Annual distribution, and this distribution directly affects the actual mobile cloud application availability experienced of user.3) the continuous electricity time (relevant with application using forestland), application power consumption and the dump energy in certain moment of application directly affect the availability that mobile cloud is applied, and meanwhile, these two factors are closely related with the using forestland of user.

Therefore, that applies along with mobile cloud enriches, need for mobile subscriber and cloud application provider provide a kind of usability analyses method and system, better Quality of Service Experience is obtained to make user, these make mobile cloud application provider improve its application availability becomes a urgent need to solve the problem, and carries out Predicting and analysis according to the time-space behavior pattern of mobile subscriber's personalization to the availability that cloud is applied and become a kind of new approaches.

Summary of the invention

For prior art Problems existing, the present invention proposes a kind of cloud application availability interpretation and application method and system based on the personalized time-space behavior pattern analysis of mobile subscriber, object is to experience for Intelligent mobile equipment user improves cloud application service quality, helps cloud application provider to optimize its application product.The present invention is described from Time and place two dimensions user behavior, time describes the use wish and custom that user applies cloud, the transfer of spatial description user locations and network environment, by the analysis of mobile subscriber's time-space behavior pattern of applying cloud, in conjunction with the change in time and space of energy consumption of mobile equipment state connected to the network, analysis and prediction cloud is applied in the availability of mobile terminal.

Technical scheme provided by the invention is:

A kind of mobile device end cloud application availability interpretation and application method and system, as shown in Figure 1.This system contains intelligent movable equipment end and high in the clouds two parts.

Intelligent movable equipment end main modular has: cloud application apparatus end, user interface, time-based cloud application using forestland logging modle, time-based electric quantity consumption monitoring and statistical module, network connection type monitoring and statistical module and mobile subscriber's time-space behavior data collector based on geographical position:

1) user interface, major function is to provide the reciprocal process of mobile terminal and mobile subscriber.User specifies the cloud application wishing analysis and prediction by interface, select wish to use the time period of cloud application; Meanwhile, user interface provides user to check the availability functionality that cloud is applied, and provides Data support to lower floor's time-based cloud application using forestland logging modle.

2) time-based cloud application using forestland logging modle, function is that record move user (obtained from user interface the wish service time that certain cloud is applied, specified by user), be add up in the cycle with a period of time, and statistics is passed to mobile subscriber's time-space behavior data collector.

3) time-based electric quantity consumption monitoring and statistical module, to user's mobile device end electricity Real-Time Monitoring, { time, battery}, this tuple represents the correspondence of certain moment and dump energy, is equivalent to a list item of relational database during storage to generate tuple.Meanwhile, test subscriber uses the power consumption speed of mobile cloud device procedures, finally monitoring result is passed to mobile subscriber's time-space behavior data collector.

4) based on network connection type monitoring and the statistical module in geographical position, Real-Time Monitoring is carried out to the network state of user and with one day for the cycle is added up, finally statistics is passed to mobile subscriber's time-space behavior data collector.The network state of user is divided into Four types: 3g network, wifi network, without network, available the while of 3g and wifi network.

5) mobile subscriber's time-space behavior data collector, collects user's using forestland from above three modules, electric quantity consumption and real-time condition and network state pattern, three is arranged and packed be sent to high in the clouds.

6) cloud application apparatus end, refers to the equipment end application program of the mobile cloud application of to be analyzed and monitoring, mainly with app client form, i.e. and the cloud application apparatus end chosen of user interface.This module, mainly as monitored object, directly reacts user behavior and cloud application performance.

Add up based on the time with upper module, for sake of convenience, give tacit consent to one day for the cycle, one day time slice.Set up the mapping of user network state with the time period, the monitoring result of every day is recorded as the number of times being in each mapping status.

Mobile high in the clouds facing cloud application provider, is responsible for carrying out analysis and prediction to user's time-space behavior pattern of mobile device end statistics, and mutual with mobile device end.Mobile high in the clouds comprises with lower module: cloud application high in the clouds, cloud application availability administration interface, cloud application availability model module, mobile subscriber's time-space behavior data sink, the cloud application availability interpretation and application module based on mobile subscriber's time-space behavior pattern:

1) cloud application availability administration interface, is mainly cloud application provider and provides visual user behavior pattern and availability situation, and availability predictions and analysis result are passed to user interface.

2) mobile subscriber's time-space behavior data sink, the data that the mobile subscriber's time-space behavior data collector accepting intelligent movable equipment end transmits, and by data preparation, pass to the cloud application availability interpretation and application module based on mobile subscriber's time-space behavior pattern.

3) cloud application availability model module, deposit simple cloud application availability model, and transmit naive model parameter to the cloud application availability interpretation and application module based on mobile subscriber's time-space behavior pattern, the such as probability of availability of four kinds of networks on physical condition (as disposed, quorum sensing inhibitor etc.).This module deposits the availability parameters irrelevant with user's time-space behavior, leaves high in the clouds in.

4) based on the cloud application availability interpretation and application module of mobile subscriber's time-space behavior pattern, receive parameters from mobile subscriber's time-space behavior data sink and cloud application availability model module, set up user's time-space behavior model and by certain algorithm, analysis and prediction carried out to cloud application availability.

5) cloud application high in the clouds, refers to that cloud application provides the high in the clouds part of service.This module is as a part for system, and its function is that same cloud application apparatus end is mutual, and considers by other modules the change in availability that its reciprocal process is brought.Meanwhile, cloud application provider can be optimized this module according to analysis result.

Compared with prior art, good effect of the present invention is:

The availability that native system and method can help user in predicting future time section mobile device end to apply, for the more reasonable comfortable use application of user is offered help; Native system and method can help cloud application provider to analyze the behavioural habits of customer group and the application availability of user's real experiences, help application provider's optimizing product.

Accompanying drawing explanation

Fig. 1 is the cloud application availability analysing and predicting system structural representation based on mobile subscriber's time-space behavior pattern of the present invention.

Fig. 2 is schematic flow sheet in embodiment.

Embodiment

Below by specific embodiments and the drawings, the present invention will be further described.

The present invention, based on the analysis to user's time-space behavior pattern, adopts monitoring and statistical, calculates and prediction cloud application availability.Main feature has following three aspects:

1, the focus of native system and method is that the availability that Intelligent mobile equipment end is applied, angle are from user behavior pattern analysis.

2, native system and method are described from Time and place two aspects user behavior, and the time describes user and uses wish and custom, the transfer of spatial description user locations and environment for use.

3, native system and method are by the monitoring of user behavior and wish and acquisition, and the main change in time and space obtaining Intelligent mobile equipment power consumption and network connection state, calculates the availability of the comparatively actual application of Intelligent mobile equipment end accurately.

Fig. 1 is the structural representation of the Intelligent mobile equipment cloud application availability prognoses system based on mobile subscriber's time-space behavior pattern of the present invention.Native system adopts modular mode to build, and carries out mutual and pass-along message between disparate modules by interface, has relative independence and accomplish the loose coupling of intermodule between each module.As shown in Figure 1, wherein mobile device end comprises cloud application mobile terminal, user interface, time-based cloud application using forestland logging modle, time-based electric quantity consumption detects and statistical module, based on network connection type monitoring and statistical module and mobile subscriber's time-space behavior data collector in geographical position; Mobile high in the clouds (mainly facing cloud application provider) comprises cloud application, cloud application availability administration interface, mobile subscriber's time-space behavior data sink, based on the cloud application availability interpretation and application module of mobile subscriber's time-space behavior pattern, cloud application availability model.Respectively different modules is specifically described below.

1. mobile device end

Mobile device end is statistics and the monitoring center of whole system, it achieves the quantitative and qualitative analysis statistical analysis to user's time-space behavior pattern, is delivered to the function in high in the clouds with the mutual of user and data packing.Divide module according to functional characteristic and design interaction path, the service condition apply cloud and environment carry out comprehensive Simultaneous Monitoring.

Mobile device end contains concrete support to this method and system core thought and embodiment, mainly comprise cloud application mobile terminal, user interface, time-based cloud application using forestland logging modle, time-based electric quantity consumption monitoring and statistical module, based on network connection type monitoring and statistical module and mobile subscriber's time-space behavior data collector six modules in geographical position.Modules all completes allomeric function by carrying out information interaction with other correlation modules and mobile high in the clouds.Set forth the major function of each module below.

1> user interface

User interface is that user is directly visual can operational module, and the major function provided comprises the selection function of user to the application of wish monitoring cloud, and user is to the appointed function of wish usage time interval and calculate gained usability results to the feedback function of user.

The realization of user interface, based on app interface, mainly comprises display layout and operation layout, mobile terminal operating system DLL (dynamic link library) can be used to realize.As in android system, textview can be used, the assemblies such as bar.

2> time-based cloud application using forestland logging modle

Time-based cloud application using forestland logging modle major function is that the cloud of specifying user is applied and the wish usage time interval of this application records (obtaining from user interface), and passes to mobile subscriber's time-space behavior data collector.

Recording method is by time slice, maintains two marks (as shaping variable) to each time period, represent respectively following this time period whether wish to use and in the past at the number of times that this time period wish uses.

The time-based electric quantity consumption monitoring of 3> and statistical module

Time-based electric quantity consumption monitoring and statistical module mainly realize two functions: measuring and calculating mobile terminal power consumption and record real time electrical quantity.Mobile terminal power consumption measuring and calculating point two parts, Part I is system power consumption, does not namely use the power consumption speed that mobile cloud is applied; Part II is power consumption speed when using the application of mobile cloud.The measuring and calculating way of mobile terminal cloud application power consumption speed has multiple, for android system, even has part Android flavor to carry application power consumption and detects.Conveniently way reads battery log file (wrapping as used internal in API), the cpu that also can read when the application of mobile cloud runs distributes the time, therefrom obtain the percentage of its service time, and represent with (total power consumption * percentage/total time).When user wishes system prediction availability, this module obtains current electric quantity and consigns to mobile subscriber's time-space behavior data collector.

Because power consumption is closely related with network state, need to calculate respectively the power consumption speed under heterogeneous networks state.In the network connection type monitoring based on geographical position with statistical module, network state is divided into Four types, then calculating speed has four values with network state one_to_one corresponding.

4> monitors and statistical module based on the network connection type in geographical position

Network connection type monitoring based on geographical position is network state with statistical module focus.This module is divided into four kinds network state: 3g, wifi, 3g with wifi and no 3g with no wifi, time slice, sets up record (table 1).

Table 1 is based on time-space behavior statistical form

The statistics number of table 1 hollow white square representative office under certain time period and certain network state.Can be that (being segmentation by the hour here) obtains current network state at set intervals to the monitoring of user network state, as having available API to judge network type in Android exploitation; Also can trigger be set, the change of monitoring network state.Here, the problem needing solution is unstable networks situation.When in the short time, network state continuously changes, usually have two reasons: the first, network signal is weak; The second, mobile device end breaks down.For both of these case, the result of use of mobile cloud application or card or fault, therefore directly will can be divided into no 3g with no wifi, i.e. network down state during this period of time.

5> mobile subscriber time-space behavior data collector

The major function of mobile subscriber's time-space behavior data collector is the data of collection three monitoring modulars, and it is sent to mobile subscriber's time-space behavior data sink in high in the clouds with certain data structure.This module bears two functions, to the packing function of data and the sending function of data flow.Meanwhile, this module in charge ensures the stable transfer of data, the transmission state of record Monitoring Data and realize retransmitting, and the function such as to resume.

6> cloud application mobile terminal

Cloud application mobile terminal is the monitoring of native system and method and the object of analysis, is the application program of intelligent movable equipment end.

2. move high in the clouds

Mobile high in the clouds Main Function is for cloud application provider provides the pattern analysis of user's time-space behavior and the cloud application availability based on user's time-space behavior pattern.It encompasses nearly all evaluation work, and calculate the cloud application availability of user's real experiences according to certain algorithm.Mobile high in the clouds comprises cloud application, cloud application availability administration interface, mobile subscriber's time-space behavior data sink, based on the cloud application availability interpretation and application module of mobile subscriber's time-space behavior pattern, and cloud application availability model.

1> mobile subscriber time-space behavior data sink

Mobile subscriber's time-space behavior data sink is as the interactive module of mobile high in the clouds with intelligent movable equipment end, and major function is exactly obtain the information of mobile subscriber time-space behavior data collector and in addition Simple Calculation and structuring.

The information that this module receives has four parts: cloud application using forestland record (time that time-based user intention access times and user specify wish to use), cloud application power consumption speed, current real time electrical quantity and the network type statistic record based on time-space behavior.As follows to the process computational methods of information:

Suppose: with one day for the cycle, the scope of time t is interval (0:01,24:00).User intention uses cloud Applicative time u (t) (value is the piecewise function of 0 or 1), and namely in t, if user intention uses, functional value is set to 1, otherwise is set to 0; Function Nu (t) represents the wish access times at the cloud user application of t, and here owing to being to time slice record, then function is piecewise function, and value is integer; System power consumption speed in heterogeneous networks situation is b ₁, b ₂, b ₃, b ₄; Cloud application power consumption speed is B ₁, B ₂, B ₃, B ₄; Current real time electrical quantity Bn; The number of times statistics being in different network type state in t is function Nn ₁(t), Nn ₂(t), Nn ₃(t), Nn ₄t (), these four functions are piecewise function equally and value is shaping.

Calculate: for function Nu (t), by hour in units of, we wish that user intention participates in availability calculations process as weighting parameters, and therefore we define probability density function f (t) and are:

$f\left(t\right)=\frac{\mathrm{Nu}\left(t\right)}{\underset{0}{\overset{24}{\∫}}\mathrm{Nu}\left(t\right)\mathrm{dt}},$

Meet here, the value of f (t) is determined by Nu (t).

For function Nn ₁(t), Nn ₂(t), Nn ₃(t), Nn ₄(t), equally with one day for the cycle, by hour in units of, we calculate based on the user network distributions probability density function of space-time:

$p_i\left(t\right)=\frac{{\mathrm{Nn}}_i\left(t\right)}{\underset{1}{\overset{4}{\mathrm{\Σ}}}{\mathrm{Nn}}_i\left(t\right)},$

Wherein { 1,2,3,4} meets i ∈

Finally, mobile subscriber's time-space behavior data sink transmits data u (t), f (t), B to the cloud application availability interpretation and application module based on mobile subscriber's time-space behavior pattern _i, b _i, Bn, p _i(t), wherein i ∈ { 1,2,3,4}.

2> cloud application availability model

The simple and easy availability calculations model of part preserved by cloud application availability model, and Main Function is to provide call parameter to the cloud application availability interpretation and application module based on mobile subscriber's time-space behavior pattern.Parameter mainly comprises: be in the mobile cloud application availability A under four kinds of network type states _i, wherein i ∈ { 1,2,3,4}; Availability (the high availability index as illustrated in the SLA SLA) As that cloud application provider ensures.

3> is based on the cloud application availability interpretation and application module of mobile subscriber's time-space behavior pattern

Based on the cloud application availability interpretation and application module of mobile subscriber's time-space behavior pattern as the core in mobile high in the clouds, major function is the parameter according to obtaining, analytical calculation cloud application availability by result feedback to cloud application provider.Wherein, usability analyses calculates and is described below:

First, the parameter that this module obtains from other modules is listed below:

User intention uses cloud Applicative time u (t); User arranges the moment t of wish service time ₀; Probability function f (t) of user intention frequency of utilization; Based on the user network distributions probability function p of space-time _i(t); System power consumption speed b _i; Cloud application power consumption speed B _icurrent real time electrical quantity Bn; Rely on the mobile cloud application availability A of network state _i; The availability indexes As that cloud application provider ensures; Wherein, i ∈ { 1,2,3,4}.Defined parameters P{battery_charge>0} represent mobile terminal battery electric quantity can probability.

Step1: obviously, p _it () and f (t) are separate, both simultaneous probability are p _i(t) * f (t), this probability of happening distribution function is:

$F_i\left(t\right)=\underset{0}{\overset{t}{\∫}}f\left(t\right)p_i\left(t\right)\mathrm{dt}$

With one day for the cycle, by hour in units of, the value that we calculate this function is as follows:

$F_i\left(24\right)=\underset{0}{\overset{t}{\∫}}f\left(t\right)p_i\left(t\right)\mathrm{dt}$

Below by it referred to as F _i.

Step2: suppose the moment t of user in setting wish service time ₀rear until finish using and can not charge.Calculate and be accustomed to trying to achieve expending electricity and up duration in theory according to user's requirement and user's time-space behavior, method following (representing with false code), wherein initializing variable t is t ₀.

Suppose that user wishes that the future time period predicted is T ₀(being here one day), above false code is finished and records gained moment t is a numerical value, represents in future time, according to user intention, predicts the moment that gained mobile device electricity exhausts or through the moment of a time cycle, is designated as T.Calculate time Tb that user before moment T wishes to use and user wish the total time hypothesis used be Tt (by hour in units of, both are numerical value):

$\mathrm{Tb}=\underset{t0}{\overset{T}{\∫}}u\left(t\right)\mathrm{dt}$

$\mathrm{Tt}=\underset{t0}{\overset{t0+T0}{\∫}}u\left(t\right)\mathrm{dt}$

Calculate P{battery_charge>0}=Tb/Tt

Step3: the final computational methods of availability for cloud application are as follows:

We represent mobile cloud application availability with Probability Forms, availability comprises two aspects: mobile cloud application system availability A _mobilewith mobile terminal battery electric quantity availability A _battery.

In Step2, the calculating of P{battery_charge>0} is by the P of representative of consumer behavioural habits _ijointly determine with power consumption speed, therefore A _batterydirectly can get P{battery_charge>0} as rational characterising parameter.

For A _mobile, we know in Step1 and calculate gained F _idescribe for different four kinds of network states, the using forestland of a user and user intention comprehensive probability of availability in time (can be understood like this, within the scope of one day, user intention uses and satisfied i-th the network-like probability of state in residing geographical position), when calculating user intention f (t) to it in time to interval (0,1) map, calculate network state p residing for user _itime (t), it is spatially mapped to interval (0,1), namely at a time by four kinds of network states with probability description, so when calculating F _itime, we to its in time integration obtain the probability of availability based on user's time-space behavior pattern.Mathematically have:

$\underset{1}{\overset{4}{\mathrm{\Σ}}}{F}_i=1$

We provide brief logical and prove:

Due to $\underset{1}{\overset{4}{\mathrm{\Σ}}}{p}_i\left(t\right)=1,$ Therefore t at any time, has $f\left(t\right)=\underset{1}{\overset{4}{\mathrm{\Σ}}}{p}_i\left(t\right)*f\left(t\right),$ Have again $\underset{0}{\overset{24}{\∫}}f\left(t\right)\mathrm{dt}=1,$ So integration is carried out on peer-to-peer both sides:

$\underset{0}{\overset{24}{\∫}}f\left(t\right)\underset{1}{\overset{4}{\mathrm{\Σ}}}{p}_i\left(t\right)\mathrm{dt}=1$

According to the addition property of integration, we extract superposition number:

$\underset{1}{\overset{4}{\mathrm{\Σ}}}\underset{0}{\overset{24}{\∫}}f\left(t\right)p_i\left(t\right)\mathrm{dt}=1$ That is:

$\underset{1}{\overset{4}{\mathrm{\Σ}}}{F}_i=1$

Card is finished.

A _irepresent the availability of four kinds of networks on physical condition (as disposed, quorum sensing inhibitor etc.), these parameters obtain from cloud application availability model module.A _i* F _irepresentative is based on the usable probability under certain network condition of user behavior pattern, and we also need the availability indexes As that consideration cloud application provider ensures simultaneously, final A _mobilebe calculated as follows formula:

$A_{\mathrm{mobile}}=\mathrm{As}*\underset{i=1}{\overset{4}{\mathrm{\Σ}}}{A}_i{F}_i$

Finally, our comprehensive mobile cloud application system availability A _mobilewith mobile terminal battery electric quantity availability A _battery,thus obtain the final mobile cloud application availability based on user's time-space behavior pattern:

$\begin{array}{c}A=A_{\mathrm{mobile}}*A_{\mathrm{battery}}\\ =\mathrm{As}*\underset{i=1}{\overset{4}{\mathrm{\Σ}}}{A}_i{F}_i*A_{\mathrm{battery}.}\end{array}$

4> cloud application availability administration interface

Cloud application availability administration interface Main Function is for cloud application provider provides visualized data and result.Meanwhile, availability predictions and analysis result are passed to user interface by it, with thinking that user offers suggestions the services such as propelling movement.

5> cloud is applied

Cloud application module is the analytic target of native system and method, and it is the high in the clouds part of paid close attention to cloud application, and Main Function is user and provides service, has certain function.

Below by instantiation, whole system flow process is described:

As shown in Figure 2, first, user selects cloud application to be predicted by user interface to whole system flow process, and specifies the time wishing to use the application of this cloud in the following time.Time-based cloud application using forestland logging modle obtains the information record that user arranges, and passes to mobile subscriber's time-space behavior data collector afterwards.Meanwhile, time-based electric quantity consumption is monitored with statistical module and is monitored the information of monitoring with statistical module separately to the transmission of mobile subscriber's time-space behavior data collector and recording based on the network connection type in geographical position.

The data packing obtained arranges by mobile subscriber's time-space behavior data collector, by Internet Transmission to high in the clouds, and monitors and fault-tolerant processing transmitting procedure.When transfer of data is complete, mobile subscriber's time-space behavior data collector carries out simple process and calculating to data, simultaneously, its parameter of preserving of cloud application availability model preparation, afterwards, the every data needed for availability calculations and parameter transmission are given the cloud application availability interpretation and application module based on mobile subscriber's time-space behavior pattern by two modules.

Based on the cloud application availability interpretation and application module of mobile subscriber's time-space behavior pattern according to the data obtained, calculated based on the cloud application availability of user's time-space behavior pattern by parser, and result is transferred to cloud application availability administration interface.Cloud application availability administration interface can integrated multiple function, and mainly mutual with cloud application provider, meanwhile, result is transferred to user interface, this process is also a network transmission process, and data volume is less.

An application example is provided below.

In daily life, video/audio class moves cloud application as a kind of integration of Internet resources, the extreme enrichment life of people.But the mobile cloud application environment for use based on the behavior pattern of user is but subject to each side restriction, such as network environment, battery electric quantity etc.Like this, in a lot of situation, user wishes but cannot use when using cloud application, causes cloud application availability greatly by user's conditioning.

We wish that on intelligent family moving platform, set up a video/audio class moves cloud application availability analytical system, be intended to for user provides the usability analyses and forecast model be accustomed to for self, for user's reasonable arrangement uses cloud Applicative time and place to offer help; Meanwhile, we wish that cloud application provider can obtain the actual availability of its cloud application service by the analysis to user behavior pattern and mobile cloud application availability, and analyze user behavior pattern distribution, provide Data support for optimizing cloud application service.

Based on this method and system, move the feature of cloud application for video/audio class, we have made two aspect amendments to system.The first, the use of video/audio class cloud application requires very high to fluency, and fluency is one of decisive factor of availability.Therefore, 3g network state, arranging in cloud application availability model, divides into unavailable by we, and concrete reason is two aspects: first, and 3g network signal and level of coverage do not reach the requirement seeing video; Secondly, economically see, use 3g flow downloading video files cost too high, therefore almost nobody uses 3g network to see video.The second, the application of general video/audio class cloud provides download caching function, so, if but network state unavailable local existing buffer memory time, cloud application still can be used user.Like this, we add buffer memory push function, namely after specifying the time wishing to use this application user, system is according to the migration and variation of user network state in a day, find out user to wish to use application but the network state unallowed time period, shift to an earlier date PUSH message to user and inform that user should download or caching audio files in advance.

This application in design for video/audio class move cloud application characteristic build a set of entirety availability parameters collect, interpretation and application scheme.By the monitoring to user behavior pattern, for user better experience cloud application offer suggestions push and Shi Yun application provider for customer group behavior pattern design service provision, cloud application availability is got a promotion in user level.

Above embodiment is only in order to illustrate technical scheme of the present invention but not to be limited; those of ordinary skill in the art can modify to technical scheme of the present invention or equivalent replacement; and not departing from the spirit and scope of the present invention, protection scope of the present invention should be as the criterion with described in claim.

Claims (10)

1., based on a cloud application availability Forecasting Methodology for mobile subscriber's time-space behavior, the steps include:

1) the cloud to be predicted application that user is chosen of intelligent movable equipment end and service time t, i.e. user's using forestland, sends to mobile subscriber's time-space behavior data collector of this intelligent movable equipment end;

2) the user's using forestland information that will receive of described mobile subscriber's time-space behavior data collector, electric quantity consumption statistical information and the network state mode statistical information of this intelligent movable equipment end send to high in the clouds;

3) described high in the clouds calculates according to the data that this mobile subscriber's time-space behavior data collector sends over the availability that described cloud to be predicted is applied in setting usage time interval, and result of calculation is sent to this intelligent movable equipment end.

2. the method for claim 1, is characterized in that described electric quantity consumption statistical information comprises system power consumption counting rate information, cloud application power consumption speed and electricity Real-Time Monitoring two tuple { time, battery}; Wherein, time represents certain moment, and battery represents dump energy.

3. the method for claim 1, is characterized in that described network state mode statistical information comprises the network state statistical information of each time period in one-period; Wherein network state comprises: 3g network can be used, and wifi network can be used, and can use without network, available the while of 3g and wifi network.

4. the method as described in claim 1 or 2 or 3, is characterized in that described high in the clouds is according to formula calculate probability density function f (t), and according to formula calculate a user network distributions probability density function p _i(t); Wherein, nu (t) represents the wish access times at the cloud user application of t, and T1 is cycle service time, and the number of times statistics being in i-th network type state in t is function Nn _it (), network state adds up to n.

5. method as claimed in claim 4, is characterized in that described availability calculations method is:

51) according to formula calculate the mobile cloud application system availability A of this intelligent movable equipment end _mobile, and calculate the battery electric quantity availability A of this intelligent movable equipment end _battery;

52) according to formula A=A _mobile* A _batterycalculate cloud to be predicted described in each be applied in user select predicted time in availability;

Wherein, A _battery=Tb/Tt, Tt are the total time that user wishes to use, and Tb is that the user before the intelligent movable equipment electricity of prediction exhausts wishes service time; The availability indexes that As ensures for cloud application provider, A _ithe probability of availability in i-th kind of network state is in, F for this intelligent movable equipment end _ifor user intention within the scope of one-period uses and satisfied i-th the network-like probability of state in residing geographical position, and n is network state sum.

6., based on a cloud application availability prognoses system for mobile subscriber's time-space behavior, it is characterized in that comprising the intelligent movable equipment end and high in the clouds that are connected by network; Described intelligent movable equipment end comprises time-based cloud application using forestland logging modle, time-based electric quantity consumption monitoring and statistical module, based on network connection type monitoring and the statistical module in geographical position, and mobile subscriber's time-space behavior data collector; Described high in the clouds comprises mobile subscriber's time-space behavior data sink; Wherein,

Described time-based cloud application using forestland logging modle, for cloud to be predicted application that user is chosen and service time t, i.e. user's using forestland, sends to described mobile subscriber's time-space behavior data collector;

Described time-based electric quantity consumption monitoring and statistical module, for the electricity of intelligent movable equipment end described in Real-Time Monitoring, generate electric quantity consumption statistical information, and send it to described mobile subscriber's time-space behavior data collector;

The described monitoring of the network connection type based on geographical position and statistical module, for the network state pattern of intelligent movable equipment end described in Real-Time Monitoring, generating network state model statistical information, and send it to described mobile subscriber's time-space behavior data collector;

Described mobile subscriber's time-space behavior data collector, for the user's using forestland information that will receive, the electric quantity consumption statistical information of this intelligent movable equipment end and network state mode statistical information send to described mobile subscriber's time-space behavior data sink;

The data that described high in the clouds receives according to described mobile subscriber's time-space behavior data sink, calculate the availability that described cloud to be predicted is applied in setting usage time interval, and result of calculation are sent to this intelligent movable equipment end.

7. system as claimed in claim 6, is characterized in that described electric quantity consumption statistical information comprises system power consumption counting rate information, cloud application power consumption speed and electricity Real-Time Monitoring two tuple { time, battery}; Wherein, time represents certain moment, and battery represents dump energy.

8. system as claimed in claim 6, is characterized in that described network state mode statistical information comprises the network state statistical information of each time period in one-period; Wherein network state comprises: 3g network can be used, and wifi network can be used, and can use without network, 3g and wifi network available four kinds of states simultaneously.

9. the system as described in claim 6 or 7 or 8, is characterized in that described mobile subscriber's time-space behavior data sink is according to formula calculate probability density function f (t), and according to formula calculate a user network distributions probability density function p _i(t), and by p _it (), f (t) send to described high in the clouds; Wherein, nu (t) represents the wish access times at the cloud user application of t, and T1 is cycle service time, and the number of times statistics of i-th network type state is function Nni (t), and network state adds up to n.

10. system as claimed in claim 9, is characterized in that described high in the clouds is according to formula calculate the mobile cloud application system availability A of this intelligent movable equipment end _mobile, and calculate the battery electric quantity availability A of this intelligent movable equipment end _battery; According to formula A=Amobile*Abattery calculate cloud to be predicted described in each be applied in user select predicted time in availability; Wherein, Abattery=Tb/Tt, Tt are the total time that user wishes to use, and Tb is that the user before the electricity of prediction exhausts wishes service time; The availability indexes that As ensures for cloud application provider, A _ithe probability of availability in i-th kind of network state is in, F for this intelligent movable equipment end _ifor user intention within the scope of one-period uses and satisfied i-th the network-like probability of state in residing geographical position, and n is network state sum.