CN104994581A - Multi-mode wireless resource scheduling method based on SARS mechanism - Google Patents

Abstract

The invention discloses a multi-mode wireless resource scheduling method based on an SARS (service-awarded resource scheduler) mechanism, concretely comprising: in a wireless interface mode, a multi-mode user wireless MT (Media Terminal) obtaining a wireless bandwidth RP (Resource Pool) better than single selecting wireless signals through the virtual scheduling of an SARS server in obtaining wireless video business, and forming a new wireless VB (virtual bus); meanwhile the MT capturing parameters such as self user business characteristics, video content popularity, and signals of present multi-mode system base stations to generate an SAF (Service-awared Factor) set, and sending SARS messages to an SARS server, the SARS server performing cyclic scheduling on SARS requirements, and distributing mapped multi-mode system base stations (including WIFI), thereby effectively improving the perception satisfaction degree of wireless on-line video users, and realizing load balancing of base station working load.

Description

A kind of multi-mode wireless resource regulating method based on SARS mechanism

Technical field

The present invention relates to a kind of a kind of scheduling method for wireless resource based on SARS mechanism for carrying out during the access to netwoks of 3G/4G user radio, belonging to wireless communication across layers field.

Background technology

In current moving communicating field, 3G/4G pattern can be adopted to get online without being tethered to a cable, because 3G standard has four kinds of patterns, China have employed 3 kinds of 3G patterns wherein, and current 4G network is also in carrying out building and distribute telephone numbers, and traditional speech data still carries out at use 2G channel, therefore in the near future, 3G will be there is simultaneously, multiple types 3G, multiple types 4G the wireless communication technology of depositing, also the generation of multi mode terminal terminal has been expedited the emergence of thus, the multi-mode radio data link of operator, the multi-mode support of manufacturer, such as five mould terminals, five moulds are exactly that a chip supports TD-LTE, FDD-LTE, TD-SCDMA, WCDMA, the communication pattern that GSM five kinds is different moves 3G/4G/WiMAX/WIFI multiplex roles, and be known as at present " the whole network leads to ", the network formats of all operators can be supported, comprise movement, UNICOM, telecommunications, 2G, 3G, 4G, each pattern due to five mould terminals is divided into again 13 frequency ranges, therefore be otherwise known as " five moulds 13 frequently ".

Problem just creates, the operating characteristic of current multi-modal mobile phone, is substantially all that under namely opening present mode, other patterns are all in standby or resting state, except non-user is manually selected to switch to required mode in switching preference pattern.This just means, under selected pattern, the Radio Resource of other patterns is not effectively discharged.

Current and following wireless user's main flow is wireless video on-demand and two-way video call, larger wireless bandwidth must be needed, current operator with the satisfaction of user for the best examines key element, therefore fixed mode connection, and the switching of fixed cellular community, still can bring the speed shake in single cellular cell and the switching sluggishness across cellular cell.

Current physical facility, there is multiple 3G/4G aerial array the cellular cell of a steel tower covering, such as

3G：CDMA EVDO 800；

3G：TD-SCDMA B34/39；

3G：WCDMA B1/8；

4G：TD-LTE B38/39/40/41；

4G：FDD-LTE B1/3MHz；

Coverage is not from 3 kilometers ~ 10 kilometers etc.

WIFI, IEEE 802.11n/b/g, coverage 5 meters ~ 50 meters is not etc.

With regard to the use of wireless user; along with the attention of the copyright protection of Internet video; do not reoffer the local download service of video; hot broadcast video is gradually to online charging order video; a large amount of videos is along with the change of popularity in a network by dynamic program request, and the user perceptibility of the order video that user's pop degree is high requires to improve.And 3G/4G business is also for real-time radio video call service in the future proposes new requirement, i.e. real-time (Real time) and Bandwidth guaranteed (Bandwidth Guaranteed).

Because ISP (Internet Service Provider) progressively carries out evolution to ICP (Internet Content Provider), and future network is also software defined network (SDN) and the content delivery networks (CDN) of customer-centric, therefore from the content of current video request program, popularity is theoretical: user's request content meets " 2/8 " law according to statistics, namely user's request of 80% concentrates on the content of 20%, especially the film of hot broadcast and TV play, find thus, although the behavior of customer access network content has randomness, but meet certain Mathematical Modeling, be defined as content popularit.

With regard in current wireless resource scheduling technology, under all concentrating on multi-user's competition, how allocation optimized is carried out to the Radio Resource that current base station provides, to the self attributes of user, the concrete traffic performance of user does not carry out deep excavation and perception, focuses in the link properties of concrete wireless link.

Summary of the invention

Technical problem to be solved by this invention is that wireless user is under present mode work, when cannot obtain bandwidth needed for user, can by starting SARS (Service-awared Resource Scheduler, service-aware scheduling of resource) service, collect other mode wireless base stations optional, thus obtain flow compensated bandwidth.And under part scene, by the prediction of community residence time, the maximum effective bandwidth resource of user can be distributed to.

The present invention is for solving the problems of the technologies described above by the following technical solutions:

A kind of multi-mode wireless resource regulating method based on SARS mechanism, by starting service-aware scheduling of resource SARS service, collect other mode wireless base stations optional by SARS server, thus be the wireless terminal acquisition flow compensated bandwidth of user, concrete steps are as follows:

The media termination MT of steps A 1, user selects active wireless network, and carries out registration and the connection of wireless link;

Steps A 2, when the bandwidth resources of user to active wireless network are unsatisfied with, this user is registered to SARS server by MT, and enters next step;

Steps A 3, search for other optional wireless network by user MT, and the ability sequence of the alternative network that sorted according to signal power;

Steps A 4, the acquisition current different kinds of business of user MT and perceptual parameters thereof, for video stream media business, also need to catch its popularity parameter, generates customer service perceptual parameters SAF collection;

The rate bandwidth AB of steps A 5, acquisition user MT current wireless interface, and rate bandwidth EB corresponding to the different kinds of business that steps A 4 obtained, the SAF collection that the ability sequence of the alternative network obtained together with steps A 3, steps A 4 obtain, is sent to SARS server by single or continuous print SARS request message;

Steps A 6, SARS server according to obtain the ability of the provided resource of base station is provided, the user sending SARS request message is sorted, then scheduling of resource distribution is carried out to ordering user, and be returned as by SARS message the scheduling of resource mapping scheme RSS that each user MT formulates;

Steps A 7, user MT obtains the Radio Resource of having authorized according to RSS, start new Radio Resource bandwidth, supplementing and gain as wireless network selected by steps A 1.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, the acquisition process of customer service perceptual parameters SAF collection is as follows:

Step B1, carry out the priority classification of SARS service registry user;

Step B2, acquisition user present type of service;

Step B3, carry out user's current available wireless network search and can providing capability sequence;

Step B4, carry out the popularity classification of user video content, obtain the popularity basic parameter of Streaming Media.

Further, multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, in step B1, the priority classification of SARS service registry user adopts grade of service agreement SLA pattern, 1 byte is taken time in SARS message body, namely 8 grades of classification are adopted, from 0 ~ 7 distribution, on SARS server, register authority of obtaining the authorization by MT;

Its wireless signal strength need be recorded respectively, as the first minor sort that follow-up wireless network provides SARS to serve when carrying out available wireless network search in step B3.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, whether in step B4, MT terminal needs to identify that current network accesses content is that streaming media video plays FLV service, and step is as follows:

Step B4.1, caught the URL of HTTP request in real time by MT, and record is as the entry condition of Detection of content;

Whether step B4.2, to be detected in wireless IP message flow have FLV to indicate by MT, if continuous print IP all has FLV to indicate in flowing, then tentatively regarding as this URL is the streaming medium content that user selects;

Step B4.3, judging that this URL and FLV flows whether more than certain time, as exceeded, then recording this URL, as a part for SARS request message.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention is add SARS request message again after selected URL is first generated HASH hash in step B4.3.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, in steps A 6, SARS server sorts to the user sending SARS request message, specifically comprises:

SAF parameter entrained by the SARS request message that step C1, SARS server send the multiple MT users received continuously is resolved, and { SLA, SerType, URL} gather to extract basic parameter; Record other parameters simultaneously, comprise { NetType, IsWork, Intensity}, wherein, SLA represents grade of service agreement, and SerType represents COS, and URL represents network address, NetType represents network type, and whether IsWork representative can work, Intensity representation signal intensity;

Step C2, employing two-dimensional coordinate axle, using SLA and SerType parameter as X, Y-axis parameter, adopts formula: tentatively sort, obtain the preliminary priority sequence of SARS request message in the T cycle;

Step C3, be the user of Online Video to type of service, by comparing to the popularity of its SAF parameter, and be worth PopularValue according to the popularity returned and carry out sequence adjustment.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, in steps A 6, SARS server carries out scheduling of resource distribution to ordering user, is specially:

D.1, SARS server by the connection of the RPC base station controller of each wireless network, the provided radio resource capacity collection of current all wireless base stations is provided;

D.2, to the resource capability that each base station reports assess, prevent overload or slow down the possibility entering overload, the base station of sequence overload simultaneously;

D.3, add up current available optional wireless network base station, the optional network specific digit sequence that respective user is submitted to is filtered, only retain user can base station sequence number;

D.4, active user's average satisfaction is calculated: $\mathrm{SATIS}=\frac{\mathrm{\ΣL}\left(i\right)*\mathrm{Satis}\left(i\right)}{\mathrm{\ΣL}\left(i\right)}=\frac{\mathrm{\ΣL}\left(i\right)*\mathrm{AB}\left(i\right)/\mathrm{EB}\left(i\right)}{\mathrm{\ΣL}\left(i\right)},$

Wherein AB (i) is the bandwidth distributing to user i, and EB (i) is the bandwidth that user i expects; L (i) is user i acquisition priority length; Satis (i) is the satisfaction of user i;

D.5, to ordering user carry out resource mapping, the bandwidth obtained as each user's actual allocated is:

$B\left(i\right)=\frac{L\left(i\right)*[\mathrm{EB}\left(i\right)-\mathrm{AB}\left(i\right)]}{\mathrm{\ΣL}\left(i\right)}\mathrm{\Σ\δ}\left(n\right)*\mathrm{BS}\left(n\right);$

Wherein, BS (n) is optional base station, the acailable factor that δ (n) is this base station;

D.6, by SARS return messages return to MT, MT selects the bandwidth resources of base station and correspondence to carry out wireless network speedup according to obtained acailable factor δ (n);

D.7 the SARS of the sequence request message in next round cycle, is processed.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, also comprises the dynamic prediction by the residence time of MT in cellular cell, distributes to the maximum effective bandwidth resource of user, is specially:

E.1, for the LBS being ready to provide position-based to serve, by the residence time in the movement velocity of MT in current area and direction of motion estimation Zhong community;

E.2, to base station property be AP, whether prediction residence time is greater than Preset Time, SARS server complete D.6 step time, by data flow priority allocation at WIFI, and the radio network information channel originally selected only does SARS control message channel.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, step C3 concrete steps are as follows:

C3.1, URL or the HASH parameter value extracted in SARS request message;

Whether C3.2, Query Database content have existed this HASH value, if existed, then return corresponding popularity parameter PopularValue, popularity time PopularTime are increased progressively 1 simultaneously; If there is no this URL, then this URL juxtaposition PopularTime=1 newly-increased; Wherein PopularValue is defined between (0,100);

C3.3, to sort seasonal SerType (new)=SerType (old) * (1+PopularValue) user, then again substitute into formula recalculate L (i) sequence, and sort according to L (i), the user MT of identical L (i) is postponed by request time order.

Further, the multi-mode wireless resource regulating method based on SARS mechanism proposed by the invention, step D.2 in the resource capability that each base station reports is assessed, evaluation points computing formula is as follows:

$\mathrm{Capa}\left(\mathrm{BS}\left(n\right)\right)=\mathrm{\α}\·\frac{\mathrm{RR}\left(\mathrm{BS}\left(n\right)\right)}{\frac{2}{3}\·\mathrm{AR}\left(\mathrm{BS}\left(n\right)\right)}+\mathrm{\β}\·S\left(\mathrm{UEm}\right),(\mathrm{\α}+\mathrm{\β}=1)$

Wherein, RR (BS (n), t) be the remaining stock number of t base station BS (n), AR (BS (n)) is the total stock number of base station BS (n), the satisfaction that S (UEm) is user UEm;

The base station of sequence overload simultaneously, exclusion standard is:

$(\mathrm{RR}\left(\mathrm{BS}\left(n\right)\right)>\frac{2}{3}\mathrm{AR}\left(\mathrm{BS}\left(n\right)\right)).$

The present invention adopts above technical scheme compared with prior art, has following technique effect:

1, change the single system work of multimode terminal MT, make full use of the available resources of other heterogeneous wireless base stations, the Radio Resource bandwidth that can reach user's optimal satisfaction is provided;

2, send the sequence of SARS request message according to MT, to optional wireless base station capability set, carry out the virtual mapping of capability set;

3, to providing the user of LBS to carry out residence time prediction in community, in predicted time, the optimum bandwidth of best base station is provided to distribute.

Accompanying drawing explanation

Fig. 1 is allomeric function application scenarios figure of the present invention.

Fig. 2 is the work overview flow chart of SARS service.

Fig. 3 is that MT detects user video business process graph.

Fig. 4 is the interaction sequences figure between MT and SARS server and each BS.

Fig. 5 is that SARS server asks user to sort to SARS.

Fig. 6 is the process figure of SARS server to video popularity.

Fig. 7 is SARS server virtual distributing radio resource procedure chart.

Motion residence time computation model in Tu8Shi MT community.

Embodiment

Below in conjunction with accompanying drawing, technical scheme of the present invention is described in further detail:

The invention belongs to cross-layer radio communication, by the multi-mode antenna resource making full use of user MT and have, effectively bandwidth is improved to current carried out business (mainly Online Video play service), thus the fluency of raising user, definition and user satisfaction.

As Fig. 1 describes, the present invention is applicable to support multimode (dual-mode handset on application scenarios, the whole network leads to) terminal MT (Media Terminal) is under fixed wireless work pattern, start SARS service, by to own service perception, current can the perception of connecting wireless network ability, and the parameter such as the popularity of video content generates SAF collection, SARS server is arrived by SARS service request information, distribute and scheduling resource after SARS RQ cycle being sorted by SARS server, for each SARS user distributes a new Radio Resource pond RP (Resource Pool) better than single selected wireless signal, an i.e. virtual wireless bus VB (Virtual Bus), thus effectively improve the user awareness satisfaction of wireless online video, and make base station maneuvering load reach load balancing.

Specifically be implemented as follows:

The concrete startup service process of A, SARS service, as shown in Figure 2:

Steps A 1, user manually select current 3G/4G network to be Master pattern, and carry out registration and the connection of wireless link, wireless link concrete herein connects relevant to concrete base station signal pattern, for maximum possible does not destroy the link setup agreement of original wireless link.

Steps A 2, user are when carrying out the individual business of oneself, such as WAP browses, when Online Video program request, video calling, if be unsatisfied with wireless bandwidth resource current, user selects to run and starts SARS bandwidth enhancement service processes, and prerequisite is that this MT have registered SARS service.

If steps A 3 starts SARS service, then open other optional 3G/4G/WIFI networks of search that current MT supports, and complete the power according to signal, recommend to get the stronger BS signal of front 3 signals.

The service-aware parameter of steps A 4, acquisition active user MT, it is such as pure WAP page browse service, or game on line business, if carry out video stream media business, also need to catch its popularity, generate current business perceptual parameters SAF (Service-awared Factor, service-aware parameter) collection.

The rate bandwidth AB (Assigned Bandwidth) of steps A 5, acquisition MT current wireless interface, and rate bandwidth EB (Excepted Bandwidth) corresponding to the different kinds of business that step 4 obtained, the SAF collection that the stronger BS network of front 3 signals obtained together with steps A 3, steps A 4 obtain, is sent to SARS server by single or continuous print SARS request message.

Steps A 6, SARS server according to obtain the ability of the provided resource of base station is provided, carry out scheduling of resource distribution to ordering user, the concrete scheduling of resource mapping scheme RSS (Resource Scheduling Solution) of generation is exactly the virtual bus VB made to measure for every 1 MT.

The Radio Resource scheme that steps A 7, user MT authorize according to VB in the SARS message received, start new Radio Resource bandwidth, close not by the wireless BS interface of use, the Radio Resource started is for supplementing and gain the original Master of formulation wireless mode, current business all can be switched under special circumstances, Master signal only retains controlling functions.

The acquisition that B, customer service perceptual parameters SAF (Service-awared Factor) collect is the SARS user property to current MT, the data mining of concrete type of service, video content and analysis, the acquisition of SAF parameter set is based on the extraction of the content of user MT, and MT can only select normal priority carry out the speed-raising of Radio Resource:

The priority of step B1, SARS service registry user adopts the grade of service to contract SLA mould (Service Level Agreement) formula, 1 byte is taken time in message body, namely adopt 8 grades of classification, from 0 ~ 7 distribution, on SARS server, register obtained authorization privilege by MT.

Step B2, acquisition for present type of service, only comprise 3 grades at present, only comprise 3 grades at present: WAP browses, Online Video program request, and two-way video is conversed.Due to what intercept based on IP message, if having Streaming Media mark in browsing WAP, be also judged to be Online Video program request.

Step B3, user's current network are searched for and can providing capability be sorted, the multiple wireless mode that the quantity of current alternative network is supported by MT terminal determined, the conventional priority of available network is WIFI, 4G, 3G, after web search, also wireless signal strength will be recorded, as the first minor sort that follow-up wireless network provides SARS to serve, message format is { NetType (network type), IsWork (whether can work), Intensity (signal strength signal intensity) }.

The popularity classification of step B4, user video content, obtain the popularity basic parameter of Streaming Media, as shown in Figure 3, in the type of service of application SARS service, main is be exactly the video playback of Streaming Media, whether MT terminal needs to identify that current network accesses content is Streaming Media FLV (Flash Video) service, and step as shown in Figure 3.

Step B4.1, MT catch the URL of HTTP request in real time, and record is as the entry condition of Detection of content;

Whether step B4.2, MT detect in wireless IP message flow has FLV to indicate, all has FLV to indicate, then tentatively regard as the streaming medium content that this URL may like for user if flowed in (encapsulation of UDP message) by continuous print IP;

Step B4.3, judging that this URL and FLV flows and whether continue more than 5 minutes, as exceeded, then recording this URL, as a part for SARS request message; As being less than 5 minutes, then think that this user loses the interest flowed this FLV, or be one section of commercial advertisement video flowing;

Step B4.4, for ensureing the user privacy Cookie of user (be equivalent to reveal) of URL, after selected URL can being generated HASH hash, add SARS request message.

User's queuing strategy of C, SARS server, reciprocal process as shown in Figure 4, specifically describes as follows:

Step C1, SARS server receive the SARS request that many MT user sends, and carry out user's qualification and mandate, and reply " service confirms ".

Step C2, also to change along with the change of the movement position of MT due to the wireless network gone out of MT, therefore SARS request message intermittently can send supplemental messages, in time current network state is upgraded, SARS server needs to ask user to sort to the SARS of up-to-date networking state, as shown in Figure 5.

Step C.3, SARS server resolves the SAF parameter entrained by the SARS request message received continuously, and { SLA, SerType, URL} collection records other parameters simultaneously, comprises { NetType, IsWork, Intensity} to extract basic parameter.

Step C4, preliminary sequence adopt two-dimensional coordinate axle, select the SARS request message in the T cycle, with SLA and SerType for X, and Y-axis parameter, $L\left(i\right)=\sqrt{\mathrm{SLA}{\left(i\right)}^2+\mathrm{SerType}{\left(i\right)}^2},$ As shown in Figure 5.

Step C5, Online Video also to be needed by carrying out popularity comparison to the Popularity of SAF parameter, and carry out sequence adjustment according to the PopularValue returned, as Fig. 6 signal, carry out the fine setting of sorting according to popularity.

Step C5.1, URL or the HASH parameter value extracted in SARS request message;

Step C5.2, inquire about whether data-base content has existed this HASH value, if existed, then return corresponding popularity parameter PopularValue, simultaneously because this URL accessed mistake once, the popularity time, PopularTime increased progressively 1; If there is no this URL, then increase this URL newly and PopularTime=1;

Step C5.3, be defined in (0 due to PopularValue, 100) between, then will when user sorts by SerType (new)=SerType (old) * (1+PopularValue), then the formula of step C4 is again substituted into, recalculate L (i) sequence, and sort according to L (i), the MT user of identical L (i) is postponed by request time order.

D, SARS, to the allocation schedule method of wireless isomer resource, as shown in Figure 7, are specially:

D.1, SARS server knows the provided radio resource capacity collection of current all wireless base stations by the connection of the RPC base station controller (RPCX n, n=1 ~ 7) of each wireless network.

D.2, to the resource capability that each base station reports assess, evaluation points computing formula is as follows:

$\mathrm{Capa}\left(\mathrm{BS}\left(n\right)\right)=\mathrm{\α}\·\frac{\mathrm{RR}\left(\mathrm{BS}\left(n\right)\right)}{\frac{2}{3}\·\mathrm{AR}\left(\mathrm{BS}\left(n\right)\right)}+\mathrm{\β}\·S\left(\mathrm{UEm}\right),(\mathrm{\α}+\mathrm{\β}=1)$

Wherein RR (BS (n), t) be t base station BS n surplus yield, AR (BS (n)) is base station BS (n) (1≤n≤N) total resources, S (UEm) is user UEm satisfaction, desirable α=0.6, β=0.4.

The base station of sequence overload simultaneously, exclusion standard is:

$(\mathrm{RR}\left(\mathrm{BS}\left(n\right)\right)>\frac{2}{3}\mathrm{AR}\left(\mathrm{BS}\left(n\right)\right)).$

D.3, add up current available optional wireless network base station, calculate residue base station according to base station ability evaluation points expression formula

Capa(BS(n))，

And sort from high to low according to result of calculation, generate current SAR S server available resources sequence { BS (n) };

The optional network specific digit sequence that respective user is submitted to, filter, introduce acailable factor δ (n), δ (n)=0 represents that the base station of the n-th sequence number is not in optional network specific digit sequence, and δ (n)=1 represents that the base station of user n-th sequence number can be used.Therefore the base station sequence for subsequent use of final each user is { δ (n) * BS (n) } (1≤n≤N).

D.4, user satisfaction is calculated, as unique user only consider individual satisfaction Satis (i)=AB (i)/EB (i) wherein AB (i) be the bandwidth distributing to user i, EB (i) be user i expect bandwidth;

As system, the average satisfaction weighing all users is just meaningful,

$\mathrm{SATIS}=\frac{\mathrm{\ΣL}\left(i\right)*\mathrm{Satis}\left(i\right)}{\mathrm{\ΣL}\left(i\right)}=\frac{\mathrm{\ΣL}\left(i\right)*\mathrm{AB}\left(i\right)/\mathrm{EB}\left(i\right)}{\mathrm{\ΣL}\left(i\right)},$

Wherein L (i) is for user i is in step C5.3 acquisition priority length.

D.5, to ordering user carry out resource mapping, need to be divided into two kinds of situations, when SARS server available bandwidth is greater than bandwidth on demand, then for each user i directly returns the bandwidth EB (i) of maximum satisfaction;

And actual conditions often can not meet all desired bandwidth EB, the bandwidth that therefore each user's actual allocated also can increase in time is: $B\left(i\right)=\frac{L\left(i\right)*[\mathrm{EB}\left(i\right)-\mathrm{AB}\left(i\right)]}{\mathrm{\ΣL}\left(i\right)}\mathrm{\Σ\δ}\left(n\right)*\mathrm{BS}\left(n\right).$

D.6, by SARS return messages return to MT, MT selects to determine that the bandwidth resources of that base station and correspondence carry out wireless network speedup according to obtained δ (n).

D.7 the SARS of the sequence request message in next round cycle, is processed.

The dynamic prediction of E, MT residence time in cellular cell;

E.1, position-based is provided to serve LBS (Location Based Service) for being ready, can pass through the movement velocity of MT in current area and the residence time in direction of motion estimation Zhong community, thus better bandwidth service is provided, as shown in Fig. 8 model residence time, this service is mainly useful in current area motion slowly, and wishes the user obtaining optimum bandwidth resource.

E1.1, average speed prediction to user terminal MT:

${\tilde{v}}_t=\underset{i=1}{\overset{t}{\mathrm{\&Sigma;}}}(1-\mathrm{\&gamma;}){\mathrm{\&gamma;}}^{i-1}{v}_{t-i+1},(0<\mathrm{\&alpha;}<1)$

Wherein, the MT average speed of t prediction; v _tthe actual speed of MT at moment t; γ: smoothing factor, 0< γ <1;

E1.2, the direction of motion of user MT to be predicted:

$E\left(\mathrm{\&alpha;}\right)=\underset{i=1}{\overset{t}{\mathrm{\&Sigma;}}}(1-\mathrm{\&beta;}){\mathrm{\&beta;}}^{i-1}{\mathrm{\&alpha;}}_{t-i+1},(0<\mathrm{\&beta;}<1)$

Wherein, E (α): UE in the predicted motion direction of moment t; α _t: UE is in the actual motion direction of moment t; β: smoothing factor, 0< β <1; K: k moment before current time t.

E1.3, the movement locus estimated in Mathematical Modeling as shown in Figure 8, if this base station cell radius is r, the translational speed of user MT is v; Moving direction and the angle between user and base station line are α, 0≤α≤π; Distance between user MT and base station is r ₁, 0≤r ₁≤ r; Variable d represents mobile subscriber's total displacement in the cell, 0≤d≤2r;

$h=\left\{\begin{array}{c}{r}_{1}sin\mathrm{\&alpha;},0\&le;\mathrm{\&alpha;}\&le;\mathrm{\&pi;}/2\\ r_{1}sin(\mathrm{\&pi;}-\mathrm{\&alpha;})=r_{1}sin\mathrm{\&alpha;},\mathrm{\&pi;}/2\&le;\mathrm{\&alpha;}\&le;\mathrm{\&pi;}\end{array}\right.$

Comprehensive is h=r ₁sin α, 0≤α≤π;

E1.4, displacement d in user MT community can be calculated thus

$d=\sqrt{r^2-h^2}+r_{1}cos\mathrm{\&alpha;},0\&le;\mathrm{\&alpha;}\&le;\mathrm{\&pi;};$

The speed of the user of E1.5, integrating step E1.1 and E1.2 and the prediction in direction, can obtain final Target cell residence time estimation equation is:

$\begin{array}{c}{t}_s=\frac{d}{v}=\frac{\sqrt{r^2-{r_1}^2{\sin }^2\left(E\left(\mathrm{\&alpha;}\right)\right)}+r_1*\cos \left(E\left(\mathrm{\&alpha;}\right)\right)}{\underset{i=1}{\overset{k}{\mathrm{\&Sigma;}}}(1-\mathrm{\&gamma;}){\mathrm{\&gamma;}}^{i-1}{v}_{t-i+1}}\\ =\frac{\sqrt{r^2-{r_1}^2{\sin }^2\left(\underset{i=1}{\overset{t}{\mathrm{\&Sigma;}}}(1-\mathrm{\&beta;}){\mathrm{\&beta;}}^{i-1}{\mathrm{\&alpha;}}_{t-i+1}\right)}+r_1*\cos \left(\underset{i=1}{\overset{t}{\mathrm{\&Sigma;}}}(1-\mathrm{\&beta;}){\mathrm{\&beta;}}^{i-1}{\mathrm{\&alpha;}}_{t-i+1}\right)}{\underset{i=1}{\overset{k}{\mathrm{\&Sigma;}}}(1-\mathrm{\&gamma;}){\mathrm{\&gamma;}}^{i-1}{v}_{t-i+1}}\end{array}$

Wherein 0≤γ≤1,0< β <1.

E.2, to base station property be AP, prediction residence time >5 minute, SARS server complete D.6 step time, by data flow priority allocation at WIFI, and original Master channel only does SARS control message channel.

F, part SARS message illustrate:

BIP (Bandwidth Increased Pacing) message is a special SARS message, is that MT is unsatisfied with current distributing radio resource, and request SARS server is slightly hurried up and increased some wireless bandwidths in original distribution basis.

It is λ that definition BIP increases progressively demand, first setting parameter λ=10%, if when SARS distributes bandwidth in step D5, this user does not stop paying out and send BIP message, parameter L (i) '=L (i) * (1+ λ) of this user, makes the priority of this user have small size raising, certainly, if all use sends BIP message per family, then there is not the effect of raising speed in final result when equal available bandwidth is constant.So increase BIP message mainly for for MT in Online Video, by the current dedicated bandwidth of the continuous small elevation of BIP message, improve user's visual experience.

Thus, the present invention proposes a kind of user MT based on multi-mode interface under current multi-modal wireless base station environment, devise the SARS server with cross-layer scheduling function, what can obtain the Radio Resource of multiple base station provides bandwidth resources simultaneously, under the environment of wireless MAC access not changing various complicated standard, employing has perception User Priority attribute, perception user multiple business type, the attributes such as the popularity of perception user streaming media service content, for the MT worked under single standard provides effective allocated bandwidth gain.SARS server, also to the motor pattern of MT in community and predict residence time, provides good solution for selecting optimum network signal.The present invention, when 3G, 4G base station signal overlap of WIFI soon and multi-modulation scheme, experiences with the optimal user obtained under load balancing condition and provides effective processing scheme.

The above is only some embodiments of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (10)

1. the multi-mode wireless resource regulating method based on SARS mechanism, it is characterized in that, by starting service-aware scheduling of resource SARS service, collecting other mode wireless base stations optional by SARS server, thus be the wireless terminal acquisition flow compensated bandwidth of user, concrete steps are as follows:

The media termination MT of steps A 1, user selects active wireless network, and carries out registration and the connection of wireless link;

Steps A 2, when the bandwidth resources of user to active wireless network are unsatisfied with, this user is registered to SARS server by MT, and enters next step;

Steps A 3, search for other optional wireless network by user MT, and the ability sequence of the alternative network that sorted according to signal power;

Steps A 4, the acquisition current different kinds of business of user MT and perceptual parameters thereof, for video stream media business, also need to catch its popularity parameter, generates customer service perceptual parameters SAF collection;

The rate bandwidth AB of steps A 5, acquisition user MT current wireless interface, and rate bandwidth EB corresponding to the different kinds of business that steps A 4 obtained, the SAF collection that the ability sequence of the alternative network obtained together with steps A 3, steps A 4 obtain, is sent to SARS server by single or continuous print SARS request message;

Steps A 6, SARS server according to obtain the ability of the provided resource of base station is provided, the user sending SARS request message is sorted, then scheduling of resource distribution is carried out to ordering user, and be returned as by SARS message the scheduling of resource mapping scheme RSS that each user MT formulates;

Steps A 7, user MT obtains the Radio Resource of having authorized according to RSS, start new Radio Resource bandwidth, supplementing and gain as wireless network selected by steps A 1.

2. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 1, is characterized in that, the acquisition process of customer service perceptual parameters SAF collection is as follows:

Step B1, carry out the priority classification of SARS service registry user;

Step B2, acquisition user present type of service;

Step B3, carry out user's current available wireless network search and can providing capability sequence;

Step B4, carry out the popularity classification of user video content, obtain the popularity basic parameter of Streaming Media.

3. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 2, is characterized in that,

In step B1, the priority classification of SARS service registry user adopts grade of service agreement SLA pattern, takies 1 byte, namely adopts 8 grades of classification, from 0 ~ 7 distribution, register authority of obtaining the authorization by MT on SARS server time in SARS message body;

Its wireless signal strength need be recorded respectively, as the first minor sort that follow-up wireless network provides SARS to serve when carrying out available wireless network search in step B3.

4. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 2, is characterized in that, whether in step B4, MT terminal needs to identify that current network accesses content is that streaming media video plays FLV service, and step is as follows:

Step B4.1, caught the URL of HTTP request in real time by MT, and record is as the entry condition of Detection of content;

Whether step B4.2, to be detected in wireless IP message flow have FLV to indicate by MT, if continuous print IP all has FLV to indicate in flowing, then tentatively regarding as this URL is the streaming medium content that user selects;

Step B4.3, judging that this URL and FLV flows whether more than certain time, as exceeded, then recording this URL, as a part for SARS request message.

5. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 4, is characterized in that, is to add SARS request message again after selected URL is first generated HASH hash in step B4.3.

6. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 1, is characterized in that, in steps A 6, SARS server sorts to the user sending SARS request message, specifically comprises:

SAF parameter entrained by the SARS request message that step C1, SARS server send the multiple MT users received continuously is resolved, and { SLA, SerType, URL} gather to extract basic parameter; Record other parameters simultaneously, comprise { NetType, IsWork, Intensity}, wherein, SLA represents grade of service agreement, and SerType represents COS, and URL represents network address, NetType represents network type, and whether IsWork representative can work, Intensity representation signal intensity;

Step C2, employing two-dimensional coordinate axle, using SLA and SerType parameter as X, Y-axis parameter, adopts formula: tentatively sort, obtain the preliminary priority sequence of SARS request message in the T cycle;

Step C3, be the user of Online Video to type of service, by comparing to the popularity of its SAF parameter, and be worth PopularValue according to the popularity returned and carry out sequence adjustment.

7. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 6, is characterized in that, in steps A 6, SARS server carries out scheduling of resource distribution to ordering user, is specially:

D.1, SARS server by the connection of the RPC base station controller of each wireless network, the provided radio resource capacity collection of current all wireless base stations is provided;

D.2, to the resource capability that each base station reports assess, prevent overload or slow down the possibility entering overload, the base station of sequence overload simultaneously;

D.3, add up current available optional wireless network base station, the optional network specific digit sequence that respective user is submitted to is filtered, only retain user can base station sequence number;

D.4, active user's average satisfaction is calculated: $\mathrm{SATIS}=\frac{\mathrm{\&Sigma;L}\left(i\right)*\mathrm{Satis}\left(i\right)}{\mathrm{\&Sigma;L}\left(i\right)}=\frac{\mathrm{\&Sigma;L}\left(i\right)*\mathrm{AB}\left(i\right)/\mathrm{EB}\left(i\right)}{\mathrm{\&Sigma;L}\left(i\right)}.$

Wherein AB (i) is the bandwidth distributing to user i, and EB (i) is the bandwidth that user i expects; L (i) is user i acquisition priority length; Satis (i) is the satisfaction of user i;

D.5, to ordering user carry out resource mapping, the bandwidth obtained as each user's actual allocated is:

$B\left(i\right)=\frac{L\left(i\right)*[\mathrm{EB}\left(i\right)-\mathrm{AB}\left(i\right)]}{\mathrm{\&Sigma;L}\left(i\right)}\mathrm{\&Sigma;\&delta;}\left(n\right)*\mathrm{BS}\left(n\right);$

Wherein, BS (n) is optional base station, the acailable factor that δ (n) is this base station;

D.6, by SARS return messages return to MT, MT selects the bandwidth resources of base station and correspondence to carry out wireless network speedup according to obtained acailable factor δ (n);

D.7 the SARS of the sequence request message in next round cycle, is processed.

8. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 7, is characterized in that, also comprise the dynamic prediction by the residence time of MT in cellular cell, distributes to the maximum effective bandwidth resource of user, is specially:

E.1, for the LBS being ready to provide position-based to serve, by the residence time in the movement velocity of MT in current area and direction of motion estimation Zhong community;

E.2, to base station property be AP, whether prediction residence time is greater than Preset Time, SARS server complete D.6 step time, by data flow priority allocation at WIFI, and the radio network information channel originally selected only does SARS control message channel.

9. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 6, it is characterized in that, step C3 concrete steps are as follows:

C3.1, URL or the HASH parameter value extracted in SARS request message;

Whether C3.2, Query Database content have existed this HASH value, if existed, then return corresponding popularity parameter PopularValue, popularity time PopularTime are increased progressively 1 simultaneously; If there is no this URL, then this URL juxtaposition PopularTime=1 newly-increased; Wherein PopularValue is defined between (0,100);

C3.3, to sort seasonal SerType (new)=SerType (old) * (1+PopularValue) user, then again substitute into formula recalculate L (i) sequence, and sort according to L (i), the user MT of identical L (i) is postponed by request time order.

10. a kind of multi-mode wireless resource regulating method based on SARS mechanism according to claim 7, is characterized in that, step D.2 in the resource capability that each base station reports is assessed, evaluation points computing formula is as follows:

$\mathrm{Capa}\left(\mathrm{BS}\left(n\right)\right)=\mathrm{\&alpha;}\&CenterDot;\frac{\mathrm{RR}\left(\mathrm{BS}\left(n\right)\right)}{\frac{2}{3}\&CenterDot;\mathrm{AR}\left(\mathrm{BS}\left(n\right)\right)}+\mathrm{\&beta;}\&CenterDot;S\left(\mathrm{UEm}\right)(\mathrm{\&alpha;}+\mathrm{\&beta;}=1)$

Wherein, RR (BS (n), t) be the remaining stock number of t base station BS (n), AR (BS (n)) is the total stock number of base station BS (n), the satisfaction that S (UEm) is user UEm;

The base station of sequence overload simultaneously, exclusion standard is:

$(\mathrm{RR}\left(\mathrm{BS}\left(n\right)\right)>\frac{2}{3}\mathrm{AR}\left(\mathrm{BS}\left(n\right)\right)).$