CN103248500A - Real-time on-demand data broadcast scheduling method in consideration of size of data item - Google Patents

Abstract

The invention provides a real-time on-demand data broadcast scheduling method in consideration of sizes of data items, adopts a split idea to process the situation that the sizes of the data items are different, and additionally provides a scheduling method for equal-size data item split (ES-LxRxW) and a scheduling method for unequal-size data item split (US-LxRxW), split stragetries of equal-size data items and adjustment strategies of deadlines, and split strategies of unequal-size data items and adjustment strategies of unequal-size data items of two kinds, i.e. PF and PxBF. Both the scheduling method for the equal-size data items and the scheduling method for the unequal-size data items provided by the invention consider bandwidth and period. A split standard size for the split of the equal-size data items is provided aiming at distribution of different deadlines and distribution of different data item sizes and a split standard size for the split of the unequal-size data items is provided aiming at different data item sizes and different scheduling strategies, so that the real-time situation can be well adapted.

Description

A kind of data broadcast scheduling method as required in real time of considering the data item size

Technical field

Data broadcast scheduling of the present invention field is specifically related to data broadcast scheduling method under a kind of mobile environment, especially relates to a kind of real-time cycle data broadcasting as required of considering.

Background technology

The dispatching algorithm of data broadcasting is most important, most crucial content in the data broadcasting.Data broadcast scheduling algorithm refers to study in the broadcasting process which data of this selection is broadcasted away, and the algorithm of the priority problem of broadcasting data.In general, be divided into the broadcasting of static cycle data, data broadcasting and mix three kinds of broadcasting as required.This section is broadcasted, data broadcasting as required, is mixed the focus that broadcasting and four parts of index are introduced current research respectively from static cycle data.

(1) static data broadcasting.Major part has been studied static broadcast mode, and wherein the algorithm of influence power maximum is the polydisc dispatching algorithm, and it has proposed the data item that access probability is big and has been put into than upward broadcasting of dish faster, and the data item that access probability is smaller is put on the slow dish to be broadcasted.Azer Bestavros has proposed the AIDA algorithm based on the polydisc dispatching algorithm, and this algorithm has improved promptness and serious forgiveness, allows the multipriority of better control data item simultaneously.Sanjoy Baruah etc. change the real-time task system at wheel and have proposed to improve dispatching method based on the another kind of AIDA polydisc dispatching algorithm, defined the model of real-time fault tolerance broadcasting disk, experimental result shows that the design of designated broadcast disk in this model and the scheduling that wheel changes the real-time task system are closely-related.

(2) data broadcasting as required.Because static broadcasting can not in time respond user's request, so on-demand broadcasting becomes hot topic gradually.Ping Xuan etc. has proposed the BoD model, and it can be broadcasted timely according to user's request, has proposed the EDF algorithm also with having experimental results show that its validity, and experimental result shows, the time-constrain of the satisfied request that BoD can be successful.Demet Aksoy etc. has clearly defined the performance of dispatching algorithm under large-scale on-demand broadcasting environment and the standard of scale, the RxW algorithm has been proposed, it shows functional in all these standards, can well adapt on the demand of a specific application or system; RxW is strong in user's colony and the change of workload, and scheduling selects because the access probability that it is based on the state of current request formation rather than estimated data's item is done.Qiu Fang etc. has studied traditional non-moving real-time policy and non real-time shift strategy in great detail, has proved that traditional real time algorithm not necessarily still is that performance is optimum in mobile environment; Proposed based on pull-type real time broadcasting system, and proposed a dispatching algorithm ACR efficiently, it is in order in time to transmit data to the mobile terminal and to design.The quantity that Jianliang Xu etc. considers the urgency of time and do not finish request has proposed a kind of new dispatching algorithm SIN-α, and a kind of effective realization of this algorithm proposed, also analyzed the theoretical boundary that the speed that arrives when request is increased to request failure rate when infinite in the literary composition, experimental results show that SIN-α is better than existing algorithm in most operating load, and when high request rate close to the theoretical boundary of analyzing out.Jun Chen etc. has studied the request of multidata item in the broadcasting of data as required of being with time restriction, algorithm DTIU has been proposed, it combines the advantage of data item scheduling and request scheduling, The performance test results shows, this algorithm has not only reduced the request failure rate, save the bandwidth of broadcast channel again, and under influence of various factors, be better than other classic algorithm.Joseph Kee-Yin Ng etc. has proposed algorithm MRS and it is with the mutation of buffer memory real-time transmission data in the real time information dispatching patcher.This algorithm family has considered the off period of request, the popularizations big or small and data in scheduling decision of data object.Experiment showed, that this algorithm can not only broadcast data on time, be better than traditional dispatching algorithm 20% aspect the response time simultaneously.Kai Liu has studied the multinomial purpose request scheduling of multichannel under the on-demand broadcasting environment, proposes algorithm DUP, has solved the problem of asking hungry problem and bandwidth availability ratio, and experimental results show that the superiority of algorithm.Jingsong Lv etc. has proved that the multinomial request of broadcast scheduling in real time is np problem under single channel and multichannel environment, two algorithm PVC and SSA based on " profit " have been proposed, be respectively at single channel and multi channel scheduling, simulation result shows, compare with traditional algorithm, great improvement is arranged.

(3) mix broadcasting.Mix broadcasting refer to static periodic broadcasting and as required data broadcast a kind of broadcast scheduling algorithm that combines, it is a part of hot spot data periodic broadcasting, the non-hot spot data on-demand broadcasting of another part.Propositions such as Sunho Kim, statistics collection client's request, data item is divided into hot and cold two data sets carries out mixed scheduling, the regular broadcasting of data based its access probability of heat, cold data based stand-by period and uncompleted request number are broadcasted.According to average response time, the best separation of the cold and hot data item of having analyzed and researched, and use the emulation of Web data query to prove the validity of scheme.JungHwan Oh has inquired into a kind of method of ADAPTIVE MIXED, only broadcast relatively more popular data, transmit data not too commonly used to save the network bandwidth in the mode of program request, a kind of new technology has been proposed, with integrated mode serve as broadcasting and the unicast allocated bandwidth, simulation result shows that this technology significantly is better than recent technology.

(4) index.The purpose of introducing index is in order to reduce the energy consumption of mobile device, because after the adding index, mobile device checks that index just can know when the data item that oneself needs is broadcasted, thereby switch to park mode during this period of time what wait for, save the energy, when data item is come, just switch to listen mode and go to receive data.Because the limitation of the mobile device energy, a lot of all index in the mobile data broadcasting of research in recent years.Wherein, major part all is the index in the static broadcasting of research, and Jianliang Xu proposes a kind of index with parameter----index index, can optimize access delay under the condition of the boundary of given tuning period, and vice versa.Jianliang Xu has proposed a new index data structure Dtree, is used to provide the inquiry of the data that the position is relevant in the mobile environment, and experimental result shows that Dtree has the better overall performance than existing structure.The research of the data as required broadcasting of a part all is that Swarup Acharya has proposed algorithm MAX according to the standard of optimizing single request under the worst case at the algorithm of broadcast scheduling, and this algorithm is done well in experiment.Demet Aksoy introduces a parametrization algorithm, all shows well on all evaluation criterions, and can adjust time under the average or the worst wait situation, has also proposed simulation model in the literary composition.Minority with indexes applications in the broadcasting of data as required.Jiun-Long Huang proposed one adapt to index and data organize algorithm AIDOA, can adjust the length of index dynamically, experimental result has also proved the validity of this algorithm.

Consult pertinent literature, before finding research in real time as required the document of data broadcast scheduling all be to be under the consistent situation at the data item size, the validity of research dispatching algorithm.And for the setting of data item, all be to make a search under for fixing situation in tentation data item size.But in practice, the data item size of request is inconsistent, the situation of very big data item may occur, the situation of very little data item also may occur, and the distribution of data item size also has different situations.As user request if webpage or text, the data item size probably has only several KB so; If request is video or big pictorial information, data item just may have the size of several MB or tens MB so.

Owing to technical equipment or wireless transmission medium, bandwidth size under different environment is different simultaneously.Can only reach 10Kbps as bandwidth in the 2G environment, and amount of bandwidth can reach 10Mbps in the 3G environment, bandwidth can reach 100Mbps in the 4G environment.And for 3G network, from technical standpoint, 3G network only is the broadband network of last kilometer use wireless access, and its actual bandwidth depends on signal strength signal intensity, and network capacity.

In data broadcasting before, each data item only could be other mobile subscriber's services after broadcasting is finished.If broadcast big data item, under the certain situation of bandwidth, so the time of this data item busy channel long, make that the number of users of service reduces in the unit interval, service quality reduces.Equally, if data item size is certain, under the less situation of bandwidth ratio, the time of data item busy channel also is long.

Summary of the invention

At the research of broadcasting for real-time data as required in the prior art, all be the validity of data broadcast scheduling algorithm, the data item size that all is hypothesis is fixing situation, but all differing of data item size do not taken into account, and do not consider bandwidth and the factor that varies in size in cycle in the different environment; The present invention proposes a kind of data broadcast scheduling method as required in real time of considering the data item size.

Technical scheme of the present invention is 1. 1 kinds of data broadcast scheduling methods as required in real time of considering the data item size, it is characterized in that: comprise following process,

Mobile client is by after uploading channel and submitting the request of oneself to, server receives request, check earlier this request whether in broadcast queue or prepare in the formation, if, then directly will ask to add in the broadcast queue or in the preparation formation in the request queue of corresponding data subitem; If do not exist, then will ask to press request time order from small to large and add in the request queue of system, and be request distribution ID;

Server takes out the request of current time from the request queue of system, search the data item that satisfies request, obtains groups of data items and is made into data item information, and described data item information comprises the ID of data item and the content of data item;

Server splits according to the data item that the big young pathbreaker of data item gets access to, and the data subitem that splits into is added in the preparation formation; A data item splits into an above data subitem, and the data item ID of these data subitems is identical, but data subitem ID difference; Preparing formation at first arranges by the ID of data item, the data subitem of same data item ID is arranged by data subitem ID, and a current request of data item is kept at the request queue that splits each data subitem of gained, and the data subitem of preparing in the formation is called the preparation item;

Server is periodically dispatched tissue by dispatching algorithm to the data of preparing in the formation, selects the data subitem to add in the broadcast queue dynamically, and the data subitem in the broadcast queue is broadcast, and the data subitem in the broadcast queue is called the broadcasting item.

And employing waited the size data item to split mode when the data item that gets access to was split, and the fractionation and the scheduling flow that wait the size data item to split mode may further comprise the steps,

Step a1, initialization;

Step a2 is if not request of current time then jumps to step a7, otherwise continues to enter step a3;

Step a3, the request of taking-up current time;

Step a4 obtains the data item of this request, according to the fractionation strategy that waits the size data item it is split, and according to the adjustment strategy of off period the off period of the data item of fractionation is adjusted;

Step a5 handles respectively for all the data subitems after splitting, and processing mode is as follows,

To arbitrary data subitem wherein, if at the broadcasting item of broadcast queue and do not have broadcasting, then this request is added in the request queue of this broadcasting item; Otherwise, if the preparation item in the preparation formation then adds this request in the request queue of this preparation item; Otherwise, this data subitem is added the preparation formation as new preparation item, and will ask to add and prepare newly to prepare in the formation in the request queue of item;

Step a6 waits new request to be generated, jumps to step a2;

Step a7 broadcasts according to broadcast queue;

Step a8 judges whether just to have broadcasted not empty broadcasting item in the broadcast queue, otherwise directly enters step a9, be that the successful solicited message that then will broadcast item writes successfully the solicited message file, delete this broadcasting all requests, delete this broadcasting item, enter step a9 then;

Step a9 judges whether that broadcast queue is empty, if otherwise directly enter step a10, if broadcast queue is empty, then generate new broadcast queue according to the dispatching algorithm that waits size to split data item, enter step a10 then;

Step a10, the time point that advances if surpassed the running time of regulation, then jumps to step a11; Otherwise, jump procedure a2;

Step a11 reads successfully the solicited message file, the statistics solicited message.

And, split according to the fractionation strategy that waits the size data item among the step a4, comprise following substep,

Step a4.1, initialization comprises the big or small remainsize=datasize of residue, subitem sequence number subitemorder=1 are set; Datasize is the data item size;

Step a4.2, it is as follows to determine to split benchmark size SplitStandardSize,

$\mathrm{SplitS}\tan \mathrm{dardSize}=\frac{{\mathrm{Bw}}_k\×{\mathrm{Cy}}_k}{\mathrm{SplitK}}$

Wherein, Bw _kBe the broadcast bandwidth of the k time broadcast cycle, Cy _kIt is the broadcast cycle of the k time broadcast cycle; SplitK is a positive integer, and the data item size of the interior broadcasting of expression one-period is the number of the data subitem of fractionation benchmark size SplitStandardSize;

Step a4.3 if remainsize＞SplitStandardSize then continues step a4.4, otherwise jumps to step a4.5;

Step a4.4, add a data subitem, data subitem size is SplitStandardSize, filling size is SplitStandardSize, sequence number is subitemorder, subitemorder=subitemorder+1, remainsize=remainsize-SplitStandardSize jumps to step a4.3;

Step a4.5 adds a data subitem, and data subitem size is SplitStandardSize, and filling size is remainsize, and sequence number is subitemorder, remainsize=0;

Step a4.6 returns and splits good data subitem.

And, generate new broadcast queue according to the dispatching algorithm that waits size to split data item among the step a9, comprise following substep,

Step a9.1, import the k time broadcasting before the preparation formation after enter circulation,

Step a9.2 upgrades and prepares each data subitem d in the formation _{X, y}Current request number, off period and stand-by period;

Step a9.3 calculates each data subitem d _{X, y}Dispatching priority;

Step a9.4, the data subitem d that will have the greatest priority value _{X, y}Be designated as d _{N, m}And compose and give Bd _j

Step a9.5 judges that all data subitem sizes add d in the broadcast queue _{N, m}Size whether greater than Bw _k* Cy _kIf, greater than then withdraw from circulation and enter step a9.8,, otherwise continue execution in step a9.6;

Step a9.6 is with Bd _jInsert the B of broadcast queue _kAfterbody returns step a9.2;

Step a9.7, end loop;

Step a9.8 returns the B of broadcast queue the k time _k

Perhaps, adopt when the data item that gets access to is split and do not wait the size data item to split mode, the fractionation and the scheduling flow that do not wait the size data item to split mode may further comprise the steps,

Step b1, initialization;

Step b2 is if not request of current time then jumps to step b7, otherwise continues to enter step b3;

Step b3, the request of taking-up current time;

Step b4 obtains the data item of this request, according to the fractionation strategy that does not wait the size data item it is split, and according to the adjustment strategy of off period the off period of the data item of fractionation is adjusted;

Step b5 handles respectively for all the data subitems after splitting, and processing mode is as follows,

To arbitrary data subitem wherein, if at the broadcasting item of broadcast queue and do not have broadcasting, then this request is added in the request queue of this broadcasting item; Otherwise, if the preparation item in the preparation formation then adds this request in the request queue of this preparation item; Otherwise, this data subitem is added the preparation formation as new preparation item, and will ask to add and prepare newly to prepare in the formation in the request queue of item;

Step b6 waits new request to be generated, jumps to step b2;

Step b7 broadcasts according to broadcast queue;

Step b8 judges whether just to have broadcasted not empty broadcasting item in the broadcast queue, otherwise directly enters step b9, be that the successful solicited message that then will broadcast item writes successfully the solicited message file, delete this broadcasting all requests, delete this broadcasting item, enter step b9 then;

Step b9 judges whether that broadcast queue is empty, if otherwise directly enter step b10, if broadcast queue is empty, then generate new broadcast queue according to the dispatching algorithm that does not wait size to split data item, enter step b10 then;

Step b10, the time point that advances if surpassed the running time of regulation, then jumps to step b11; Otherwise, jump procedure b2;

Step b11 reads successfully the solicited message file, the statistics solicited message.

And, according to not waiting fractionation strategy of size data item to split, comprise following substep among the step b4,

Step b4.1, input data item d _i, data item d _iThe data item size

, data item priority Weight (d _{I, j}), bandwidth B w _k, cycle Cy _kBack initialization comprises that remaining data item size is set is data item size, subitem sequence number subitemorder=1;

Step b4.2 calculates the fractionation benchmark size SplitStandardSize of this data item,

$\mathrm{SplitS}\tan \mathrm{dardSize}=\frac{\mathrm{\α}\×{\mathrm{Bw}}_k\×{\mathrm{Cy}}_k}{S_{d_i}\×\mathrm{Weight}\left(d_{i,j}\right)}$

Wherein, α is parameter;

Step b4.3 if current residual data item size then continues step b4.4 greater than SplitStandardSize, otherwise jumps to step b4.5;

Step b4.4 adds a data subitem, and data subitem size is SplitStandardSize, and sequence number is subitemorder, subitemorder=subitemorder+1, and remaining data item size deducts SplitStandardSize, jumps to step b4.3; Step b4.5 adds a data subitem, and data subitem size is remaining data item size, and sequence number is subitemorder;

Step b4.6 returns and splits good data subitem.

And, generate new broadcast queue according to the dispatching algorithm that does not wait size to split data item among the step b9, adopt the scheduling strategy of priority maximum, comprise following substep,

Step 9.1.1, after formation was prepared in input, initialization broadcast queue was empty;

Step 9.1.2 carries out following circulation,

Step 9.1.2.1, traversal is prepared formation, chooses the data subitem with greatest priority LxRxW value;

Step 9.1.2.2 does not still surpass the data subitem size that one-period can be broadcasted if the total size of the data subitem in the broadcast queue adds this data subitem size, then adds this data subitem to broadcast queue, returns step 9.1.2.1 and circulates; Otherwise, withdraw from circulation, output broadcast queue.

Perhaps, generate new broadcast queue according to the dispatching algorithm that does not wait size to split data item among the step b9, adopt the scheduling strategy of taking all factors into consideration priority and bandwidth availability ratio, comprise following substep,

Step 9.2.1, after formation was prepared in input, initialization broadcast queue was empty;

Step 9.2.2 carries out following circulation,

Step 9.2.2.1, traversal is prepared formation, chooses to have minimum

The data subitem of value;

Step 9.2.2.2 does not still surpass the data subitem size that one-period can be broadcasted if the total size of the data subitem in the broadcast queue adds this data subitem size, then adds this data subitem to broadcast queue, returns step 9.2.2.1 and circulates; Otherwise, withdraw from circulation and enter step 9.2.3;

Step 9.2.3, initialization residue size;

Step 9.2.4 carries out following circulation,

Step 9.2.4.1, traversal is prepared formation, is chosen at data subitem size less than data subitem maximum under the prerequisite of residue size, if find then enter step 9.2.4.2, if can not find such data subitem, then withdrawing from circulation output broadcast queue;

Step 9.2.4.2 adds the data subitem that finds to broadcast queue, returns step 9.2.4.1 and circulates.

And the adjustment strategy according to the off period among step a4 or the step b4 is adjusted the off period of the data item of fractionation, comprises following substep,

(1) input data item d _iH data item gone on the air the off period that constantly r asks during the k time broadcast cycle

, data item d _iData subitem d _{I, 1}, d _{I, 2}..., d _{I, P}Initializing variable

, j=1;

(2) carry out following circulation,

(2.1) setting data subitem d _{I, j}Off period

$\mathrm{Deadline}=\mathrm{Deadline}+S_{d_{i,j}}/{\mathrm{Bw}}_k$

(2.2) if j equals data item d _iTotal number P of the data subitem that splits jumps out circulation output and adjusts the data subitem of off period, otherwise j=j+1 returns (2.1) and continues circulation.

And, add up solicited message among step a11 or the step b11, comprise following operation,

(1) reads successfully in the solicited message file all solicited messages to dynamic array vector;

(2) ID of the successful solicited message among the dynamic array vector according to request sorted;

(3) read the identical successful solicited message of ID of asking among the dynamic array vector successively, if the data subitem sum of the bar number of the successful solicited message of judgement and the data item of request equates, the request of then judging this ID is successfully request, adds up the access time of this success request; Otherwise this asks to be the request of losing efficacy;

(4) read the end of dynamic array vector, then withdraw from, export and successfully ask number, total access time.

The present invention's data broadcasting as required in real time can receive the request of mobile client fully,, the broadcasting of data in real time is gone out real-time response user's request according to real-time situation.Considered in real time differing of data item size in the data broadcasting as required, proposed to solve problem with the method for fractionation.Fractionation can make bigger data item and smaller data item can both take turns changeing and dispatch, and is unlikely to allow the data item size interfere with the effect of data item scheduling.Comprise two kinds of fractionation schemes:

(1) scheduling that size splits such as the present invention proposes, can be good at solving data item data broadcast scheduling as required in real time not of uniform size.It has taken the size in bandwidth and cycle into consideration, with respect to the different distributions of average off period and the distribution situation of data item mean size, combines the LxRxW dispatching algorithm, has proposed the ES-LxRxW scheduling.Experimental result proves, the ES-LxRxW scheduling is effectively, obeys under the situation of different distributions effect even can be better than some algorithm of not considering the data item size in the average off period.

(2) the present invention proposes the scheduling that does not wait size to split, can not only consider cycle and bandwidth in fractionation, can also consider the priority of data item size and scheduling, is to have considered that the attribute of data item self and the real-time status in the scheduling split.Equally, it combines the LxRxW dispatching algorithm with respect to the distribution of different data item mean size, has proposed the US-LxRxW scheduling.Having proposed in the literary composition to come the data item that does not wait the size data item to split is dispatched with two kinds of scheduling strategies, is respectively PF strategy and PxBF strategy.Experiment showed, the US-LxRxW scheduling no matter be to adopt PF strategy or PxBF strategy, effect all is better than the ES-LxRxW scheduling, and the request failure rate is lower than ES-LxRxW, and mean access time is lower than ES-LxRxW, and it is lower than ES-LxRxW on average to split number of times.And two kinds of strategies in the US-LxRxW scheduling, both are very approaching on request failure rate and average fractionation number of times, but the PxBF strategy is better than the PF strategy on mean access time.

Description of drawings

Fig. 1 is the system construction drawing of the broadcasting of cycle data as required of the embodiment of the invention.

Fig. 2 is the periodic broadcasting model schematic diagram of the embodiment of the invention.

Fig. 3 is the fractionation strategy schematic diagram that waits the size data item of the embodiment of the invention.

Embodiment

Describe technical solution of the present invention in detail below in conjunction with drawings and Examples.

1, method provided by the present invention increases on the basis of existing technology and has realized below describing split process:

Each process can adopt the software modularity technology to be embodied as corresponding module.The system configuration of the broadcasting of cycle data as required that the embodiment of the invention proposes as shown in Figure 1.Supposing the system comprises a server (Server) and a plurality of mobile client, and mobile client is submitted the request of oneself to by uploading channel (Up Channel), and server is by the formation of download channels (Down Channel) broadcasting data item.The system of server mainly comprises reception, obtain, split and dispatch four modules, receiver module is safeguarded the request queue (Req Queue) that the request of a system is formed, scheduler module safeguards that is prepared a formation (Ready Queue), safeguard a broadcast queue (Broadcast Queue) simultaneously, preparing formation is made up of the data subitem, each data subitem comprises a request queue, the request queue of a data subitem is the formation that all requests of this data subitem constitute, can be designed to when therefore specifically implementing: prepare the concrete data structure of formation and be (data item ID, the data item size, split size, data subitem sum, data subitem sequence number, the request queue of this data subitem).The data subitem of preparing formation adds broadcast queue after dispatching, but the concrete data structure of respective design broadcast queue is (data item ID, data item size, split size, data subitem sum, data subitem ID, go on the air the request queue of this data subitem the time).ID represents sequence number.Because data item is split as the data subitem, request also can be divided in each data subitem, i.e. request can be asked all data subitems of the data item that it asks.The request queue of a data subitem is the formation that all requests of this data subitem constitute.In all request queues, embodiment designs the concrete data structure of each information requested and is (ID of request, the data item ID that asks, the time of request, the off period of request).

The operation principle of method provided by the present invention comprises following process:

(1) receives (Accept).Mobile client is submitted the request of oneself to by uploading channel, and each request comprises the data item information of request, the information such as off period of request.Server receives request, checks this request earlier whether in broadcast queue or prepare in the formation, if, then directly will ask to add in the broadcast queue or in the preparation formation in the request queue of corresponding data subitem; If do not exist, then will ask to press request time order from small to large and add in the request queue of system, and be request distribution ID.This process can mainly be finished by receiver module.

(2) obtain (Fetch).Server takes out the request of current time from the request queue of system, (Database DB) or search the data item that satisfies request in the buffer memory of server, obtains groups of data items and is made into data item information for (Internet), database from the internet.The present invention does not consider that server obtains the time of data and the cache policy of server.Data item information comprises the ID of data item and the content of data item.This process can mainly be finished by acquisition module.

(3) split (Split).Server splits the data item that gets access to (waiting big or small big or small fractionation that split or do not wait) according to splitting strategy, and the data subitem that splits into is added in the preparation formation.Such data item has one or several data subitem, and the data item ID of these data subitems is identical, but data subitem ID difference.Prepare formation and at first arrange by data item ID, the data subitem of same data item ID is arranged by data subitem ID, and a current request of data item is kept in the request queue that splits each data subitem of gained.This process can mainly be finished by splitting module.

(4) scheduling (Schedule).Server is periodically dispatched tissue by dispatching algorithm to the data of preparing in the formation, the dynamic data subitem of selecting adds in the broadcast queue, server sends to satellite or base station with broadcast queue, and broadcast away the data subitem in the broadcast queue by download channels satellite or base station.This process can mainly be finished by scheduler module.

Split and dispatch two steps shown in Fig. 1, divide fractionation (Equal Split) and scheduling and the fractionation (UnEqual Split) that does not wait the size data item and scheduling in order to wait size data item.

Mobile client (MC) conduct and server separate equipment according to the demand of oneself, constantly produce request and send to server by uploading channel, constantly monitor download channels afterwards and obtain own data item of asking.

During scheduling, server splits into the data subitem with data item; And mobile client is assembled the data subitem after having received the data subitem, is reduced into original data item, and the order of assembling can be determined according to the numbering of data subitem.

2, the data broadcast model to the embodiment of the invention describes:

(2.1) model agreement

For the ease of research, the present invention does following hypothesis and agreement for real-time data broadcast model as required:

(1) periodic broadcasting scheduling mode is as required adopted in tentation data broadcasting, namely server according to the RQ cycle of mobile client choose the data item that will broadcast according to dispatching algorithm, and broadcast away.

(2) data processing time agreement: suppose that server all is 0 handling the time of asking, obtaining the data item, calculating priority level and the generation broadcast queue that ask, the broadcasting that is data item can not have because of the processing time at interval in time, all is to broadcast continuously.

(3) satisfiability of mobile client: the mobile client data item of asking of having supposed server broadcast, all ask the mobile client of this data item all can be satisfied, and ignore the replacement problem of data item during the broadcasting.

(4) existence of data item: all data item of supposing the mobile client application all exist on server, and namely the data item of all mobile client applications is all in the data in server storehouse, as long as dispatching algorithm reaches.

(5) Guang Bo data item identifiability: namely mobile client can be known the data item whether it will visit itself according to identification code before receiving any one data item.

(6) data item of broadcasting overall situation uniqueness: namely Guang Bo each data item has a keyword, it can identify a broadcasting item uniquely, and mobile client visits data item in the data broadcasting according to this keyword, and this identifiability with the data item of broadcasting is consistent.

(7) Guang Bo data item position in broadcast queue is determined: when define the position of a data item in data broadcasting and is this data item being broadcasted in broadcast channel from the side-play amount of certain zero-time.All data item all are that order is broadcasted, and do not have multichannel or concurrent broadcasting situation.

(8) consistency of data hypothesis: all data item can only wait until that all data subitems of this data item could not upgrade and revise when not existing in preparing formation, the data subitem that each request of each mobile device receives is that the data item of same version is split.

(9) forms data request agreement: each mobile client is once only being asked a data item in the request, the repeatedly onrelevant between the requested data item of mobile client, and also the user can send next request before not obtaining a request msg.

(10) mobile client is to broadcasting the independence of data access: the data item of all broadcasting is visible to all mobile clients on the server, and all mobile clients all are separate, non-interfering to the visit of data item.

Because the present invention is split as the data subitem with data item, below broadcasting item and preparation item refer to the data subitem in broadcast queue and the preparation ordered series of numbers respectively.

(2.2) off period model

In concrete enforcement, when assessing the dispatching algorithm performance, the present invention during the off period, can select specific implementation as required by those skilled in the art.Three kinds of different off periods for example commonly used distribute, and namely are evenly distributed, exponential distribution and stationary distribution.

DeadLine=ArrivalTime+Uniform (0,2*MeanSlack) (formula 1)

DeadLine=ArrivalTime+Exponential (0,2*MeanSlack) (formula 2)

DeadLine=ArrivalTime+Fixed (MeanSlack) (formula 3)

Wherein Arrivaltime is the time of advent of request, and MeanSlack is average off period skew primary quantity, does not comprise request processing time.Formula 1 plants the off period for evenly dividing, and Uniform is even distribution random function, generate at random one (0,2*MeanSlack] between value.Formula 2 is the exponential distribution off period, and wherein Exponential is the exponential distribution random function, generate at random one (0,2*MeanSlack] between value.Formula 3 is the stationary distribution off period, and wherein, Fixed is the stationary distribution random function, and steady state value is MeanSlack.

(2.3) the big mini Mod of data item

In concrete enforcement, when the present invention assesses the dispatching algorithm performance, can select concrete data item size to calculate implementation as required by those skilled in the art.Chang Yong three kinds of different data item size distribution, i.e. evenly distribution, exponential distribution and stationary distribution for example.

DataSize=Uniform (0,2*MeanSize) (formula 4)

DataSize=Exponential (0,2*MeanSize) (formula 5)

DataSize=Fixed (MeanSize) (formula 6)

Wherein, MeanSize is the skew primary quantity of data item mean size.Formula 4 is for evenly dividing the data item size that plants, and Uniform is equally distributed random function, generate at random one (0,2*MeanSize] between value.Formula 5 is the data item size under the exponential distribution, Exponential exponential distribution random function, generate at random one (0,2*MeanSize] between value.Formula 6 is the data item size of stationary distribution, and wherein Fixed is the random function of stationary distribution, and steady state value is MeanSize, and namely the data item size is fixing special circumstances.

(2.4) broadcast model definition

Figure 2 shows that data item splits the broadcast scheduling cycle afterwards, establish and carried out the k-1 time scheduling, the k time scheduling, the k+1 time scheduling ..., the broadcast cycle of the k time broadcast scheduling wherein, 1,2,3,4 ..., h ..., the respectively corresponding data item of M.For studying conveniently, the present invention does following hypothesis and definition, and is as shown in table 1.

The hypothesis of table 1 model and definition

3, appraisement system

In real time as required in the Radio Data System, the quality of the stand-by period measure algorithm of should be not only not considering with traditional algorithm, and should under the situation of meeting consumers' demand in real time of algorithm, quality of services for users be taken into account.And the real-time demand that satisfies the user namely is not the data item that simple broadcasting user is asked, and must satisfy the off period of user's request.Simultaneously, the efficient of algorithm also is the response time that is directly connected to broadcasting, so the efficient of algorithm also should consider.

The present invention is from asking failure rate, mean access time and on average split three aspects of number of times to come dispatching algorithm is estimated.The request failure rate is to estimate the situation that is satisfied of user's request, and the access time is used for reflecting quality of services for users, on average splits number of times and be the efficient of estimating fractionation.

(1) request failure rate (Loss Rate, LR).

The request failure rate refers to the request of losing efficacy shared ratio in all requests, can reflect the validity of scheduling, i.e. the situation about being satisfied of all requests.Wherein, total request number refers to the number sum of the request of the number of request of total inefficacy and total success.

If in the processing procedure of server, the off period of certain request to but the data item of its request be not broadcasted, the broadcasting of this data item was lost efficacy to this request so.On the contrary, if data item can be broadcasted before the off period of request, this request is effective so.After data item splits into the data subitem, for each mobile client, have only him all data subitem broadcasting of requested data item to finish, this request could be satisfied, and namely this asks to be successfully request; On the other hand, if the off period of any one data subitem wherein of his request msg has been missed, the broadcasting of this data item had just been lost efficacy so, namely this to ask be the request of losing efficacy.

(2) mean access time (Average Access Time, AAT).

Mean access time refers to the mean value of the access time that all are successfully asked, and namely the summation of all access times of successfully asking is divided by total successful request number.Access time refers to that issuing a request to mobile client from mobile client begins to receive the time that data continue between this.

(3) on average split number of times (Average Split Times, AST).

The present invention considers is to the fractionation that waits size of data item and does not wait the fractionation of size, will take server CPU during fractionation, and therefore the number of times of fractionation is more many, and server spends the time in fractionation can be more many, and the efficient of fractionation can be more low.Therefore the present invention comes the speed of service of measure algorithm, the i.e. efficient of algorithm with the fractionation number of times.The present invention is defined as the number of the data subitem that each data item splits into the fractionation number of times of data item, and so on average splitting number of times is that total fractionation number of times of data item is divided by the result of total number of the data item that splits.

4, the size data item such as provide to split among the embodiment and not wait the size data item to split dual mode selective, below the specific implementation that splits of reciprocity size data item describe.

(4.1) fractionation and the scheduling principle that split of size data item such as

In actual applications, the data item size of user's request is not of uniform size, and bigger data item is arranged, and smaller data item is also arranged.If the algorithm before just simply applying mechanically, and data item is not split and handles, following inconvenience may be with.(1) bandwidth is not enough.Simultaneously because technical equipment or wireless transmission medium, bandwidth size under different environment is different, so may cause in practice that bandwidth is not enough under a lot of situations.Under the not enough situation of bandwidth, if the cycle size is fixed, then may cause big data item in one-period, can not broadcast Wan.(2) efficient is not high.In data broadcasting before, each data item only could be other mobile subscriber's services after broadcasting is finished.Under the certain situation of bandwidth, if broadcast big data item, so the time of this data item busy channel long, make the number of users of service in the unit interval reduce, service quality reduces.Equally, if data item size is certain, under the less situation of bandwidth ratio, the time of data item busy channel is also with long.(3) not to be fit to very much for periodic broadcasting.Periodic broadcasting is that server is periodically chosen data item and broadcasted, compare opposite extreme situations, under the situation about all determining in bandwidth and cycle, if certain data item size has surpassed the data item size of the maximum that can broadcast in the broadcast cycle, this data item then can not be broadcasted in this cycle so, and this has just runed counter to the theory of periodic broadcasting.

The present invention proposes the data subitem with sizes such as data item split into.After size such as data item split, with respect to not waiting big or small fractionation, be based on following some consider.(1) mobile client can conveniently be predicted the time that next data item arrives.In practice, under amount of bandwidth was certain situation, if data item is to wait size, then mobile client can be calculated the arrival time of next data item when intercepting download channels.When data item arrives, judge the data of whether oneself asking earlier, if then can continue to intercept, and obtain the data that oneself need; If not, then can switch to park mode, when next data item arrives, switch to enable mode again and intercept channel, and continue to intercept.Therefore, etc. the data item size time, the present invention just can well predict the time that next data item arrives according to existing condition, thereby saves the energy of mobile client.(2) can be good at using achievement before.The real-time algorithm of data broadcast scheduling as required before all is to study under the situation of sizes such as data item, and some good algorithms occurred, as EDF of the prior art, MRF, RxW, SIN-α, LxRxW.If data item such as is split at size, the data broadcast scheduling algorithm before then can well using for reference comes data item is dispatched.

(4.2) fractionation and the scheduling flow that split of size data item such as

Deng fractionation and the scheduling that the size data item splits, be on the basis of real-time digital broadcasting dispatch system as required, earlier data item is waited the fractionation of size, dispatch again and broadcast.When splitting Deng size, pre-set SplitK value, and then according to etc. the fractionation strategy of size data item fractionation split.

The fractionation that splits Deng the size data item and scheduling (Equal Split and LxRxW, ES-LxRxW), concrete flow process such as process ES-LxRxW.

Process: ES-LxRxW

(1) initialization generally comprises request queue, prepares initial work such as formation, broadcast queue, time point, when specifically implementing, can design concrete operations by those skilled in the art;

(2) if not request of current time then jumps to (7), otherwise continues to enter (3);

(3) take out this request, i.e. the request of current time;

(4) obtain the data item of this request, according to the fractionation strategy that waits the size data item it is split, according to the adjustment strategy of off period the off period of the data item that splits is adjusted;

(5) for all the data subitems after splitting, handle respectively: if in the broadcasting item of broadcast queue and no broadcasting, then this request is added in the request queue of this broadcasting item; Otherwise, if the preparation item in the preparation formation then adds this request in the request queue of this preparation item; Otherwise, generate new preparation item adding and prepare formation (be about to this data subitem adding and prepare formation), and will ask adding to prepare should newly prepare in the request queue of item in the formation;

(6) new request to be generated such as; Jump to (2);

(7) broadcast according to broadcast queue;

(8) judge whether just to have broadcasted not empty broadcasting item, if otherwise directly enter (9), if just broadcasted not empty broadcasting item, the successful solicited message that then will broadcast item writes successfully solicited message file (can set up successful solicited message file in advance), delete this broadcasting all requests, delete this broadcasting item, enter (9) then;

(9) judge whether that broadcast queue is empty, if otherwise directly enter (10), if broadcast queue is empty, then generate new broadcast queue according to the dispatching algorithm that waits size to split data item, enter then (10);

(10) time point that advances if surpassed the running time of regulation, then jumps to (11); Otherwise, redirect (2);

(11) read successfully the solicited message file, statistics solicited message (tactful SSRM).

Therefore the fractionation of data item is arranged among the present invention, dispatch at the data subitem, when adding request, each data subitem can add request to, therefore loses efficacy when statistics and during successful request, can not simply calculate by method before.

For the request of losing efficacy, missed when the off period of certain data subitem, then the request of this ID was lost efficacy.The present invention is for the request of losing efficacy, on one side dispatch during scheduling, Yi Bian add up, and when having run into the request of losing efficacy, the off period of i.e. this request has been missed, and can delete this solicited message, seeks and delete the request of same request ID in the preparation item of preparing in the formation other simultaneously.In the reality, when the off period has been missed in certain request of certain data subitem, be that off period of this request of this data subitem is less than the current time, can infer so, the data subitem that the sub-sequence number of the data item that this request is asked is littler than the sub-sequence number of current data subitem has been broadcasted success, otherwise the off period of this request early will be missed.Therefore, here as long as delete sub-sequence number than the request of request ID together in the big data subitem of the sub-sequence number of current data item.

When adding up successful solicited message, judge whether the data subitem of each request has successfully broadcasted in its off period, if successfully broadcasted all, then this to ask be successfully request; If any one data subitem has missed the off period, then should request still be the request of losing efficacy.

Add up strategy (the Statistic Success Request Message strategy of successful solicited message, SSRM), this strategy is only added up the number of successfully request and the summation of access time, because the request of losing efficacy is before just by statistics, so the request that will lose efficacy in this strategy all neglects.

Vector represents container or dynamic array, is the Essential Terms in the existing C++ STL, and the embodiment of the invention adopts dynamic array.Sum for the data subitem of the data item of the bar number of solicited message and request is situation about equating, the i.e. request of this ID is the situation of successfully asking, when adding up its access time, the time that the present invention takes first data subitem that the family receives is its time of reception, rather than the subitem sequence number is the time of reception of 1 data subitem.Because, explanation by preamble can be known, the data subitem is not necessarily broadcasted from small to large according to its subitem sequence number, and to be the user calculate by this time point of beginning receiving data item the access time that is related to time of reception, therefore, the present invention is the timing definition of first data subitem that mobile client receives the time of mobile client receiving data item, and the access time of mobile client is request time to the time that is continued between the time of reception.

(4.3) the fractionation strategy of size data item such as

Cutting and packaging Problems (Cutting and Packing) are the problems of classics, and it is the combinatorial problem of difficulty, under the background of many reality application is arranged, generally industry and service industry's appearance often.In general, it is that " big object " split into " small articles ", comes to make waste to minimize in this way.This type of problem more typically has: the cutting paper logs of paper industry become the big plank of the cutting of narrow volume, furniture industry be divided into littler rectangular slab, the fashion industry from irregular assembly of fabric warrping cutting profile etc.

The present invention uses the thought of cutting and packaging Problems, adopts the mode that splits will be assembled into one-period and broadcasts away, and mobile client is assembled into original data item again with the data that split after receiving data.

For the fractionation of data item, the present invention is its fractionation that fractionation of size data item such as is divided into and does not wait the size data item, the fractionation of size data item such as this chapter introductions earlier.

Deng the fractionation of size data item, be the data subitem that all data item is split into equal size, dispatch broadcasting again.Namely as long as the size of a fractionation of definition, then all data item all will be split into the equirotal data subitem of fractionation of this definition.

The present invention defines the size of fractionation for splitting benchmark size (SplitStandardSize), waits the fractionation of size data item to be about to the data subitem that all data item all split into fractionation benchmark size.Because the data item size may be less than splitting the benchmark size, so the filling size that the present invention defines the data subitem is the data item size of data subitem actual load.The concrete fractionation strategy that waits the size data item (Equal Split, ES) as follows.

As shown in Figure 3, be 3M if definition splits benchmark, the data item size is the data item d of 8M so _iWill be split into 3 data subitems: data subitem d _{I, 1}, data subitem d _{I, 2}, data subitem d _{I, 3}, the size of data subitem all is 3M.Wherein, data subitem d _{I, 1}All filled data subitem d by data item _{I, 2}All filled by data item, and data subitem d _{I, 3}Only filled the data item of 2M, also having 1M is in order to keep the size that waits of data item, and reserves the zone of not filling up of coming out, and data item size is 3M, and the filling size is 2M.For reserving the zone of coming out, it can not play effect for request, but it can take the bandwidth of broadcasting.

In order to study conveniently, the present invention does following hypothesis and explanation for the fractionation strategy that waits the size data item:

(1) the data subitem after data item splits need not broadcasted according to sequencing, i.e. the data subitem broadcast order of data item di d not necessarily _{I, 1}, d _{I, 2}, d _{I, 3}

(2) at each request, have only this request in the off period, to receive the request that all data subitems are only success.If any one data subitem has wherein surpassed the off period, then should request lose efficacy.

(3) the inefficacy request number of data item and successfully to ask number all be at original data item, the inefficacy request number of data subitem and successfully ask not double counting of number.

(4) deadline of data subitem broadcasting is that data item by its actual filling decides, i.e. data subitem d _{I, 3}The broadcasting deadline refer to broadcasting deadline of the data item of the 2M that fills.

(4.4) the adjustment strategy of off period

After data item split, to the definition of off period, then the off period of data item had just satisfied as long as data item has been gone on the air according to before.Data item d as mentioned above _iSplit into data subitem d _{I, 1}, d _{I, 2}, d _{I, 3}, according to definition before, as long as data item d _iGo on the air, i.e. data subitem d _{I, 1}Go on the air data item

Off period just satisfy and asked.But, if server is being broadcasted data subitem d _{I, 1}Do not broadcast data subitem d afterwards, and always _{I, 2}, d _{I, 3}, this request just slowly is not being met so.Obviously, if the time of waiting for is of a specified duration excessively, it is the off period that can surpass request.Therefore, after data item split into the data subitem, the off period will change through corresponding.Here, the present invention will be divided in the corresponding data subitem off period.

In the fixed-size situation of data item, under the given situation of bandwidth, the off period of request (deadline of the data item of certain request of going on the air) determined, request off period (finishing the deadline of these data of broadcasting) of finishing is determined so.Same, under the given situation of bandwidth, when waiting size to split, the airtime of data subitem determines, data item is broadcasted the time of finishing and determined so.If determined the off period of request, then determined to ask to finish the off period.According to request off period deadline, can push away to such an extent that this request of broadcasting is finished between the off period finishline pass of satisfying and is

The off period is finished in request according to the current data subitem, just can push away request off period of next data subitem.

The present invention defines the new off period with second off period to last data subitem according to the time of transmitting, i.e. second data subitem d _{I, 2}Off period be first data subitem d _{I, 1}Off period add time of first data subitem broadcasting, the 3rd data subitem d _{I, 3}Off period be second data subitem d _{I, 2}Off period add time of second data subitem broadcasting, the rest may be inferred.Concrete off period adjustment strategy (DeadlineAdjust, DA) as follows:

According to the setting rule of table 1,

Expression data item d _iH data item gone on the air the off period that constantly r asks during the k time broadcast cycle.Illustrate: if data item d _iAny one data subitem surpassed the off period and be not broadcasted, data item d then _iLost efficacy, and all data item d that broadcasts and do not broadcast _iThe data subitem all lost efficacy.

(4.5) scheduling strategy that splits of size data item such as

In on-demand broadcasting scheduling in real time, in not considering the broadcast scheduling algorithm of data item size, all be that the data item size of hypothesis is fixing situation, identical with the model of size after splitting that wait of the present invention.Do not consider that the data item size has following several typical dispatching algorithm.Do not split the data subitem in these algorithms, for the purpose of convenient relatively optimization, data item is designated as d _{I, j}

(1) EDF will be used for representing the urgency of data item the off period, get off period data item the earliest and broadcast at first.The tentation data item is d _{I, j}, current k broadcast cycle dispatched h data item, d _{I, j}The priority calculating formula is

It is more high to be worth more little data item priority.The shortcoming of algorithm is only to have considered the off period situation, may cause the data item that the few data item of request is Duoed than request preferentially to broadcast, and causes failure rate low inadequately.

(2) MRF will ask number as urgency and the importance of data item, get the data item that has largest request number and broadcast at first.The tentation data item is d _{I, j}, current k broadcast cycle dispatched h data item, d _{I, j}The priority computing formula is

It is more high to be worth more little data item priority.The shortcoming of algorithm is only to have considered the request situation of data item, if be applied in the real-time on-demand broadcasting, to the not clear sense of urgency of data item request, so failure rate may be very big.

(3) RxW will ask number and stand-by period as the urgency of data item, get have the request number and maximum wait time the product maximum data item broadcast at first.The tentation data item is d _{I, j}, current k broadcast cycle dispatched h data item, d _{I, j}The priority calculating formula is

It is more high to be worth more little data item priority.Though the shortcoming of algorithm is to reduce mean access time and average tuning period, can not be at real-time on-demand broadcasting problem, the same with the MRF algorithm, failure rate may be very big.

(4) SIN-α as the urgency of data item, gets SIN value (the residue slack time of request to be solved) data item that has minimum SIN value and broadcasts at first.The tentation data item is d _{I, j}, current k broadcast cycle dispatched h data item, d _{I, j}The priority calculating formula is

It is more high to be worth more little data item priority, and α is the weights factor of weighing off period and request number in the algorithm, and experiment shows that effect and the optimal situation of α=1 o'clock calculating formula is close.The shortcoming of algorithm is the whole urgency that reflects data item with the off period, and is perfect inadequately, and do not consider the stand-by period situation.

(5) LxRxW will be about to failure number, request number and stand-by period as the urgency of data item, get the request number that is about to lose efficacy, ask the data item of number and maximum wait time three's achievement maximum to be broadcasted at first.The tentation data item is d _{I, j}, current k broadcast cycle dispatched h data item, d _{I, j}The priority calculating formula is , it is more high to be worth more little data item priority.Algorithm reflects the urgency of off period with being about to failure number, will ask number and maximum wait time well to combine again, has reflected the urgency of data item integral body.

The LxRxW dispatching algorithm combines the request of be about to losing efficacy number, current request number and the maximum wait time of data item as the main policies factor of scheduling, the whole urgency that the request number that wherein is about to lose efficacy more can the representative data item than off period of data item.

For the priority of quantitatively evaluating data item, in each broadcast cycle, determine before the data item of each medium wave that data item all can be calculated a LxRxW value.The k time broadcast cycle, during h the medium wave, data item d _{I, j}The LxRxW value computing formula as shown in Equation 7.

$\mathrm{Weight}\left(d_{i,j}\right)=\frac{1}{L_{d_{i,j},k,h}\×R_{d_{i,j},k,h}\×W_{d_{i,j},k,h,r}}$ (formula 7)

Wherein, the data item priority with minimum Weight value is the highest, is broadcasted at first.But in the Weight value is calculated

May be 0, just use this moment

Expression data item d _{I, j}Urgency, therefore be revised as formula 8

$\mathrm{Weight}\left(d_{i,j}\right)=\{\begin{array}{c}\frac{1}{L_{d_{i,j},k,h}\×R_{d_{i,j},k,h}\×W_{d_{i,j},k,h,r}}\mathrm{if}(L_{d_{i,j},k,h}\≠0)\\ \frac{{\mathrm{StartD}}_{d_{i,j},k,h,r}}{R_{d_{i,j},k,h}}\mathrm{if}(L_{d_{i,j},k,h}=0)\end{array}$ (formula 8)

For the sake of ease of implementation, the present invention will wait the fractionation of size data item and LxRxW dispatching algorithm well to combine, and split according to ES earlier and will carry out broadcast scheduling by LxRxW again, concrete dispatching algorithm such as process Schedule, but d wherein _{I, j}Expression splits gained data subitem.

(4.6) the value strategy of fractionation benchmark size

Deng the fractionation of size data item, the number of fractionation decides by splitting benchmark size SplitStandardSize fully.And split the benchmark size, should consider in conjunction with the actual conditions of amount of bandwidth and cycle size.

(1) amount of bandwidth (Bandwidth).Bandwidth refers to transmission data item size in the server unit interval.Bandwidth is more big, and the data item of broadcasting is more many in the unit interval, and the throughput of server is more big, and the time of onesize data item broadcasting, still less other conditions one regularly load of system diminished.

(2) cycle size (Cycle).Cycle refers to the time size that server is once broadcasted.Cycle is more big, and the data item of broadcasting under the certain situation of bandwidth is more.

The size that data item splits should be that the data item that can broadcast with each cycle is big or small relevant, namely with the product (Bw in bandwidth and cycle _k* Cy _k) size relevant.And in one-period, the data item number of broadcasting should be that integer is individual, and the fractionation benchmark size of data item should be to be divided exactly by the product in bandwidth and cycle so.Because the data item size of broadcasting in the one-period, if not an integral multiple data item size, can there be waste in bandwidth so, also can waste the speech amplifier meeting of data item accordingly, thereby makes failure rate rise.Therefore, the size of data item fractionation should be Bw _k* Cy _k/ SplitK, wherein SplitK is a positive integer, the data item size of the interior broadcasting of expression one-period is the number of the data subitem of fractionation benchmark size SplitStandardSize.

$\mathrm{SplitS}\tan \mathrm{dardSize}=\frac{{\mathrm{Bw}}_k\×{\mathrm{Cy}}_k}{\mathrm{SplitK}}(\mathrm{SplitK}\∈N^{+})$ (formula 9)

Wherein, N ⁺The expression positive integer, the problem that splits for size such as data item, data item splits into much portions and is only preferably actually, has only by experiment and could illustrate.And for the fractionation scheme, need an evaluation criterion, estimate the quality that this splits scheme.

Because the present invention considers from service quality, satisfy user, the best demand that satisfies the user namely at most, so the present invention comes dispatching algorithm is estimated from request failure rate and two aspects of mean access time.In addition, because the fractionation of data item will occupy the regular hour, therefore the efficient that splits also is a key point that should be noted that.The present invention adopts average fractionation number of times to weigh the efficient of fractionation, and it is more little on average to split number of times, and the number of times of fractionation is more few, and the efficient of dispatching algorithm is more high.

Concrete enforcement can obtain the selected value roughly of SplitK from experimental result, table 2 be distribute the off period and the situation of the various combination of data item size distribution under the summary sheet of suggestion value of SplitK value.

The best SplitK of table 2 gathers

5, in the specific implementation that does not wait the fractionation of size data item, except not waiting big or small fractionation strategy and the scheduling strategy of data item, other are the same with fractionation such as item such as size data such as grade, for example the adjustment strategy of off period.Below to do not wait the size the fractionation strategy and the scheduling strategy of data item be specifically described:

(5.1) fractionation and the scheduling principle that does not wait the size data item to split

Though split airtime and the good achievement before of using that can be good at predicting next data item Deng big or small data item, the scheduling that waits the size data item to split has following some deficiency.

(1) do not take into full account bandwidth utilization.Fractionation Deng the size data item, owing to there are a lot of data subitems not to be filled, this makes most of data item split the waste that all can have living space afterwards, thereby make the bandwidth of server not be utilized fully, thereby a part of user's stand-by period can become big relatively, and the service quality of server can corresponding step-down.

(2) server efficient is low.When the data item size was bigger, the number of times of fractionation was many, be the CPU that need take server and split data item, so server can cause efficient very low because of the fractionation of data item.

(3) do not consider that the data item size is to the influence of algorithm.If the data item size is too little, the phenomenon of a lot of space wastes may appear, may cause that bandwidth utilization above-mentioned is very low.On the contrary, if data item is too big, then may make the efficient step-down of server.Therefore, should choose a better method, make the two compromise.

(4) do not consider data item urgent influence to splitting in time.Temporal urgency is directly relevant with service quality, when therefore splitting, should also take into account it.

This chapter has proposed data item is split into the data subitem that does not wait size.Fractionation with respect to waiting size data item does not wait the fractionation of size data item to have following advantage.

(1) can be good at interior bandwidth utilization of lifting cycle.To the fractionation that does not wait size of data item, the bandwidth that can take into account each cycle can take full advantage of.

(2) can be good at considering that the data item size is to the influence of algorithm.The data item that the data item size is bigger, if broadcasting always, the bandwidth resources that take are many, and namely in bandwidth one regularly, the time of broadcasting is long, and the number of users of unit interval service reduces.Therefore, the data item that the data item size is bigger should repeatedly split, repeatedly broadcasting.

(3) can adjust the real-time condition of data item dynamically.During fractionation, in conjunction with the dispatching priority of data item, can consider that real-time being about to of data item lost efficacy request number, request number and maximum wait time, make fractionation can directly follow service quality correlation.

(5.2) fractionation and the scheduling flow that does not wait the size data item to split

Do not wait the fractionation of size data item, as its name suggests, be about to different data item and split according to different fractionation benchmark, also namely wait the size data item split in defined fractionation benchmark size can be according to different data item and different.

For the equirotal data item of data item and the data item that waits size to split, data item itself is of a size, and just in time be complete several data item of broadcasting in the cycle, therefore the situation before all is worth the priority of having considered data item, and the factor of data item size is not taken into account when scheduling.And for the data item that does not wait size to split, the data subitem size that each data item splits into does not wait, and therefore, when scheduling, not only will consider the priority of data item, also must consider the data item size.

The present invention will not wait the scheduling of size data item to be divided into two parts, and namely elder generation carries out not waiting the fractionation of size to data item after getting access to data item, and the data subitem that splits into is added in the preparation formation.Afterwards, will split good data item at scheduling phase and dispatch broadcasting.

The scheduling that does not wait the size data item to split is on the basis of real-time digital broadcasting dispatch system as required, earlier data item is carried out not waiting the fractionation of size, dispatches and broadcasts.Do not wait the fractionation of size data item, similar with the data item fractionation that waits size, also be that one of definition earlier splits the benchmark size.Different is, the benchmark that splits Deng the size data item is to calculate according to the SplitK value at the very start, but do not wait the fractionation benchmark size meeting of size data item different because of the difference of data item, the fractionation benchmark size of each data item is calculated before fractionation in real time.

(Unequal Split and LxRxW, US-LxRxW), concrete flow process is similar with the fractionation and the scheduling flow that wait the size data item to split, as process US-LxRxW not wait fractionation that the size data item splits and scheduling.

Process: US-LxRxW

(1) initialization generally comprises request queue, prepares initial work such as formation, broadcast queue, time point, when specifically implementing, can design the idiographic flow operation by those skilled in the art;

(2) if not request of current time then jumps to (7), otherwise continues to enter (3);

(3) take out this request;

(4) obtain the data item of request, according to the fractionation strategy that does not wait the size data item it is split, according to the adjustment strategy of off period the off period of the data item that splits is adjusted, adjust strategy as DA;

(5) for all the data subitems after splitting, if in broadcast queue and no broadcasting, then this request is added in the request queue of this broadcasting item; Otherwise, if in preparing formation, then this request is added in the request queue of this preparation item; Otherwise the preparation formation is added in new of generation to, and will ask to add in the request queue of being somebody's turn to do new in the preparation formation;

(6) generate new request and the formation that joins request; Jump to (2);

(7) broadcasting;

(8) judge whether just to have broadcasted not empty broadcasting item, if otherwise directly enter (9), if just broadcasted not empty broadcasting item, the successful solicited message that then will broadcast item writes file, delete all requests of this broadcasting item, delete this broadcasting item, enter (9) then;

(9) judge whether that broadcast queue is empty, if otherwise directly enter (10), if broadcast queue is empty, then generate new broadcast queue according to the dispatching algorithm that does not wait size to split data item, enter (10) then;

(10) time point that advances if surpassed the running time of regulation, then jumps to (11); Otherwise, jump to (2);

(11) read successfully the solicited message file, the statistics solicited message.Statistics solicited message specific implementation is seen tactful SSRM.

(5.3) the fractionation strategy that does not wait the size data item to split

The data item size is carried out not waiting the fractionation of size, and the most direct evaluation is exactly the service quality of each broadcasting.And request failure rate and mean access time are the standards of estimating service quality.

(1) cycle size and amount of bandwidth.The data item size that the product representation one-period of cycle and bandwidth can be broadcasted.Seizing the opportunity of cycle and bandwidth is more big, the data item size that one-period can be broadcasted is more big, and the throughput of system is more big, and the size that data item splits can be more big, thereby in the request failure rate that guarantees data item, make that the fractionation number of times of data item is few as much as possible.

(2) data item size.Here the data item size that does not split that refers to, it is the attribute of data item self, splitting scheduling should consider.Different data item varies in size, and the size of fractionation also should be different.The data item size is more big, and the fractionation number of times that data item is possible is more many, and the failure rate of data item can increase, and the efficient of server can reduce.

(3) time urgency degree.The urgent degree of time is directly connected to total failure rate and mean access time, and it should be directly reflected into and split in the benchmark size.The dispatching priority that the present invention considers adopts the LxRxW algorithm, and it has taken all factors into consideration soon lost efficacy request number, request number and maximum wait time three.The dispatching priority size is more big, and namely the value of LxRxW is more big, and the expression off period is more approaching, and data item should preferentially be broadcasted, so this data item can not be torn open too for a short time.Because if that tears open is too little, the data subitem that then splits into can be a lot, thereby make failure rate increase, and the efficient of server reduces.

Identical with the fractionation that waits the size data item, the present invention defines and splits the benchmark size is SplitStandardSize.During fractionation, calculate real-time SplitStandardSize earlier, again data item is split into the data subitem.

(1) by to splitting the analysis of the relevant factor of size, finds data item d _iThe data item size

The data item size that can broadcast with one-period is more big, namely

Be worth more greatly, the broadcasting of data item is more big to the pressure of bandwidth, and the unit interval number of users of service is more few, and namely the size of Chai Fening should be more little, could make the number of users of serving obtain equilibrium.Namely

Size be inversely proportional to splitting size.

(2) and the dispatching priority of data item

Value more big, the expression data item off period more approaching, the request number more many, the stand-by period is more long, data item more should preferentially be broadcasted, so the fractionation size of data item should be more big.Namely

Size be directly proportional with splitting size.

Therefore, be formula 10 according to top 2 formula that can obtain the fractionation size of data item.

$\mathrm{SplitS}\tan \mathrm{dardSize}=\mathrm{\α}\×\frac{{\mathrm{Bw}}_k\×{\mathrm{Cy}}_k\×L_{d_{i,j},k,h}\×R_{d_{i,j},k,h}\×W_{d_{i,j},k,h,r}}{S_{d_i}}$ (formula 10)

Wherein, the α in the formula is parameter to be set, and when specifically implementing, value can be selected default by The simulation experiment result.Amendment type by LxRxW in the formula 8 can be modified to following formula formula 11.

$\mathrm{SplitS}\tan \mathrm{dardSize}=\frac{\mathrm{\α}\×{\mathrm{Bw}}_k\×{\mathrm{Cy}}_k}{S_{d_i}\×\mathrm{Weight}\left(d_{i,j}\right)}$ (formula 11)

What split the benchmark definition is the size that split of each data item.In the reality, how each piece should split, and specifically splits muchly so, and the present invention has adopted following strategy.

Do not wait the size the fractionation strategy (Unequalsize Split, US).Computing formula according to the fractionation benchmark size that does not wait size, each data item is according to its actual conditions, calculate and split after the benchmark size, from data item, constantly split out the data subitem that splits the benchmark size, up to the data item size less than splitting the benchmark size, separately as a data subitem, the data item size is the residue size with the data item that remains.The concrete fractionation strategy that does not wait the size data item is as follows.

For the request in the real system, obtained after the data item of request, whether the data item of judging earlier request is in broadcast queue or prepare in the formation, if, then take out the fractionation size of this data item, if do not exist, then recomputate the weight value, and calculate the SplitStandardSize of this data item, it is split according to this standard, and the off period of upgrading each data subitem.

Each data subitem to data item is judged successively: judge this data subitem whether in broadcast queue, if, then should ask to add its request queue; Otherwise judging this data subitem whether in preparing formation, if, then should ask in its request queue of adding; Otherwise add this data subitem and arrive the preparation formation as new.All judge to finish up to all data subitems of this data item like this and namely finished this data item and obtain, split and add.

Table 3 does not wait the size data item to split the example of strategy

As shown in table 3 is an example that does not wait the size data item to split strategy, d ₁Fractionation benchmark size be 3M, then size is the d of 8M ₁Split into size and be the d of 3M _1,1, size is the d of 3M _1,2With size be the d of 2M _1,3Three data subitems; d ₂Fractionation benchmark size be 3.3M, then size is the d of 5M ₂Split into size and be the d of 3.3M _2,1With size be the d of 1.7M _2,2Two data subitems.

(5.4) scheduling strategy that does not wait the size data item to split

In the periodic broadcasting scheduling of having considered the data item size, the data item size differs, if do not adopt the described size data item that waits of last chapter to split, can occur in each cycle bandwidth waste being arranged when scheduling.Therefore, the present invention is when scheduling, can select to adopt concrete scheduling strategy as required by those skilled in the art, these two kinds of different scheduling strategies for example: first kind is the scheduling strategy that the highest data item of dispatching priority LxRxW is satisfied in preferential broadcasting, and second kind is the scheduling strategy of taking all factors into consideration bandwidth availability ratio and dispatching priority.

(1) scheduling strategy of priority maximum (Priority First, PF).What dispatching priority represented is the urgency of time.Each cycle is when beginning to dispatch, server constantly chooses the data subitem of preparing to have in the formation dispatching priority LxRxW and joins broadcast queue, and the big or small summation of data subitem has surpassed that data item that this cycle can broadcast is big or small just to be stopped in broadcast queue.This tactful target is to make that the data subitem in the broadcast queue that at every turn chooses all is the data subitem of priority maximum, thereby satisfies more user's request as much as possible, makes service quality higher.

(2) the preferential the highest data item of broadcast scheduling time urgency of the scheduling strategy of priority maximum, but have the waste of bandwidth.And if be the simple bandwidth availability ratio maximum that makes, though taken full advantage of bandwidth, do not consider fully for the time factor that is directly connected to scheduling validity, so it is not an efficient strategy.The present invention takes all factors into consideration priority and the bandwidth utilization of scheduling, has proposed to take all factors into consideration the scheduling strategy PxBF(Priority and Bandwidth First of priority and bandwidth availability ratio, PxBF).

The present invention uses for reference the thought of 0-1 knapsack algorithm, the dispatching priority (LxRxW) of data item is mapped as the value of article, data item size (amount of bandwidth that namely takies) is mapped as the weight of article, the data item size that one-period can be broadcasted is the total weight that knapsack can bear, target is to be no more than under the prerequisite of the total weight that knapsack can bear in total weight, makes the total value maximum of the data item of getting.When fetching data, to each data item d _{I, j}Have only 2 kinds of selections, namely pack into knapsack (1) or the knapsack (0) of not packing into.Can not be with data item d _{I, j}Knapsack pack into repeatedly, can not be with the partial data item d that only packs into _{I, j}The typical algorithm that solves the 0-1 knapsack problem has dynamic programming algorithm and greedy algorithm.Dynamic Programming solution scheme is the optimal solution that solves the 0-1 knapsack problem, and it represents whether choosing of article with two-dimensional array.And the scale of data item of the present invention is bigger, if solve with Dynamic Programming, then space complexity is very high, so the present invention takes greedy algorithm.Calculating the dispatching priority of each data item and the ratio of data item size, again by the descending sort of this ratio, adorn knapsack since first, is second then, and the like, put as much as possible up to filling knapsack more.

Concrete algorithm when each cycle begins to dispatch, is chosen

Maximum data item is added broadcast queue to, and data item size summation has surpassed that data item that this cycle can broadcast is big or small just to be stopped in broadcast queue.If remaining bandwidth is arranged, the present invention adopts the method for supplying of bandwidth availability ratio maximum, namely selects the maximum data item that can broadcast.By formula 3-1 as can be known, $L_{d_{i,j},k,h}\×R_{d_{i,j},k,h}\×W_{d_{i,j},k,h,r}/S_{d_{i,j}}=1/S_{d_i}\×\mathrm{Weight}\left(d_{i,j}\right)$ , namely choose

The minimal data item gets final product.Specific strategy is described below:

It is as shown in table 4 to obtain gathering of optimum alpha value by experiment.

Table 4 optimum alpha value gathers

Specific embodiment described in the present invention only is that the present invention's spirit is illustrated.Those skilled in the art can make various modifications or replenish or adopt similar mode to substitute described specific embodiment, but can't depart from spirit of the present invention or surmount the defined scope of appended claims.

Claims (10)

1. data broadcast scheduling method as required in real time of considering the data item size is characterized in that: comprise following process,

Mobile client is by after uploading channel and submitting the request of oneself to, server receives request, check earlier this request whether in broadcast queue or prepare in the formation, if, then directly will ask to add in the broadcast queue or in the preparation formation in the request queue of corresponding data subitem; If do not exist, then will ask to press request time order from small to large and add in the request queue of system, and be request distribution ID;

Server takes out the request of current time from the request queue of system, search the data item that satisfies request, obtains groups of data items and is made into data item information, and described data item information comprises the ID of data item and the content of data item;

Server splits according to the data item that the big young pathbreaker of data item gets access to, and the data subitem that splits into is added in the preparation formation; A data item splits into an above data subitem, and the data item ID of these data subitems is identical, but data subitem ID difference; Preparing formation at first arranges by the ID of data item, the data subitem of same data item ID is arranged by data subitem ID, and a current request of data item is kept at the request queue that splits each data subitem of gained, and the data subitem of preparing in the formation is called the preparation item;

Server is periodically dispatched tissue by dispatching algorithm to the data of preparing in the formation, selects the data subitem to add in the broadcast queue dynamically, and the data subitem in the broadcast queue is broadcast, and the data subitem in the broadcast queue is called the broadcasting item.

2. according to the described a kind of data broadcast scheduling method as required in real time of considering the data item size of claim 1, it is characterized in that: employing waited the size data item to split mode when the data item that gets access to was split, the fractionation and the scheduling flow that split mode Deng the size data item may further comprise the steps

Step a1, initialization;

Step a2 is if not request of current time then jumps to step a7, otherwise continues to enter step a3;

Step a3, the request of taking-up current time;

Step a4 obtains the data item of this request, according to the fractionation strategy that waits the size data item it is split, and according to the adjustment strategy of off period the off period of the data item of fractionation is adjusted;

Step a5 handles respectively for all the data subitems after splitting, and processing mode is as follows,

To arbitrary data subitem wherein, if at the broadcasting item of broadcast queue and do not have broadcasting, then this request is added in the request queue of this broadcasting item; Otherwise, if the preparation item in the preparation formation then adds this request in the request queue of this preparation item; Otherwise, this data subitem is added the preparation formation as new preparation item, and will ask to add and prepare newly to prepare in the formation in the request queue of item;

Step a6 waits new request to be generated, jumps to step a2;

Step a7 broadcasts according to broadcast queue;

Step a8 judges whether just to have broadcasted not empty broadcasting item in the broadcast queue, otherwise directly enters step a9, be that the successful solicited message that then will broadcast item writes successfully the solicited message file, delete this broadcasting all requests, delete this broadcasting item, enter step a9 then;

Step a9 judges whether that broadcast queue is empty, if otherwise directly enter step a10, if broadcast queue is empty, then generate new broadcast queue according to the dispatching algorithm that waits size to split data item, enter step a10 then;

Step a10, the time point that advances if surpassed the running time of regulation, then jumps to step a11; Otherwise, jump procedure a2;

Step a11 reads successfully the solicited message file, the statistics solicited message.

3. according to the data broadcast scheduling method as required in real time of the described consideration data item of claim 2 size, its feature exists: split according to the fractionation strategy that waits the size data item among the step a4, comprise following substep,

Step a4.1, initialization comprises the big or small remainsize=datasize of residue, subitem sequence number subitemorder=1 are set; Datasize is the data item size;

Step a4.2, it is as follows to determine to split benchmark size SplitStandardSize,

$\mathrm{SplitS}\tan \mathrm{dardSize}=\frac{{\mathrm{Bw}}_k\×{\mathrm{Cy}}_k}{\mathrm{SplitK}}$

Wherein, Bw _kBe the broadcast bandwidth of the k time broadcast cycle, Cy _kIt is the broadcast cycle of the k time broadcast cycle; SplitK is a positive integer, and the data item size of the interior broadcasting of expression one-period is the number of the data subitem of fractionation benchmark size SplitStandardSize;

Step a4.3 if remainsize＞SplitStandardSize then continues step a4.4, otherwise jumps to step a4.5;

Step a4.4, add a data subitem, data subitem size is SplitStandardSize, filling size is SplitStandardSize, sequence number is subitemorder, subitemorder=subitemorder+1, remainsize=remainsize-SplitStandardSize jumps to step a4.3;

Step a4.5 adds a data subitem, and data subitem size is SplitStandardSize, and filling size is remainsize, and sequence number is subitemorder, remainsize=0;

Step a4.6 returns and splits good data subitem.

4. according to the data broadcast scheduling method as required in real time of the described consideration data item of claim 3 size, it is characterized in that: generate new broadcast queue according to the dispatching algorithm that waits size to split data item among the step a9, comprise following substep,

Step a9.1, import the k time broadcasting before the preparation formation after enter circulation,

Step a9.2 upgrades and prepares each data subitem d in the formation _{X, y}Current request number, off period and stand-by period;

Step a9.3 calculates each data subitem d _{X, y}Dispatching priority;

Step a9.4, the data subitem d that will have the greatest priority value _{X, y}Be designated as d _{N, m}And compose and give Bd _j

Step a9.5 judges that all data subitem sizes add d in the broadcast queue _{N, m}Size whether greater than Bw _k* Cy _kIf, greater than then withdraw from circulation and enter step a9.8,, otherwise continue execution in step a9.6;

Step a9.6 is with Bd _jInsert the B of broadcast queue _kAfterbody returns step a9.2;

Step a9.7, end loop;

Step a9.8 returns the B of broadcast queue the k time _k

5. according to the data broadcast scheduling method as required in real time of the described consideration data item of claim 1 size, it is characterized in that: adopt when the data item that gets access to is split and do not wait the size data item to split mode, the fractionation and the scheduling flow that do not wait the size data item to split mode may further comprise the steps

Step b1, initialization;

Step b2 is if not request of current time then jumps to step b7, otherwise continues to enter step b3;

Step b3, the request of taking-up current time;

Step b4 obtains the data item of this request, according to the fractionation strategy that does not wait the size data item it is split, and according to the adjustment strategy of off period the off period of the data item of fractionation is adjusted;

Step b5 handles respectively for all the data subitems after splitting, and processing mode is as follows,

To arbitrary data subitem wherein, if at the broadcasting item of broadcast queue and do not have broadcasting, then this request is added in the request queue of this broadcasting item; Otherwise, if the preparation item in the preparation formation then adds this request in the request queue of this preparation item; Otherwise, this data subitem is added the preparation formation as new preparation item, and will ask to add and prepare newly to prepare in the formation in the request queue of item;

Step b6 waits new request to be generated, jumps to step b2;

Step b7 broadcasts according to broadcast queue;

Step b8 judges whether just to have broadcasted not empty broadcasting item in the broadcast queue, otherwise directly enters step b9, be that the successful solicited message that then will broadcast item writes successfully the solicited message file, delete this broadcasting all requests, delete this broadcasting item, enter step b9 then;

Step b9 judges whether that broadcast queue is empty, if otherwise directly enter step b10, if broadcast queue is empty, then generate new broadcast queue according to the dispatching algorithm that does not wait size to split data item, enter step b10 then;

Step b10, the time point that advances if surpassed the running time of regulation, then jumps to step b11; Otherwise, jump procedure b2;

Step b11 reads successfully the solicited message file, the statistics solicited message.

6. according to the data broadcast scheduling method as required in real time of the described consideration data item of claim 5 size, it is characterized in that: according to not waiting fractionation strategy of size data item to split, comprise following substep among the step b4,

Step b4.1, input data item d _i, data item d _iThe data item size

, data item priority Weight (d _{I, j}), bandwidth B w _k, cycle Cy _kBack initialization comprises that remaining data item size is set is data item size, subitem sequence number subitemorder=1;

Step b4.2 calculates the fractionation benchmark size SplitStandardSize of this data item,

$\mathrm{SplitS}\tan \mathrm{dardSize}=\frac{\mathrm{\α}\×{\mathrm{Bw}}_k\×{\mathrm{Cy}}_k}{S_{d_i}\×\mathrm{Weight}\left(d_{i,j}\right)}$

Wherein, α is parameter;

Step b4.3 if current residual data item size then continues step b4.4 greater than SplitStandardSize, otherwise jumps to step b4.5;

Step b4.4 adds a data subitem, and data subitem size is SplitStandardSize, and sequence number is subitemorder, subitemorder=subitemorder+1, and remaining data item size deducts SplitStandardSize, jumps to step b4.3;

Step b4.5 adds a data subitem, and data subitem size is remaining data item size, and sequence number is subitemorder;

Step b4.6 returns and splits good data subitem.

7. according to the data broadcast scheduling method as required in real time of the described consideration data item of claim 6 size, it is characterized in that: generate new broadcast queue according to the dispatching algorithm that does not wait size to split data item among the step b9, adopt the scheduling strategy of priority maximum, comprise following substep

Step 9.1.1, after formation was prepared in input, initialization broadcast queue was empty;

Step 9.1.2 carries out following circulation,

Step 9.1.2.1, traversal is prepared formation, chooses the data subitem with greatest priority LxRxW value;

Step 9.1.2.2 does not still surpass the data subitem size that one-period can be broadcasted if the total size of the data subitem in the broadcast queue adds this data subitem size, then adds this data subitem to broadcast queue, returns step 9.1.2.1 and circulates; Otherwise, withdraw from circulation, output broadcast queue.

8. according to the data broadcast scheduling method as required in real time of the described consideration data item of claim 6 size, it is characterized in that: generate new broadcast queue according to the dispatching algorithm that does not wait size to split data item among the step b9, the scheduling strategy of priority and bandwidth availability ratio is taken all factors into consideration in employing, comprise following substep

Step 9.2.1, after formation was prepared in input, initialization broadcast queue was empty;

Step 9.2.2 carries out following circulation,

Step 9.2.2.1, traversal is prepared formation, chooses to have minimum

The data subitem of value;

Step 9.2.2.2 does not still surpass the data subitem size that one-period can be broadcasted if the total size of the data subitem in the broadcast queue adds this data subitem size, then adds this data subitem to broadcast queue, returns step 9.2.2.1 and circulates; Otherwise, withdraw from circulation and enter step 9.2.3;

Step 9.2.3, initialization residue size;

Step 9.2.4 carries out following circulation,

Step 9.2.4.1, traversal is prepared formation, is chosen at data subitem size less than data subitem maximum under the prerequisite of residue size, if find then enter step 9.2.4.2, if can not find such data subitem, then withdrawing from circulation output broadcast queue;

Step 9.2.4.2 adds the data subitem that finds to broadcast queue, returns step 9.2.4.1 and circulates.

9. according to the data broadcast scheduling method as required in real time of claim 2 or 5 described consideration data item sizes, it is characterized in that: according to the adjustment strategy of off period the off period of the data item that splits is adjusted among step a4 or the step b4, comprised following substep,

(1) input data item d _iH data item gone on the air the off period that constantly r asks during the k time broadcast cycle

, data item d _iData subitem d _{I, 1}, d _{I, 2}..., d _{I, P}Initializing variable , j=1;

(2) carry out following circulation,

(2.1) setting data subitem d _{I, j}Off period

$\mathrm{Deadline}=\mathrm{Deadline}+S_{d_{i,j}}/{\mathrm{Bw}}_k$

(2.2) if j equals data item d _iTotal number P of the data subitem that splits jumps out circulation output and adjusts the data subitem of off period, otherwise j=j+1 returns (2.1) and continues circulation.

10. according to the data broadcast scheduling method as required in real time of claim 2 or 5 described consideration data item sizes, it is characterized in that: add up solicited message among step a11 or the step b11, comprise following operation,

(1) reads successfully in the solicited message file all solicited messages to dynamic array vector;

(2) ID of the successful solicited message among the dynamic array vector according to request sorted;

(3) read the identical successful solicited message of ID of asking among the dynamic array vector successively, if the data subitem sum of the bar number of the successful solicited message of judgement and the data item of request equates, the request of then judging this ID is successfully request, adds up the access time of this success request; Otherwise this asks to be the request of losing efficacy;

(4) read the end of dynamic array vector, then withdraw from, export and successfully ask number, total access time.

Images