CN102088483B - Real-time scheduling method of sonar data stream management middleware - Google Patents

Real-time scheduling method of sonar data stream management middleware Download PDF

Info

Publication number
CN102088483B
CN102088483B CN2010105930615A CN201010593061A CN102088483B CN 102088483 B CN102088483 B CN 102088483B CN 2010105930615 A CN2010105930615 A CN 2010105930615A CN 201010593061 A CN201010593061 A CN 201010593061A CN 102088483 B CN102088483 B CN 102088483B
Authority
CN
China
Prior art keywords
task
time
scheduling
real
node
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN2010105930615A
Other languages
Chinese (zh)
Other versions
CN102088483A (en
Inventor
罗琳
方世良
罗昕炜
汪磊
邓绍庆
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CN2010105930615A priority Critical patent/CN102088483B/en
Publication of CN102088483A publication Critical patent/CN102088483A/en
Application granted granted Critical
Publication of CN102088483B publication Critical patent/CN102088483B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a real-time scheduling method of a sonar data stream management middleware, which faces to a distributed system based on real-time communication control RTCC; and according to the characteristics of sonar data streams, contents aiming at the scheduling of the sonar data streams are added at the time of reserving the characteristic of the RTTC in transmitting large volumes of data in real time. The method is mainly improved in seven parts, namely, an improvement of a central controller (CC), an improvement of a schedule table generation method, an improvement of a pre-scheduling analysis method, an improvement of a terminal online adding method, an improvement of an important-level adjustment method for a channel end data source, an improvement of a dynamic scheduling method and an improvement of an automatic adjustment method for routing node Link linkage faults. The real-time scheduling method provided by the invention is utilized to meet the core requirements for the instantaneity and the openness of the sonar data stream management middleware, realize the openness and the flexible configuration of a system, and provide seamless, expandable and stable scheduling and transmission for different sonar data streams.

Description

A kind of real-time scheduling method of sonar data flow management middleware
Technical field
The present invention relates to the software sonar system, relate in particular to a kind of real-time scheduling method of sonar data flow management middleware.
Background technology
Because the number of array element is many in the sonar system, under the sampling with high precision condition, make whole system need the data volume of real-time Transmission can be very big.In the sonar system, need to realize that the data of each array element are free and open to all application end, the fixed data management mode obviously can't meet the demands.Therefore, adopting the core requirement of sonar real time data flow management middleware is real-time and opening, on the one hand, opening requires middleware to have modular structure and corresponding standards and norms, and key issue is the definition of sonar data stream standard and the selection of management switch mode; The scheduler latency of data when on the other hand, real-time is mainly reflected in communication.The real-time performance of sonar system depends on data flow real time communication mechanism, if require sonar system to have higher real time communication efficient and network transmission throughput, need to adopt suitable data stream real time communication mechanism, wherein the data flow real-time scheduling method is a most key ring, has directly influenced the interactive performance of middleware and the real-time characteristic of whole sonar system.
The kernel scheduling method CSMA/CD of Ethernet (Carrier Sense Multiple Access/Collision Detect, csma/collision detection method) is a kind of media access control method of contention type, its basic thought is when a node will send data, monitor channel at first, if channel idle just sends data, and continue to monitor; If in data transmission procedure, listened to conflict, then stop data at once and send, waits for one period at random after, restart to attempt the transmission data.The advantage of CSMA/CD is that principle is fairly simple, technical easy realization, and each work station is on an equal footing in the network, does not need centralized control, and priority control is not provided.Shortcoming is that the time delay of transmitted data on network has randomness, and transfer of data might experience unpredictalbe time-delay, even the situation that can't send for a long time, and the assurance of the service transmission of time correlation can't be provided.Therefore, this mode is not suitable for the transmission of real time data.
And RTCC(Real Time Communication Control, the real time communication control method) be a kind of bus-type hard real-time dispatching method, be suitable for the big data quantity RTTS.According to the sonar applications demand, when data volume was more big, the speed of RTCC was more close to direct transmission, and real-time performance is more good; Aspect stable, the RTCC fundamental sum is directly transmitted unanimity.Therefore the CSMA/CD method under the Ethernet, RTCC is fit to the scheduling of sonar data flow management middleware more in conjunction with the PUSH pattern, has the better real-time property energy.
Because the processing object of data stream management middleware mainly is distributed sonar data stream, therefore, need carry out system constructing in conjunction with the characteristics of distributed system and sonar data stream.Simultaneously, the RTCC real-time communication method also needs improving accordingly towards distributed system and sonar data stream characteristics according to sonar data flow management middleware, the characteristic that keeps RTCC method big data quantity, hard real-time transmission, increase the method content that is fit to the scheduling of sonar data stream, to be fit to the scheduler module of distributed sonar data flow, satisfy sonar data flow management middleware real-time and open requirement.
Summary of the invention
Goal of the invention: in order to overcome the deficiencies in the prior art, the invention provides a kind of real-time scheduling method with sonar real time data flow management middleware of real-time controllability, for the software sonar system provides higher real time communication efficient and network transmission throughput; The online adding of terminal, the adjustment of tunnel ends data source importance level, dynamic dispatching and routing node connect the employing of the methods such as automatic adjustment of fault, have realized opening and the flexible configuration of system; And can provide seamless for different sonar data stream, can expand, stable scheduling and transmission.
Technical scheme: for achieving the above object, the technical solution used in the present invention is:
A kind of real-time scheduling method of sonar data flow management middleware, on the RTCC basis, improve towards distributed system and to sonar data stream, can be referred to as DAW-RTCC(Distributed Adaptive Weight-Real Time Communication Control, the distributed self-adaption weights are adjusted the real time communication control method), DAW-RTCC is when having kept the characteristic of RTCC real-time Transmission big data quantity data, increased the content at the scheduling of sonar data stream, mainly comprise the improvement of three parts: the improvement of master station (Central Controller is called for short CC) electoral machinery, the improvement of scheduling table generating method and the improvement of pre-scheduling analytical method.
1, the CC electoral machinery among the DAW-RTCC
The processing object of sonar data flow management middleware has two classes: CC and terminal (Remote Terminal is called for short RT), and CC mainly is responsible for the operation of the whole network of control, and RT is the object that carries out data interaction; RT can be divided into tunnel ends (Channel is called for short CHA) and application end (Application is called for short APP) again; CHA refers to the data source end, and APP is data consumption/end for process, and under specific situation, if the secondary application demand is arranged, APP also becomes CHA.
The target of CC election is distributed intelligence and the distance vector according to CHA, APP, elects optimum CC, the CC electoral machinery in DAW-RTCC, and its principle of following has two: the one, the CC that elects logically distinguishes with CHA and APP; The 2nd, CC in scheduling with the communication delay minimum of each node.
2, the scheduling table generating method among the DAW-RTCC
Because the stable treated performance of Universal Signal Processor platform, can off line generate dispatch list, namely before carrying out data stream scheduling, and require to generate dispatch list with reference to static scheduling simultaneously, in a single day dispatch list generates and just can not change, each RT carries out periodic scheduling in strict accordance with dispatch list, guarantees the controlled real-time of system.
With reference to the scheduling mechanism among the RTCC, the dispatch list generating algorithm of DAW-RTCC is revised according to the sonar data properties of flow, to guarantee the real-time performance of each data flow.According to the PUSH model, dispatch list is based on CHA, with reference to the data standard of CHA, provides DAW-RTCC dispatch list generation scheme and is, for i task M in the dispatch list i, should satisfy following condition:
U(M i)≥U(A j),j=1,2,...,N AM
D(M i)≤D(A j),j=1,2,...,N AM
T S≤B(A j),j=1,2,...,N A
T E ( M i ) = T S * F S ( M i ) Σ i = 1 N M F S ( M i )
U(M i)≥U(M i+1),i=1,2,...,N M
Wherein, i represents task number, i=1, and 2 ... N M, A represents application end; U represents urgency; D represents time-delay, and unit is ms; T SExpression TDMA(time division multiplexing) dispatching cycle, unit is ms; B represents beat, and unit is ms; F SThe expression sample frequency, unit is KHz; T ETime slot is carried out in expression, and unit is ms; N ATerminal number is used in expression; N AMExpression application M iThe application terminal number of task; N MThe expression number of tasks.
3, the pre-scheduling analytical method among the DAW-RTCC
Described task M iTime response can adopt two tuple (C i, D i) sign, wherein, C iThe actual execution time of a task scheduling of expression, D iRepresent the time limit that this task is once dispatched.Because the task queue that needs to dispatch in the static scheduling pattern is fixed, the corresponding passage of a task in the dispatch list, because each channel data sample rate difference, so the C of each task iAnd D iAll be different.The C that can find each task correspondence is analyzed in pre-scheduling iAnd D iValue under the prerequisite that guarantees Real-Time Scheduling, improves the reliability of system call as much as possible, namely increases D i-C iThe time limit redundancy, prevent network burst shake.
For dispatch list, carry out time slot T EEmbodied time limit D iThought, according to the scheduling table generating method among the DAW-RTCC, dispatch list carries out the pre-scheduling analysis through after the pre-scheduling according to following steps:
(a1) judge Whether set up, if judged result is for being, task M then iThe time limit redundancy is arranged, can satisfy the real-time demand of all APP, T 0Be recorded in the dispatch list as overtime amount;
(a2) if judged result is not in the step (a1), then delete the task of urgency minimum after, return step (a1).
Because DSP has the performance of stable transmission high amount of traffic, actual execution time C iIn the pre-scheduling first time is analyzed, can use theoretical value, in the pre-scheduling of back is analyzed, can revise according to each TDMA periodic duty implementation status and routine weight value.
In the real-time scheduling method of sonar data flow management middleware provided by the invention, also the automatic adjusting method that the importance level method of adjustment of the online adding method of terminal, tunnel ends data source, dynamic dispatching method, routing node Link are connected fault has been carried out relevant improvement.
4, the online adding method of the terminal among the DAW-RTCC
Can search out the online adding that suitable time slot is realized terminal by the online adding method of terminal, behind a TDMA end cycle, CC waits for the clock control power of fetching, wait for that clock triggers the next cycle task, at this moment, the hard real-time condition of all tasks of last one-period is all satisfied, and system is in complete soft real-time status, can carry out the online adding of terminal.In the online adding method of terminal among the described DAW-RTCC, the clock of online adding triggers T constantly ASatisfy following condition:
T A = T 0 + T S , T 0 ≤ T R ≤ T 0 + T S T R , else
Wherein, T 0The expression initial moment of scheduling; T RRepresent online joining request constantly.
Clock by above-mentioned online adding triggers T constantly AThe condition that satisfies as can be seen, online joining request can be carried out at any time, realized summary responses in logic, the flexibility that has improved system; Simultaneously, the clock that has guaranteed the periodic scheduling module in real process is not subjected to the influence of online adding, and namely online joining request must wait for that at least one TDMA just can be responded cycle time, makes the Real-Time Scheduling performance of system obtain guarantee.
5, the importance level method of adjustment of the tunnel ends data source among the DAW-RTCC
In actual conditions, the significance level of a certain data flow is mainly determined by two side's factors: the one, use the urgency level to this request of data; The 2nd, the number of the application that this data flow is asked.Therefore, in the dispatch list generative process, the priority of calculating certain data flow is to take into account the factor of these two aspects simultaneously, to guarantee that important task can be performed when scheduling resource is nervous.The importance level of the tunnel ends data source among the DAW-RTCC is adjusted in the following manner:
Setting task M i(data flow) application be applied as M i={ A j| j=1,2 ..., n}, n represent application number total in this task,
Task M iThe middle urgency of using is { u j| j=1,2 ..., n} is at first to task M iThe middle urgency of using is carried out descending, makes:
{ u j| j=1,2 ..., n and u 1〉=u 2〉=... 〉=u n}
Obtain task M according to following formula again iImportance level
Figure GDA00002610511300042
Wherein, j represents to use label, and n represents application number total in this task.
6, the dynamic priority scheduling method among the DAW-RTCC
Generate in the scheduling in static state, the TDMA cycle of whole system is got minimum applications exchange beat, TDMA=Min{T i(i=1,2 ..., n), T wherein iExpression applications exchange beat; But carry out periodic scheduling according to this static priority, the efficiency of transmission of data is not high.
Adopt the dynamic dispatching priority dispatching method, can improve the efficiency of transmission of data on a certain degree, its basic thought is: in a period of time, and the task that sample rate is more high, the number of times that transmits data is more many; The task that sample rate is more low, the number of times that transmits data is more few.Adopt the dynamic dispatching priority approach, make the data of the task that sample rate is low carry out certain buffer memory earlier, after data volume acquires a certain degree number, transmit again; In general, just can reduce the number of times of shaking hands, improve efficiency of transmission.The dynamic priority scheduling method specifically comprises the steps:
(b1) pass through
Figure GDA00002610511300051
Calculation task M iDynamic priority (Priority), wherein, f iExpression task M iSample frequency,
Figure GDA00002610511300052
Expression rounds downwards, and n is application number total in this task;
(b2) all tasks are sequentially arranged on the chained list according to the descending order of dynamic priority;
(b3) gauge outfit of the chained list that obtains from step (b2) obtains a task, carry out transfer of data after, again this task is put into chained list according to the size of dynamic priority.
During initialization, the dynamic priority of all tasks equals the relative value of sample rate (being static priority).
The adaptive scheduling of task namely is that above-mentioned chained list is safeguarded, the task that the gauge outfit of chained list is placed must be the highest task of current dynamic priority.Any one task, its dynamic priority is initially the relative value of its sample rate, and its position on chained list is every by a task preemption with higher dynamic priority once, and its dynamic priority will increase.What this mechanism had guaranteed that the task of a low priority do not stopped seizes, and its dynamic priority meeting constantly increases, have eventually certain the time can be enough big, thereby obtain the chance of execution.It is after carrying out, and its dynamic priority can be set to static priority again again, begins the queuing process of a new round again.Thus, adopt the dynamic priority scheduling method as can be seen, the task that static priority is high can obtain more execution chance, and the low task of static priority can guarantee to obtain less execution chance, but will wait for the long period.
7, the routing node Link among the DAW-RTCC connects the automatic adjusting method of fault
As node N iCertain Link mouth, in running, owing to the burst reason breaks down, its adjacent node is shown as unreachable, namely message can not arrive.For example, node N 1Data to be forwarded to application end (destination), the next node of its optimal path is node N 2But, node N 1To node N 2Send data, and node N 2But do not receive, just can think node N 1To node N 2Link connect and to break down.
At above-mentioned unreachable fault, the present invention proposes a kind of automatic adjusting method of routing node Link connection fault to address the above problem, and described method is as follows: at first judge the node N in the optimal path iAnd next node N jBetween Link communication whether break down, as judged result for not, then proper communication; If judged result is for being, then at node N iOther adjacent nodes in select an optimum adjacent node N who arrives the destination k, and with node N iData be transmitted to this adjacent node N kAnd the arrival destination, and revise local routing table.
The automatic adjusting method that described routing node Link connects fault also can be considered as a kind of local auto-adaptive and adjust mechanism, and the path that chooses by this mechanism is not necessarily optimum on the whole, but it is optimum in the part.
Beneficial effect: the real-time scheduling method of a kind of sonar data flow management middleware provided by the invention compared with prior art, has following having a few:
(1) compares CSMA/CD and RTCC, DAW-RTCC flows at distributed sonar data, characteristic with big data quantity, hard real-time transmission, increased the method content of suitable sonar data stream scheduling, to be fit to the scheduler module of distributed sonar data flow, can satisfy sonar data flow management middleware real-time and open core requirement;
(2) the online adding of terminal and the Weight number adaptively method of adjustment of DAW-RTCC employing in the Real-Time Scheduling that has guaranteed system, have improved the flexibility of sonar real time data flow management middleware system, have satisfied the open requirement of sonar system;
(3) the importance level method of adjustment of the tunnel ends data source of DAW-RTCC employing, by the Weight number adaptively adjustment, implementation status according to last periodic scheduling, revise the weights of current period scheduling adaptively, make that the generation of dispatch list is more reasonable and efficient, to guarantee that important task can be performed when scheduling resource is nervous;
(4) the dynamic priority scheduling method that adopts of DAW-RTCC, transmission quantity that can balanced each data makes the low task of sample rate after carrying out a period of time buffer memory, the data of just having an opportunity to transmit; Can guarantee that each data quantity transmitted can be too not little, is conducive to improve efficiency of transmission; Can guarantee that again data in buffer transmits timely, prevent that effectively buffer memory from overflowing and cause loss of data;
(5) the routing node Link that adopts of DAW-RTCC connects the automatic adjusting method of fault, can overcome fault, guarantee the optimization of local path, not only reliably but also rapidly data has been sent to the destination.
Description of drawings
Fig. 1 is DAW-RTCC scheduling overview flow chart of the present invention;
Fig. 2 is DAW-RTCC initial frame structure of the present invention;
Fig. 3 is DAW-RTCC scheduling frame structure of the present invention;
Fig. 4 is the concrete scheduling flow of DAW-RTCC of the present invention;
Fig. 5 is DAW-RTCC periodic scheduling model of the present invention;
Fig. 6 connects the self-adjusting neighbours territory figure of fault for routing node Link.
Embodiment
Below in conjunction with accompanying drawing the present invention is done further explanation.
A kind of real-time scheduling method of sonar data flow management middleware, on the RTCC basis, improve towards distributed system and according to sonar data stream characteristics, be referred to as DAW-RTCC, when having kept the characteristic of RTCC real-time Transmission big data quantity data, increased the content at the scheduling of sonar data stream, mainly comprise the improvement of seven parts: the improvement of CC electoral machinery, the improvement of scheduling table generating method, the improvement of pre-scheduling analytical method, the improvement of the online adding method of terminal, the improvement of the importance level method of adjustment of tunnel ends data source, the improvement of dynamic dispatching method is connected the improvement of the automatic adjusting method of fault with routing node Link, it is as follows specifically to improve one's methods:
1, the CC electoral machinery among the DAW-RTCC
The principle that the CC electoral machinery is followed has two: the one, and the CC that elects logically distinguishes with CHA and APP; The 2nd, CC in scheduling with the communication delay minimum of each node.
2, the scheduling table generating method among the DAW-RTCC
Each terminal is carried out periodic scheduling according to dispatch list, and described dispatch list off line generates, for i task M in the dispatch list i, satisfy following condition:
U(M i)≥U(A j),j=1,2,...,N AM
D(M i)≤D(A j),j=1,2,...,N AM
T S≤B(A j),j=1,2,...,N A
T E ( M i ) = T S * F S ( M i ) Σ i = 1 N M F S ( M i )
U(M i)≥U(M i+1),i=1,2,...,N M
Wherein, i represents task number, i=1, and 2 ... m, A represents application end; U represents urgency; D represents time-delay, and unit is ms; T SIn the expression time division multiplexing dispatching cycle, unit is ms; B represents beat, and unit is ms; F SThe expression sample frequency, unit is KHz; T ETime slot is carried out in expression, and unit is ms; N ATerminal number is used in expression; N AMExpression application M iThe application terminal number of task; N MThe expression number of tasks.
3, the pre-scheduling analytical method among the DAW-RTCC
Described task M iTime response adopt two tuple (C i, D i) sign, wherein, C iThe actual execution time of a task scheduling of expression, D iRepresent the time limit that this task is once dispatched, dispatch list is through after the pre-scheduling, and the pre-scheduling analysis is carried out according to following steps:
(a1) judge
Figure GDA00002610511300081
Whether set up, if judged result is for being, task M then iThe time limit redundancy is arranged, can satisfy the real-time demand of all APP, T 0Be recorded in the dispatch list as overtime amount;
(a2) if judged result is not in the step (a1), then delete the task of urgency minimum after, return step (a1).
Because DSP has the performance of stable transmission high amount of traffic, actual execution time C iIn the pre-scheduling first time is analyzed, can use theoretical value, in the pre-scheduling of back is analyzed, can revise according to each TDMA periodic duty implementation status and routine weight value.
4, the online adding method of the terminal among the DAW-RTCC
The clock of online adding triggers T constantly ASatisfy following condition:
T A = T 0 + T S , T 0 ≤ T R ≤ T 0 + T S T R , else
Wherein, T 0The expression initial moment of scheduling; T RRepresent online joining request constantly.
5, the importance level method of adjustment of the tunnel ends data source among the DAW-RTCC
Setting task M i(data flow) application be applied as M i={ A j| j=1,2 ..., n}, task M iThe middle urgency of using is { u j| j=1,2 ..., n} is at first to task M iThe middle urgency of using is carried out descending, makes:
{ u j| j=1,2 ..., n and u 1〉=u 2〉=... 〉=u n}
Obtain task M according to following formula again iImportance level
Figure GDA00002610511300083
Wherein, j represents to use label, and n represents application number total in this task.
6, the dynamic priority scheduling method among the DAW-RTCC
The dynamic priority scheduling method specifically comprises the steps:
(b1) pass through
Figure GDA00002610511300084
Calculation task M iDynamic priority (Priority), wherein, f iExpression task M iSample frequency,
Figure GDA00002610511300085
Expression rounds downwards, and n is application number total in this task;
(b2) all tasks are sequentially arranged on the chained list according to the descending order of dynamic priority;
(b3) gauge outfit of the chained list that obtains from step (b2) obtains a task, carry out transfer of data after, again this task is put into chained list according to the size of dynamic priority.
During initialization, the dynamic priority of all tasks equals the relative value of sample rate (being static priority).
7, the routing node Link among the DAW-RTCC connects the automatic adjusting method of fault
The automatic adjusting method that routing node Link connects fault is as follows: at first judge the node N in the optimal path iAnd next node N jBetween Link communication whether break down, as judged result for not, then proper communication; If judged result is for being, then at node N iOther adjacent nodes in select an optimum adjacent node N who arrives the destination k, and with node N iData be transmitted to this adjacent node N kAnd the arrival destination, and revise local routing table.
Be illustrated in figure 1 as the Real-Time Scheduling schematic flow sheet of DAW-RTCC, based on the scheduling model of DAW-RTCC method three big nucleus modules arranged: start module, dispatch list generation module and scheduler module.Wherein, start module and finish initial work, comprise definition DSP number, generation pass and distance vector, channel application registration etc.; The dispatch list generation module generates the dispatch list relevant information, comprises dispatch list generation, pre-scheduling analysis, dispatch list distribution; Scheduler module is nucleus module, carries out periodic scheduling according to dispatch list.
Be illustrated in figure 2 as the DAW-RTCC initial frame structure that the present invention proposes, initial frame is to carry out initialized command frame in the dispatch list generation module.The definition of RTCC frame head decides according to the bus-type affairs, and DAW-RTCC initial frame structure must be made related amendments to the definition of RTCC frame head, the DAW-RTCC initial frame comprises 8 float types, totally 32 bytes, first three float represents frame type (Type), frame function (Code) and data packet length (Length) respectively, and back five float are that system keeps the position.
Be illustrated in figure 3 as the scheduling frame structure, the scheduling frame is the command frame of dispatching in the scheduler module, preceding four float of scheduling frame represent frame type (Type), frame function (Code), data packet length (Length) and task number (TaskID) respectively, the content of back is the scheduling time (Time) of each task, and the float number of distribution is exactly number of tasks.As seen from Figure 3,, because number of tasks is fixed, therefore dispatch the length of frame and fix in dispatching cycle at a TDMA; And because the influence of online adding can make the number of tasks of adjacent TDMA in dispatching cycle inconsistent, the length of therefore dispatching frame can be according to number of tasks and the self adaptation adjustment.
Be illustrated in figure 4 as the Real-Time Scheduling particular flow sheet of DAW-RTCC, the Real-Time Scheduling process is as follows:
(1) CC starts scheduling, takes out first task of overall scheduling table, and sends " tunnel ends data transfer command frame (being called for short the START frame) " to corresponding CHA, and clock control power is handed to this CHA;
(2) if CHA has data to send, then transmit the START frame and give route/application, send Frame (being called for short the DATA frame) afterwards, enter step (3); If there are not data, transmit the tunnel ends that the START frame is given the next task correspondence, forward step (6) to;
(3) after the START frame is received in route/application, be configured according to frame information, carry out the reception of DATA frame according to the dispatch list task attribute, and judge whether to transmit again;
(4) APP (S) is received desired data, to the tunnel ends transmission " application end finishes receiving command frame (being called for short the OVER frame) " of next task correspondence;
(5) tunnel ends of next task correspondence, and differentiate for the OVER frame, in case using number with the dispatch list application, the OVER frame number of receiving conforms to, change step (6) over to;
(6) clock control of the CHA of current task power is invalid, and the CHA of next task correspondence obtains clock control power, waits for that clock triggers next task;
After all tasks in (7) TDMA cycles finished, CC fetched clock control power, waited for that clock triggers next TDMA periodic duty.
Be illustrated in figure 5 as the periodic scheduling model, this model depends on distributed network, and dispatch list all is stored in this locality/node, has realized the distribution of information.By accompanying drawing 5 as can be seen, the direct benefit that middleware brings is to have eliminated each node physically and isomerism in logic, by loading corresponding middleware module, can be undertaken transparent mutual by distributed network.Simultaneously, Each performs its own functions in a certain distribution time slot to make each node towards the communication control process (process in the accompanying drawing 5 1., 2., 3., 4., 5.) of sonar design, guaranteed the hard real-time performance of sonar data stream demand on the basis of DSP stable transfer performance, shared for sonar data and provide convenience.
Be to connect neighbours territory figure in the automatic adjustment of fault by node Link as shown in Figure 6, suppose p (s, d)={ N 1, N 2, N 3..., N m, expression data source nodes s to the optimal path of using (destination) node d the routing node of process, and this path is N 1-N 2-N 3-...-N mOwing to the burst reason makes N I+1Break down, this shortest path just was one and opened circuit that (place of opening circuit is N this moment so i-N I+1).This paper defines adjacent node (Adjacency Node) and is DSP number of this node neighbours territory interior nodes, if certain direction does not have node, then with-1 expression.N iAdjacent node (Adjacent Node), AdtNode={N I0, N I1, N I2, N I3.This is to walk around fault point N so I+1, forward the data to the destination.So with regard to N iOther adjacent nodes in find an optimal path.This optimal path should satisfy following three conditions:
(a) the starting point N in this path Ij≠ N I+1
(b) p (N Ij, d)=Optimal{p (N Ik, d) | k=0,1,2,3}, Optimal represents optimum;
(c)
Figure GDA00002610511300111
Starting point is from N IjBegin to count.
Were it not for the path of satisfying above-mentioned condition, N then is described I+1The fault in path causes this task not finish; Because it has made N iUnreachable to destination node.
From accompanying drawing as can be seen, the present invention has carried out the structure of Real-Time Scheduling system in conjunction with the characteristics of distributed system and sonar data stream, propose the distributed self-adaption weights and adjusted real time communication control DAW-RTCC agreement, adopt the online adding of terminal, the adjustment of tunnel ends data source importance level, dynamic dispatching and routing node to connect the methods such as automatic adjustment of fault, realize open and data stream management flexibly, make whole system more efficient and rational.Through test, adopting data management middleware scheduling time of real-time scheduling method of the present invention only to account for gathers around 2.614% to 4.375% of the overall data transmission time, and adopting method of the present invention to realize the shared flexibility of sonar data and controlled real-time, the entire system efficiency of transmission is very high.
The above only is preferred implementation of the present invention; be noted that for those skilled in the art; under the prerequisite that does not break away from the principle of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (5)

1. the real-time scheduling method of a sonar data flow management middleware, based on real time communication control RTCC, it is characterized in that: the master station electoral machinery is as follows: at first select logically the node that distinguishes mutually with tunnel ends and application end, from the node of selecting, select in scheduling the node with the communication delay minimum of each node again, as final master station;
Each terminal is carried out periodic scheduling according to dispatch list, and described dispatch list off line generates, for i task M in the dispatch list i, satisfy following condition:
U(M i)≥U(A j),j=1,2,...,N AM
D(M i)≤D(A j),j=1,2,...,N AM
T S≤B(A j),j=1,2,...,N A
T E ( M i ) = T S * F S ( M i ) Σ i = 1 N M F S ( M i )
U(M i)≥U(M i+1),i=1,2,...,N M
Wherein, i represents task number, i=1, and 2 ... N M, A represents application end; U represents urgency; D represents time-delay, and unit is ms; T SIn the expression time division multiplexing dispatching cycle, unit is ms; B represents beat, and unit is ms; F SThe expression sample frequency, unit is KHz; T ETime slot is carried out in expression, and unit is ms; N ATerminal number is used in expression; N AMExpression application M iThe application terminal number of task; N MThe expression number of tasks;
Described task M iTime response adopt two tuple (C i, D i) sign, wherein, C iThe actual execution time of a task scheduling of expression, D iRepresent the time limit that this task is once dispatched, dispatch list is through after the pre-scheduling, and the pre-scheduling analysis is carried out according to following steps:
(a1) judge
Figure FDA00002610511200012
Whether set up, if judged result is for being, task M then iThe time limit redundancy is arranged, T 0Be recorded in the dispatch list as overtime amount;
(a2) if judged result is not in the step (a1), then delete the task of urgency minimum after, return step (a1).
2. the real-time scheduling method of sonar data flow management middleware according to claim 1 is characterized in that: in the described real-time scheduling method, the clock of online adding triggers T constantly ASatisfy following condition:
T A = T 0 + T S , T 0 ≤ T R ≤ T 0 + T S T R , else
Wherein, T 0The expression initial moment of scheduling; T RRepresent online joining request constantly.
3. the real-time scheduling method of sonar data flow management middleware according to claim 1, it is characterized in that: in the described real-time scheduling method, the importance level of tunnel ends data source is adjusted in the following manner: set task M iApplication be applied as M i={ A j| j=1,2 ..., n}, n represent application number total in this task, task M iThe middle urgency of using is { u j| j=1,2 ..., n} is at first to task M iThe middle urgency of using is carried out descending, makes:
{ u j| j=1,2 ..., n and u 1〉=u 2〉=... 〉=u n}
Obtain task M according to following formula again iImportance level
Figure FDA00002610511200021
Wherein, j represents to use label, and n represents application number total in this task.
4. the real-time scheduling method of sonar data flow management middleware according to claim 1 is characterized in that: in the described real-time scheduling method, adopt the dynamic priority scheduling method, described dynamic priority scheduling method comprises the steps:
(b1) pass through
Figure FDA00002610511200022
Calculation task M iDynamic priority, wherein, f iExpression task M iSample frequency,
Figure FDA00002610511200023
Expression rounds downwards, and n is application number total in this task;
(b2) all tasks are sequentially arranged on the chained list according to the descending order of dynamic priority;
(b3) gauge outfit of the chained list that obtains from step (b2) obtains a task, carry out transfer of data after, again this task is put into chained list according to the size of dynamic priority.
5. the real-time scheduling method of sonar data flow management middleware according to claim 1, it is characterized in that: in the described real-time scheduling method, adopt routing node Link to connect the automatic adjusting method of fault, described method is as follows: at first judge the node N in the optimal path iAnd next node N jBetween Link communication whether break down, as judged result for not, then proper communication; If judged result is for being, then at node N iOther adjacent nodes in select an optimum adjacent node N who arrives the destination k, and with node N iData be transmitted to this adjacent node N kAnd the arrival destination, and revise local routing table.
CN2010105930615A 2010-12-17 2010-12-17 Real-time scheduling method of sonar data stream management middleware Expired - Fee Related CN102088483B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010105930615A CN102088483B (en) 2010-12-17 2010-12-17 Real-time scheduling method of sonar data stream management middleware

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010105930615A CN102088483B (en) 2010-12-17 2010-12-17 Real-time scheduling method of sonar data stream management middleware

Publications (2)

Publication Number Publication Date
CN102088483A CN102088483A (en) 2011-06-08
CN102088483B true CN102088483B (en) 2013-07-10

Family

ID=44100095

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010105930615A Expired - Fee Related CN102088483B (en) 2010-12-17 2010-12-17 Real-time scheduling method of sonar data stream management middleware

Country Status (1)

Country Link
CN (1) CN102088483B (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101098179A (en) * 2006-06-30 2008-01-02 中国科学院声学研究所 Method for forming broadband frequency domain digital beam

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2438355C (en) * 2005-03-02 2010-02-10 John Jamieson A communication system and a method of providing communication services

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101098179A (en) * 2006-06-30 2008-01-02 中国科学院声学研究所 Method for forming broadband frequency domain digital beam

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
声纳数据流管理系统的查询策略;孙亮亮等;《声学技术》;20081031;第27卷(第5期);第442-443页 *
孙亮亮等.声纳数据流管理系统的查询策略.《声学技术》.2008,第27卷(第5期),

Also Published As

Publication number Publication date
CN102088483A (en) 2011-06-08

Similar Documents

Publication Publication Date Title
CN108777660B (en) Method for scheduling service in time-triggered FC network
Raagaard et al. Runtime reconfiguration of time-sensitive networking (TSN) schedules for fog computing
CN105245301B (en) A kind of airborne optical-fiber network analogue system based on time triggered
CN111030835B (en) Task scheduling model of TTFC network and message scheduling table generation method
CN106921591B (en) Method for processing critical service burst in airborne navigation electric network
CN109995675B (en) Self-adaptive industrial Ethernet gateway system and method based on software definition
CN102106116B (en) Technique for communicating between a plurality of nodes
CN101729386B (en) Flow control method and device based on token scheduling
CN104836750A (en) Data center network flow scheduling method based on round-robin
KR102178880B1 (en) Network system and data trasmission method based on device clustering in lorawan communication
CN104243536B (en) A kind of bullet train controls network semi-matter simulating system
CN113472811B (en) Heterogeneous service function chain forwarding protocol and method in intelligent fusion identification network
KR930001618A (en) Synchronous Node Controller for Switching Networks
CN100430847C (en) Method and device for determining time in a bus system and corresponding bus system
US20230353425A1 (en) Method, program, medium, and device for interconnecting primary network domain with secondary network domain through gateway device
CN114710571B (en) Data packet processing system
CN106411741B (en) Medium access control MAC address learning method, controller and switching equipment
Feng et al. SMT-based Task-and Network-level Static Schedule for Time Sensitive Network
CN105141489B (en) A kind of real-time data acquisition system for emulator
CN101431458A (en) CSMA industrial control network scheduling method based on P persisting policy
CN102088483B (en) Real-time scheduling method of sonar data stream management middleware
CN108462653B (en) TTE-based rapid protocol control frame sending method
CN101478826B (en) Communication scheduling method and system for wireless sensor network
CN103457815A (en) Method for constructing moment dispatch list suitable for SpaceWire network
CN109361653A (en) A kind of POWERLINK main website

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20130710

Termination date: 20151217

EXPY Termination of patent right or utility model