RU2794969C1 - Method, device and electronic device for task processing - Google Patents

Abstract

FIELD: network topology.

SUBSTANCE: method, devices and computer-readable medium for processing tasks applied to a network topology. In the method, the target data and the target algorithm are obtained, which are necessary to perform the target task; obtaining at least one first type node capable of providing said target data in said network topology, and at least one second type node capable of executing said target algorithm; selection of a node that represents a set of said target data as the first target node from at least said one node of the first type, and a node that represents a set of said target algorithms as a second target node from at least one said node of the second type; controlling said second target node to use said target algorithm to process said target data at said first target node; wherein the node that provides the set of said target data is selected from at least said one node of the first type as the first target node, and the node that provides the set of said target algorithms is selected from at least said one node of the second type as a second target node, receiving a combination formed by a node that provides a set of said target data and a node that provides a set of said target algorithms; based on the capability information of each combination, the target combination is selected, and the node that provides the set of said target data in said target combination is defined as said first target node, and the node that provides the set of said target algorithms in said target combination is defined as second target node; wherein said combination capability information includes a time delay between a node that provides a set of said target data in said combination and a node that provides a set of said target algorithm; or said combination capability information includes the time delay between the node that provides the set of said target data in said combination and the node that provides the set of said target algorithm, as well as the length of time required by the node that provides the set of said target algorithms to execute said target algorithm; or said capability information includes the time delay between the node that provides the set of said target data in said combination and the node that provides the set of said target algorithm, as well as the size of the cache space for the node that represents the set of said target algorithms.

EFFECT: increase of processing performance of computing tasks.

9 cl, 12 dwg, 1 tbl

Description

This application requires the priority of a Chinese patent application filed with the China Patent Bureau on January 23, 2020, application number: 202010076228.4, and the title of the invention: "Method, Apparatus and Electronic Apparatus for Processing Tasks", and priority of the Chinese patent application filed with the China Patent Bureau China January 23, 2020, application number: 202010076231.6, invention title: "Method, device and electronic device for data transmission", the entire content of which is given in this application.

Technical field

The present invention relates to the field of communication technology, in particular to a method, apparatus and electronic device for task processing.

State of the art

With the rapid development of the Internet, more and more data is emerging. Currently, the network architecture uses a distributed database and the data is stored in different nodes. In this case, when a certain target node needs to perform a certain task, it will look for nodes in the network that can provide the data necessary to complete this task, and then read the corresponding data from these nodes, and then the target node will perform some algorithmic manipulation of the read data.

From this, it can be seen that in the prior art, after the target node has collected all the data necessary to complete the task, the data is processed. However, when this task requires a large amount of data and complex algorithms, it is required that the target node has the computing power for the entire network topology, which increases the cost of building the network topology.

The essence of the invention

The present invention provides a task processing method, apparatus, and electronic apparatus to solve, to a certain extent, the problem that a node performing a given task must have processing power when the task to be performed includes a large amount of data, thereby increasing the cost of designing the entire network topology.

First, an embodiment of the present invention provides a task processing method that applies to a network topology. Said network topology contains a plurality of nodes; said task handling method includes:

Obtaining the target data and the target algorithm required to complete the target task;

Obtaining at least one node of the first type capable of providing said target data in the mentioned network topology, and at least one node of the second type capable of executing the said target algorithm.

Selection of a node that represents a set of said target data as the first target node from at least one node of the first type, and a node that represents a set of said target algorithms as the second target node from at least one mentioned node of the second type;

Controlling said second target node to use said target algorithm to process said target data at said first target node.

Secondly, an embodiment of the present invention also provides a task processor that is applied to a network topology, said network topology contains a plurality of nodes; said task processor includes:

The first acquisition module, used to obtain the target data and the target algorithm needed to complete the target task;

A second acquisition module used to obtain at least one first type node capable of providing said target data in said network topology and at least one second type node capable of executing said target algorithm;

A selection module used to select a node that represents the set of said target data as the first target node from at least said one node of the first type, and select a node that represents the set of said target algorithms as the second target node from, at least said one node of the second type;

A control module used to control said second target node using said target algorithm to process target data at said first target node.

Thirdly, an embodiment of the present invention also provides an electronic device that includes a processor, memory and program code stored in said memory and executed on said processor, implementing the task processing method described above.

Fourth, an embodiment of the present invention also provides a computer program including program code. executed on the computing processing device by implementing the above task processing method.

Fifth, an embodiment of the present invention also provides a computer-readable storage medium on which the above-mentioned program code is stored.

The task processing method illustrated by the exemplary embodiment of the present invention obtains the target data and the target algorithm necessary to perform the target task, and determines at least one first type node in the network topology that can provide the target data, and at least one node of the second type, which can perform the target task, thereby selecting the first type node and the second type node, and controls the selected second type node to use the target algorithm to process the target data provided by the selected first type node to achieve the target task.

From this, it can be seen that the embodiment of the present invention distributes data and algorithms separately over the network, instead of storing data evenly for the node that needs to perform the target task, and the target algorithm is no longer executed by the node that needs to perform the target task itself, but is assigned to a node with the processing power of the target algorithm to execute, that is, an embodiment of the present invention combining distributed database and distributed computing, which can reduce the capabilities of one node in the network topology, thereby reducing the cost of designing one node, and moreover, reducing the cost of design of the entire network topology.

The above description is only an overview of the technical solutions of the present invention. In order to more clearly understand the technical means of the present invention and implement them in accordance with the content of the description, and so that the above and other objects, features and advantages of the present invention are more obvious and understandable, specific embodiments of the present invention are given below.

Description of drawings

In order to more clearly illustrate the technical solution in the exemplary embodiment of the present invention or the prior art, the accompanying drawings to be used in the exemplary embodiment or description of the prior art are briefly described below.

Fig. 1 is a process flow diagram of a task processing method provided by an embodiment of the present invention;

Fig. 2 is a schematic map of a single layer network structure in a network topology in an embodiment of the present invention;

Fig. 3 is a process flow diagram of another task processing method provided by an embodiment of the present invention; Fig. 4 is a first schematic map of the distribution of nodes in the network topology involved in the addressing process in an embodiment of the present invention;

Fig. 5 is a second schematic map of the distribution of nodes in the network topology involved in the addressing process in an embodiment of the present invention;

Fig. 6 is a third schematic map of the distribution of nodes in the network topology involved in the addressing process in an embodiment of the present invention;

Fig. 7 is a fourth schematic map of the distribution of nodes in the network topology involved in the addressing process in an embodiment of the present invention;

Fig. 8 is a schematic diagram of a network topology design in an exemplary embodiment of the present invention;

Fig. 9 is a block diagram of a task processing device provided by an embodiment of the present invention;

Fig. 10 is a block diagram of another task processing device provided by an exemplary embodiment of the present invention;

Fig. 11 is a block diagram of an electronic device provided by an embodiment of the present invention.

In FIG. 12 schematically shows a memory unit for maintaining or carrying program code for implementing a method in accordance with the present application.

Specific embodiments

Below will be references to the accompanying drawings for a more detailed description of exemplary embodiments of the present invention. Although the accompanying drawings show an exemplary embodiment of the present invention, it should be understood that the present invention may be embodied in various forms and should not be limited to the exemplary embodiment described herein. On the contrary, these embodiments are provided in order to be able to more deeply understand the present invention and to be able to fully convey the scope of the present invention to a person skilled in the art.

An exemplary embodiment of the invention is a task processing method that is applied to said network topology, which includes a plurality of nodes. Alternatively, said network topology includes a control node, and said control node stores capability information of nodes other than said control node in said network topology, that is, other nodes other than the control node can regularly report their respective information about capabilities of the control node, then the task processing method of the exemplary embodiment of the present invention can be performed by said control node.

Alternatively, as shown in FIG. 2, said network topology includes: at least a single layer network structure, each layer of the network structure includes at least one ring structure, and each ring structure includes a plurality of nodes; wherein the nodes in each ring structure are connected to each other, and in accordance with the numbering order of the nodes in each ring structure, odd-numbered nodes are connected to the next node of the neighboring node, or nodes with an even number are connected to the next node of the neighboring node. Here, it is understood that the specific connection method of the ring structure is not limited to this, but also the plurality of nodes included in each ring structure are connected at the end and connected to the next node of its neighboring node.

From this, it can be seen that said network topology includes a fault-tolerant network with several consecutive layers and loops, in which each layer has the same structure, each layer uses the same loop topological structure, and the layers cross vertically. This kind of repeated superposition of the same hardware and the same structure, combined with the same software control, gradually expands the degree of resource aggregation and completes the application-oriented super project. In addition, in this structure, there is no connection between more than two levels, which simplifies the software structure and system management. Communication rings with varying level intervals ensure that if any two points in the ring fail (including PCB removal), the cycle will not be interrupted and normal communications will be maintained.

Alternatively, there are at least two connection paths between adjacent two-layer network structures. For example, the third node of the single layer network structure is connected to the fourth node of the two layer network structure, and the fifth node of the single layer network structure is connected to the sixth node of the two layer network structure, that is, there can be two connection paths between the single layer network structure and the two layer network structure.

Alternatively, the plurality of computing devices are divided into at least a single layer network structure, wherein the computing devices include:

The memory unit and at least one computing unit, said memory unit includes a first virtual control module circuit for communicating and controlling the memory unit, and said computing unit contains a second virtual control module circuit for communicating and controlling the computing unit; said first virtual control module circuit and said second virtual control module circuit are respectively connected to an external physical communication interface;

A single-board printed circuit board, said memory unit and said computing unit are located on said single-board printed circuit board;

Board frame, said single-board printed circuit board is located on the board frame;

Housing, said board frame is located in the housing;

Machine room, the mentioned building is located in the machine room;

The first data control center, said first data control center includes at least two computer rooms;

The second data control center, said second data control center includes at least two of the first data control centers.

From this, it can be seen that the computing device includes a memory unit and a computing unit, and both the memory unit and the computing unit are provided with a virtual control unit circuit for communication and control, and are connected to an external physical communication interface, so that the memory unit and the computing unit are parallel, and different computing units are parallel. From this, it can be seen that the computing device in the exemplary embodiment of the present invention no longer distinguishes between computing and networking, but integrates computing and storage into a network, and configures the computing unit to have a single interface for external communication, and computing and storage are performed in parallel. , which can improve the speed of data processing and, to a certain extent, meet the explosive demand for large amounts of data and large amounts of computing power caused by artificial intelligence and blockchain.

In addition, the plurality of said computing devices are divided into an eight-layer network structure, in which a single-layer network structure node S (silicon) is a module diagram of said computing unit, programmable internal circuit structure and control; the two-layer network structure node B (board) is said computing unit, that is, in a single-board PCB, between multiple chips, 10Gbps communication is established on one PCB, board-level control; the node of the three-layer network structure F (frame) is a single-board PCB, that is, in one frame, between several PCBs and between interconnecting boards in the PCB, 10Gbps communication is established, which is fully connected; the nodes of the four-level network structure C (cabinet) are boards and frames, in one cabinet, between the boards and frames, a 10Gbps fiber optic connection is installed, which is fully connected; nodes of the five-level network structure D (district) is a building, that is, they are divided into blocks in the data center and between several buildings. The partial on-demand fiber network in the computer room is 10 Gbps; nodes of the six-level network structure Z (zone) is a computer room, that is, they are divided into blocks in the data center and between several computer rooms. The partial on-demand fiber network in the computer room is 10 Gbps; nodes of the seven-layer network structure E (enterprise) is the first data control center mentioned, which is configured with broadband optical fiber on demand, data transmission over the Internet, and complete network management; nodes of the eight-layer network structure W (world) is the said second data control center, which is configured with broadband optical fiber on demand, data transmission over the Internet, and complete network management. It is clear that for a network topology consisting of several computing devices, the division of the network structure of each level is not limited to this.

Meanwhile, the aforementioned eight-layer network structure can be arranged in eight groups of 8 bits, for a total of 64 bits. In other words, the operating status of any device, or any part of a module's circuitry, or any part of a hard disk, can be independently queried and controlled across the entire network (worldwide). Obviously, this type of network topology design can meet the needs for a long period of time in the future. However, if one day in the future the network scale is not sufficient to meet the needs of the applications, the network topology can easily add additional layers of network structure on top of the data center (i.e. the W layer). If the number of independent entities that can be served by a superchip increases significantly in the future, resulting in insufficient network address allocation, it will be easy to expand below the chip level (i.e., the S level) and even localize to the cell level. From this, it can be seen that the scale of the network topology can be infinitely scaled to meet possible future needs.

In addition, the network topology can define three types of hardware resources and uniquely identify them using a uniform 64-bit address.

The first type, fine particle resources: refers to resources with complete functions inside the chip, which are labeled and located with the S (Silicon) address field.

The second type, Single Board Resources: Refers to the internal resources of the board with full functions, which are labeled and located in the B (Board) address field.

The third type, multi-card combined resources: usually refers to integrated multi-card resources assembled by independent boards, including combinations of various cabinets and machine rooms. At the same time, the resources are divided into 6 levels, which are indicated by the address fields F (Frame), C (Cabinet), D (District), Z (Zone), E (Enterprise) and W (World).

As shown in FIG. 1, the way to handle this task includes:

Stage 11: Obtaining the target data and the target algorithm required to complete the target task.

At the same time, said target data includes at least one data packet, for example, it may include data A and B. Said target algorithm includes at least one algorithm, for example, it may include algorithms C and D. In an exemplary embodiment of the present invention, the target data corresponds to the target algorithm, that is, the target data is the data required to execute the target algorithm. For example, the target task to be performed is to sequentially execute algorithms C and D on data A and B, then data A and B correspond to algorithms C and D.

Alternatively, before obtaining the target data and target algorithm required for said target task to be performed, it also includes:

According to the capability information of nodes in said network topology, the task to be performed is divided into at least one subtask, and one of at least said subtask is determined to be said target task.

It can be seen from this that, in the embodiment of the present invention, when a task to be performed involves a complex algorithm, it is necessary to first divide the task according to the capability information of each node in the network topology (i.e., how much computational power of the algorithm each node has) , so that multiple subproblems can be obtained, among which the subproblem algorithm is provided by the node.

For each subtask received by the partitioning route, there is a corresponding sequential execution sequence. For example, subtask 1, subtask 2, and subtask 3 are obtained by splitting path. The data and algorithms required for subtask 1 and subtask 2 do not collide with each other and can be executed in parallel. The data required for subtask 3 is the data generated by subtask 1 and subtask 2, where subtask 1 and subtask 2 can run in parallel. For subtask 1 and subtask 2, you can first get the nodes in the network topology that can provide data and algorithms, and then select one of the nodes that can provide data and one that can provide algorithms, and then the node that provides the algorithm processes the data , provided by the host that provides the data and then stores the processed data. In addition, when executing subtask 3, the nodes that execute the algorithms for subtask 1 and subtask 2 were refined. One can first identify a node in the network topology that can provide an algorithm corresponding to subtask 3, and then select a node that can provide an algorithm corresponding to subtask 3 3 and process the data required for subtask 3 stored in the nodes that execute the algorithms in subtask 1 and subtask 2 to realize the execution of the entire overall task.

In particular, for example, it is necessary to use the first model for the first 5 minutes of a 10-minute source video for processing, the second model for the next 5 minutes, and finally synthesize the video after processing the two-part model. The task can be divided into three subtasks, namely:

Subtask 1: perform processing of the first model for the first 5 minutes of the original video;

Subtask 2: perform processing of the second model for the last 5 minutes of the original video;

Subtask 3: Perform video synthesis.

For subtask 1: first get the nodes in the network topology that can provide the original video and the nodes that can provide the first model, to select the node that can provide the original video and the node that can provide the first model, and then broadcast the first 5 minutes of original video stream to the selected host that can provide the first model.

For subtask 2: since the node that can provide the original video was selected in subtask 1, during the processing of subtask 2, it is necessary to obtain a node in the network topology that can provide the second model, in order to select a node that can provide the second model, and then broadcast the last 5 minutes of the original video stream to the selected node, which can provide the second model.

For subtask 3: Since the data needed for subtask 3 is the data obtained after subtask 1 and subtask 2 are executed, during the processing of subtask 3, it is necessary to get nodes in the network topology that can perform video synthesis in order to select a node that can perform synthesis video. The data output by the node that provides the first model in subtask 1 and the data output by the node that provides the second model in subtask 2 synthesize video.

From this, it can be seen that the embodiment of the present invention can divide tasks into tasks with complex algorithms, so that the algorithms of each task are distributed over the network, rather than being executed by one node at a time, which can not only reduce the capacity requirements of one node, but also reduce the time task processing.

Step 12: obtaining at least one first type node capable of providing said target data in said network topology and at least one second type node capable of executing said target algorithm.

In the exemplary embodiment of the present invention, after receiving the target algorithm and target data necessary to perform the target task, the nodes in the network topology that can provide the target data and the nodes that can provide the target algorithm are obtained through the above step 12. From this, it can be seen that in the exemplary embodiment of the present invention, data and algorithms are distributed separately in the network, instead of storing data evenly for the node that needs to perform the target task, and the target algorithm is no longer executed by the node that needs to perform the target task itself, but is assigned to the node with processing power of the target algorithm to execute, that is, in the exemplary embodiment of the present invention, the combination of a distributed database and distributed computing can reduce the capabilities of a single node in a network topology, thereby reducing the cost of designing a single node and, moreover, the cost of designing an entire network topology.

Step 13: selecting a node that provides a set of said target data from at least said one first type node as the first target node, and a node that provides a set of said target algorithms from at least said one second type node to as the second target node.

In this case, the network topology includes several nodes, so there may be several nodes that can provide target data, and several nodes can execute target algorithms. Moreover, it is possible that the target data is distributed over several nodes, and the target algorithm is also distributed over several nodes. To complete the target, only nodes that provide a set of target data and nodes that provide a set of target algorithms are required.

Alternatively, the node that provides the set of said target data is selected from at least said one node of the first type as the first target node, and the node that provides the set of said target algorithms is selected from at least said one node of the second type as the second target node, including:

Receiving a combination formed by a node that provides a set of said target data, and a node that provides a set of said target algorithms;

According to the capability information of each combination, the selection of the target combination, and the node that provides the set of said target data in said target combination is defined as said first target node, and the node that provides the set of said target algorithms in said target combination is defined as second. target node;

At the same time, the information about the capabilities of the mentioned combination includes the time delay between the node that represents the set of the mentioned target data in the mentioned combination and the node that represents the set of the mentioned target algorithms, the length of time required by the node that provides the set of the mentioned target algorithms to execute the mentioned a target algorithm, and at least one of the cache space sizes of the nodes that provide the set of said target algorithm.

It should be noted, however, that if task division is not performed prior to step 11 above, a set of target algorithms may be provided by multiple nodes, then in this case:

First, if the node that represents the target data set also includes multiple nodes, the time delay between the node that represents the target data set and the node that represents the target algorithm set can be the sum or average of the delay between each node , which represents the target data, and each node, which represents the target algorithm. For example, the node that represents the target data set includes nodes a and b, and the node that represents the target algorithm set is node c and d, then the sum or average of the delays a-c, b-c, a-d, b-d can be used as delays between the group of nodes that represent the target data and the group of nodes that represent the target algorithm.

Secondly, the length of time required by the node that provides the set of said target algorithm to execute said target algorithm may be the sum or average of the length of time in the group required to execute the respective provided algorithms.

Thirdly, the size of the cache space of the nodes that provide the set of said target algorithm may be the sum or average of the cache space of the nodes in the group that provide the target algorithm.

If task division is performed prior to step 11, a set of target algorithms is provided by the node, then in this case:

First, if the node that represents the target data set also includes multiple nodes, the time delay between the node that represents the target data set and the node that represents the target algorithm set can be the sum or average of the delay between each node A that represents the target data, and a node that represents the target algorithm. For example, the node that represents the target data set includes nodes a and b, and the node that represents the target algorithm set is node c, then the sum or average of delays a-c, b-c can be used as the delay between a group of nodes that represent the target data, and nodes that represent the target algorithm.

Second, the length of time required by the node that provides the set of said target algorithm to execute said target algorithm is the length of time required by the node that provides the target algorithm to execute the target algorithm.

Thirdly, the size of the cache space of the nodes that provide the set of target algorithms may be the cache space of the nodes that provide the target algorithm.

In addition, for the above combination, for example, the target data packet includes two data A and B, which can be provided by nodes a and b in the network topology, and the target algorithm includes algorithm C, which can be provided by nodes c and d in network topology. The combination that can be obtained is as follows:

Combination 1: a, b, c;

Combination 2: a, b, d.

It is then possible to obtain the capability information of the above four combinations separately, and then select one combination according to the capability information of each of the four combinations, so that the nodes that represent the target algorithm in the selected combination use the target algorithm to process the target data provided by the nodes. , providing the target data in this combination.

In addition, said target combination is the most efficient combination of all combinations formed by nodes that provide a set of target data and nodes that provide a set of target algorithms. In this case, the possibility of combination can be expressed by the time delay between the node that represents the set of target data and the node that represents the set of target algorithms in combination, the length of time required by the node that provides the set of target algorithm to execute the target algorithms, the size of the cache space of the nodes , which provide a set of target algorithms. For example, the lower the latency between a node that represents a set of target data and a node that represents a set of target algorithms, the shorter the length of time it takes for a node that provides a set of target algorithms to execute the target algorithms, and the larger the cache space of nodes that provide set of target algorithms, the stronger the possibility of combination.

Alternatively, said target combination is selected based on information about the capabilities of each combination, including:

Selecting a combination corresponding to the minimum value of the time delay between the node that represents the set of said target data in said combination, and the node that represents the set of said target algorithms in said combination, as said target combination; or

Selecting a pattern corresponding to a minimum duration value required by a node that provides a set of said target algorithms to execute said target algorithms in said pattern as said target pattern; or

Selecting a combination corresponding to the maximum cache space value of the nodes that provide the set of said target algorithms in said combination as said target combination; or

Selection of the time delay between the node that represents the set of said target data and the node that represents the set of said target algorithms in said combination, and the combination corresponding to the minimum value from the sum of the duration required by the node that represents the set of said target algorithms to execute the mentioned target algorithms in combination, as mentioned target combination; or

Selecting a combination corresponding to the minimum value in the first ratio of said combination as said target combination, where said first ratio is the ratio of the time delay between the node that represents the set of said target data and the node that represents the set of said target algorithms in the mentioned combination, and the value of the size of the cache space of the nodes, which provide a set of mentioned target algorithm in combination; or

Selecting the combination corresponding to the minimum value in the second ratio of said combination as said target combination, where said second ratio is the ratio of the duration required by the node that provides the set of said target algorithms to execute the target algorithm in said combination and the cache space size value nodes that provide the set of said target algorithm in combination.

As can be seen from the above, in the exemplary embodiment of the present invention, any of the above six methods can be used to select the target combination. Here, it is understood that the specific selection method of the target combination is not limited to this, but also to other selection methods according to the combination capability information.

Step 14: managing said second target node uses said target algorithm to process said target data at said first target node.

Alternatively, managing said second target node uses said target algorithm to process said target data at said first target node, including:

Said first type node sends first indication information to said first target node, and said first indication information is used to indicate that said first target node sends said target data to said second target node;

Said first type node sends second indication information to said second target node, and said second indication information is used to indicate that said second target node uses said target algorithm to process said target data.

In an exemplary embodiment of the present invention, a node that needs to perform a target task, after selecting a node that provides a set of target data (i.e., the first target node) and a node that provides a set of target algorithms (i.e., the second target node) by of the aforementioned step 13, sends indication information to the first target node, so that the first target node sends the target data stored therein to the second target node, so that the second target node uses the target algorithm for processing, and then returns the data obtained after processing to the aforementioned node, which must complete the target.

It should be noted, however, that if task division is not performed prior to step 11 above, a set of target algorithms may be provided by multiple nodes, then in this case:

If the node that provides the target data set includes multiple nodes, the plurality of nodes that provide the target data set separately transmit corresponding data to the corresponding algorithm node. For example, nodes that provide a target data set include nodes a and b, and nodes that provide a target algorithm include nodes c and d. Data provided by node a must be processed by node c, and data provided by node b must be processed by node d. Then node a needs to send data to node c, and node b needs to send data to node d for processing.

That is, in the case where a target algorithm set is provided by multiple nodes and a target data set is provided by multiple nodes, it needs to be processed according to the correspondence between the data and the algorithm, as well as the execution order of each algorithm.

If task division is performed prior to step 11, a set of target algorithms is provided by the node, then in this case:

The nodes that provide the target data send the data provided by each of them to the node that provides the target algorithm, and the node that provides the target algorithm can process the received data.

Alternatively, when said target data includes a video stream or an audio stream, said second target control node uses said target algorithm to process said target data at said first target node, including:

The control of said second target node uses said target algorithm to process the first packet of said target data sent by the received said first target node.

At the same time, said first data packet includes a video stream or an audio stream of a given duration. Said video of a given duration may include a given number of video frames.

Alternatively, managing said second target node uses said target algorithm, and after processing the first packet of said target data sent by said first target node, also includes:

In the process of receiving said processed first data packet sent by said second target node, controlling said second target node uses said target algorithm to process said second target data packet sent by said first target node received;

At the same time, said second data packet includes a video stream or an audio stream of a given duration.

In the embodiment of the present invention, when processing a video stream or an audio stream, data packets can be processed one by one, and then the data packets being processed are sent, that is, in the embodiment of the present invention, through data-driven computing, no longer waiting for data, the collection must be completed, but in In the process of transmitting output data, calculations are performed step by step, which can improve efficiency and reduce latency.

Alternatively, the network topology includes at least a single layer network structure, with each layer of the network structure including a plurality of nodes connected in a predetermined manner. At the same time, the range of data transmission time delay values corresponding to different levels is different, and the data transmission time delay corresponding to the network structure of the first target layer includes the time delay of data transmission from the node in the mentioned network structure of the first target layer to the target node, and the mentioned the network structure of the first target level is any level of the network structure.

As shown in FIG. 3, before obtaining said network topology, at least one first type node capable of providing said target data and at least one second type node capable of executing said target algorithm also includes:

Step 21: The peer network structure addresses a second node capable of providing said target content in said peer network structure in accordance with the first corresponding target content parameters required by the first node.

At the same time, said first node is located in the structure of the first level network. The peer network structure addresses the second node capable of providing said target content in the peer network structure according to the first corresponding target content parameters required by the first node, that is, the peer network structure addresses the second node capable of providing said target content in the peer network structure. network structure according to the first relevant parameters. That is, addressing by the first matching parameter instead of addressing by the real address can increase a certain degree of security. In addition, an embodiment of the present invention performs content addressing within a layer, narrowing the content addressing scope and reducing certain overhead.

In this case, the second node may be a node of the first type or a node of the second type. When the peer network structure addresses the second node that can provide the target data according to the first corresponding parameters of the target data required by the first node, the second node is a first type node; when the peer network structure addresses the second node that can provide the target algorithm according to the first corresponding parameters of the target algorithm required by the first node, the second node is a node of the second type.

Step 22: When said peer network structure does not address said second node in said peer network structure, said peer network structure forwards said first corresponding parameters to the two-layer network structure.

In the exemplary embodiment of the present invention, when the second node that represents the target content does not perform addressing in the single layer network structure after passing step 21, the single layer network structure needs to forward the first corresponding parameters to the two layer network structure so that the two layer network structure can continue addressing to the second node. capable of providing targeted content according to the first relevant parameters.

Step 23: said single layer network structure receives the address of said second node sent by said two layer network structure.

Here, the address of said second node is obtained after said two-level network structure addresses said second node in said two-level network structure in accordance with said first corresponding parameters.

In addition, if the two-layer network structure still does not address the second node capable of providing the target content in the two-layer network structure, then the two-layer network structure continues forwarding the first relevant parameters to the other layer network structure until it addresses the second node, or until the entire addressing time to the second node reaches the specified time, and stops addressing the second node.

Step 24: The first node of said peer-to-peer network structure sends second corresponding target content parameters to said second node in accordance with the address of said second node.

In the exemplary embodiment of the present invention, when the two-layer network structure addresses the second node capable of providing target content in the two-layer network structure, the two-layer network structure sends the address of the second node to the first node of the one-layer network structure, so that the first node can send the second corresponding parameters of the target content to the second node according to the address of the second node.

Here, after the second node receives said second corresponding parameters, reports the target content to the first node according to the route from the second node to the first node.

As can be seen from the above, the task processing method provided by the embodiment of the present invention, in a single-layer network structure where the first node is located, in accordance with the first corresponding parameters of the target content required by the first node, addressing is performed to the second node, which can provide the target content, and when the second node is not addressed in the single layer network structure, the first corresponding parameters are directly forwarded to the two layer network structure to be addressed again in the two layer network structure until the second node is found, and the first node sends the second corresponding parameters of the target content to the second node according to address of the second node. From this, it can be seen that in the task processing method provided by the embodiment of the present invention, in the process of addressing to the second node, the content addressing is performed only within the layer, and the parameters on which the addressing is based are directly transferred between the layers, thereby reducing the scope of the content addressing, reducing the overhead, and due to the content addressing method, the security requirements are met to a certain extent.

Meanwhile, for each layer of the network structure in the above network topology, a control node may be installed, or a control node may not be installed. The control node stores the capability information of all nodes in the layer's network structure, that is, other nodes in the network structure other than the control node will periodically report their respective capability information to the layer's control node.

The task processing method in the exemplary embodiment establishes a control node at each layer of the network structure, and when the control node is not configured, the specific implementation methods are different.

In particular, for the situation where no control node is installed at each layer of the network structure, alternatively, said peer network structure addresses a second node capable of providing said target content in the peer network structure according to the first corresponding parameters of said target content. required by the first node.

Said first node of the peer-to-peer network structure translates said first corresponding parameters in said peer-to-peer network structure.

In the exemplary embodiment of the present invention, when the control node is not installed in the peer network structure, the first node can directly broadcast said first corresponding parameters in the peer network structure, and other nodes in the peer network structure can determine whether they can provide said target content to the first node after receiving the first corresponding parameters.

Alternatively, when said peer network structure does not address said second node in said peer network structure, said peer network structure forwards said first corresponding parameters to the bilayer network structure.

When the first node of said peer network structure does not receive the address of said second node within a predetermined period of time after the translation of said first corresponding parameters, sends indication information to the third node of said peer network structure.

The third node of said single layer network structure forwards said first corresponding parameters to the fourth node of said two layer network structure in accordance with the indication information.

At the same time, said third node is connected to said fourth node.

From this, it can be seen that in the exemplary embodiment of the present invention, the single-layer network structure and the two-layer network structure interact through the connection between the third node and the fourth node.

In addition, if the first node does not receive feedback from other nodes in the peer network structure within a given period of time after the translation of the first matching parameters, this means that there is no node in the peer network structure that can provide the target content, and the search must continue. at other levels. In this case, the first node can send indication information to the third node according to the communication channel between the first node and the third node, so that the third node forwards the first corresponding parameters to the fourth node of the two-layer network structure, so that the fourth node can continue addressing from the second node capable of providing target content to the first node in the two-layer network structure.

Alternatively, said single layer network structure receives the address of said second node sent by the two layer network structure.

The third node of said single layer network structure receives the address of said second node sent by the fourth node of said two layer network structure; wherein the address of said second node sent by said fourth node is the address sent by said second node after said fourth node broadcasts said first corresponding parameters in the two-layer network structure.

The first node of said peer network structure receives the address of said second node sent by the third node of said peer network structure, wherein the address of said second node sent by said third node is the address sent by said fourth node to said third node.

From this, it can be seen that in the exemplary embodiment of the present invention, after the fourth node receives the first matching parameters sent by the third node, it translates the first matching parameters into a two-layer network structure. After receiving the first relevant parameters, other nodes in the two-layer network structure other than the fourth node will determine whether they can provide the target content. If so, then through a direct connection from themselves to the fourth node, their address is passed back to the fourth node, so that the fourth node can be further redirected to the third node, and then the third node can be further redirected to the first node via a link from the third node to the first node.

Thus, when no control node is installed at each layer of the network topology, the process of addressing the second node capable of providing target content in the task processing method of the exemplary embodiment of the present invention can be illustrated as shown in FIG. 4. That is, the first node A broadcasts the first corresponding parameters in layer 1, and after that the first node A does not receive feedback from other nodes in layer 1 for a given time, then the first node A sends indication information to the third node C through the route to the third node C, after the third node C receives the indication information, it sends the first corresponding parameters to the fourth node D; the fourth node D broadcasts the first matching parameters at layer 2. Thereafter, the second node B receives the first matching parameters and determines that it can provide the target content. The second node B then passes its own address to the fourth node D via a route to the fourth node D; the fourth node D forwards the address of the second node B to the third node C; the third node C sends the address of the second node B to the first node A via the route with the first node A. Here, for ease of understanding, layers 1 and 2 in FIG. 4 only represent the first node to the fourth node.

Alternatively, the task handling method also includes:

Receipt by the first node of said peer-to-peer network structure of the address of said second node sent by said second node of said peer-to-peer network structure.

That is, when no control node is set in each layer of the network topology, after the first node broadcasts the first corresponding parameters in the peer network structure, if there is a second node in the peer network structure that can provide the target content, then this second node sends the address of the second node to the first node through a communication channel with the first node.

From this, it can be seen that when no control node is set at each layer of the network topology, the process of addressing the second node capable of providing target content for the first node in the task processing method of the exemplary embodiment of the present invention can also be illustrated as shown in fig. 5. That is, the first node A broadcasts the first matching parameters at layer 1. After the second node B receives the first matching parameters, determines that it can provide the target content, then the second node B transmits its own address to the first node A through the route to the first node A. However, for ease of understanding, level 1 in FIG. 5 represents only the first node and the second node.

In particular, for the situation where no control node is installed at each level of the network structure:

Alternatively, said peer network structure is provided with a first control node, and said first control node stores node capability information in said peer network structure.

Said peer network structure addresses a second node capable of providing target content in the peer network structure according to the first corresponding target content parameters required by the first node, including:

The first node of said peer-to-peer network structure sends said first corresponding parameters to the first control node of said peer-to-peer network structure;

The first control node of said peer-to-peer network structure performs addressing to said second node in accordance with said first respective parameters and node capability information in said peer-to-peer network structure.

In the exemplary embodiment of the present invention, when the peer network structure is provided with the first control node, since the first control node stores information about the capabilities of the nodes in the peer network structure, the first control node itself can determine which or which nodes in the peer network structure can provide the target content. Therefore, the first node only needs to send the first matching parameters to the first control node.

Alternatively, said two-layer network structure is provided with a second control node, and said second control node stores node capability information in said two-layer network structure;

When said peer network structure does not address said second node in said peer network structure, said peer network structure forwards said first corresponding parameters to said bilayer network structure, including:

When the first control node of said peer network structure determines that said second node does not exist in said peer network structure according to said first relevant parameters and capability information of nodes in said peer network structure, the first control node of said peer network structure sends the first corresponding parameters the second control node of the mentioned two-level network structure.

Wherein, the first control node decides according to the capability information of the nodes in the single-layer network structure, when there is no node in the single-layer network structure that can provide the target content, the first control node should send the first corresponding parameters to the second control node in the two-layer network structure, so that the second control node determines whether there is a node in the two-layer network structure that can provide the target content.

In addition, if the first control node and the second control node are not directly connected (that is, the first control node is not directly connected to the node of the two-layer network structure, or the second control node is not directly connected to the node of the single-layer network structure), for example, the third node of the single-layer network structure connected to the fourth node of the two-layer network structure, and the third node and the fourth node are not control nodes, in the process of sending the first corresponding parameters from the first control node to the second control node, the first control node must first transmit a communication channel with the third node to send the first corresponding parameters to the third node; finally, the fourth node sends the first corresponding parameters to the second control node via a communication channel with the second control node.

Alternatively, said single layer network structure receives the address of said second node sent by the two layer network structure.

The first control node of said single layer network structure receives the address of said second node sent by the second control node of said two layer network structure, wherein the address of said second node sent by said second control node is determined by said second control node in accordance with said first relevant parameters and information about node capabilities in said two-layer network structure.

The first node of said peer network structure receives the address of said second node sent by the first control node of said peer network structure.

It can be seen from this that, in the embodiment of the present invention, after the second control node receives the first relevant parameters, can decide which or which nodes in the two-layer network structure can provide the target content according to the capability information of the nodes in the two-layer network structure. , and when there is a node that can provide the target content, the address of this node is sent back to the first control node of the peer network structure, so that the first control node can be sent to the first node.

Thus, when a control node is provided in each layer of the network topology, the process of addressing the second node capable of providing target content in the task processing method of the exemplary embodiment of the present invention can be illustrated as shown in FIG. 6. That is, when the first node A sends the first corresponding parameters to the first control node E, and the first control node E decides that there is no node in this layer that can provide target content according to the node capability information in layer 1, then the first control node E sends the first corresponding parameters to the third node C via a route with the third node C; the third node C forwards the first matching parameters to the fourth node D; the fourth node D then forwards the first corresponding parameters to the second control node F of the layer 2. If the second control node F decides that there is a node in this layer that can provide the target content, i.e. the second node B, the second control node F sends the address of the second node B to the fourth node via the route with the fourth node D; the fourth node D forwards the address of the second node B to the third node C; the third node C then sends the address of the second node B to the first control node E via the route with the first control node E, so that the first control node E can send the address of the second node B to the first node A via the route with the first node A. In this case, for ease of understanding, the layers 1 and 2 in FIG. 6 only represent the first node to the fourth node, as well as the first control node and the second control node.

As can be seen from the above, when the control node is configured at each layer of the network topology, the first relevant parameters of the target content are broadcast through the control node, which can provide the address of the target content node, and feedback through the control node, the node that needs the target content may not need to know the node that represents the target content, and the node that represents the target content may not need to know for which node it is being provided, which can further increase the security of the process.

Alternatively, the task handling method also includes:

Receiving by the first node of said peer network structure the address of said second node of said peer network structure sent by the first control node of said peer network structure;

Wherein, the address of said second node sent by said first control node is determined by said first control node in accordance with the first corresponding parameters and node capability information in said single-layer network structure.

That is, when the control node is provided at each layer of the network topology, the first node sends the first corresponding parameters to the first control node of the peer network structure, and the first control node determines whether there is a second node in the peer network structure that can provide target content according to the first relevant parameters and information about the capabilities of the node in a single-layer network structure. If there is one, the first control node sends the address of the second node to the first node.

From this, it can be seen that when a control node is provided at each layer of the network topology, the process of addressing the second node capable of providing target content in the task processing method of the exemplary embodiment of the present invention can also be illustrated as shown in FIG. 7. That is, the first node A sends the first corresponding parameters to the first control node E in layer 1, and the first control node E determines that there is a second node B in layer 1 that can provide the target content, then the first control node E sends back the address the second node B to the first node A via the route to the first node A. However, for ease of understanding, layers 1 and 2 in FIG. 7 represent only the first node, the second node and the first control node.

Thus, in the exemplary embodiment of the present invention, content addressing is used within a layer, and parameters on which content addressing is based are directly transferred between layers, thereby establishing a communication connection. In this case, after the channel assembly is completed, the business content is separated according to the addressing and processing functions, and the data is marked and sent. The transit node only performs split processing and reorganizes the processed packet, and the rest of the content is transferred in full. The content is processed during the transfer, and the data is no longer transferred after each processing, but the knowledge generated by the processed data is transferred.

Thus, for the task processing method of the exemplary embodiment of the present invention, specific applications can be illustrated as follows.

For example, as shown in FIG. 8, the network topology includes a three-layer network structure in which layer 1 includes nodes 101-106, layer 2 includes nodes 201-202, and layer 3 includes node 301. Here, node 101 stores video data in in real time, node 102 stores AR materials and node 103 stores 3D scan data.

When node 301 needs real video taken at node 101 as background, AR at node 102 as interest, and scanned data at node 103 as video content, node 301 issues three requests at the beginning of the operation, which can be broadcast or sent directly to the point to a point. The request includes: operation type, resource request, data type and content description.

Node 101, node 102, and node 103 respond to the request accordingly by providing a content feedback sample. After node 301 receives feedback from node 101, node 102, and node 103, the task processing method of the exemplary embodiment of the present invention is used to plan a processing route, that is, to determine a node that provides data and a node that provides content. For example, the nodes that provide data are node 101, node 102, and node 103, and the node that provides the algorithm is node 201. Node 301 will send the required background video generation code to node 101, the AR action, and the time point to node 102, 3D extraction of time information and required content to node 103, and video synthesis code to node 201.

Thereafter, node 101 uses the tool issued by 301 to pre-process the desired video, create background video, and send it to node 201 after encoding and compression; node 102 provides the AR material to node 201 in accordance with the desired AR action and time point; the node 103 provides 3D content to the node 201 according to the desired 3D content and time point; node 201 receives the respective data provided by node 101, node 102, and video node 103 for video synthesis, transcodes it, synthesizes new video, and sends it to node 301.

Thus, in the exemplary embodiment of the present invention, data and algorithms are distributed separately over the network, instead of storing data uniformly for the node that needs to perform the target task, and the target algorithm is no longer executed by the node that needs to perform the target task itself, but is assigned to a node with the processing power of the target algorithm to execute, that is, an embodiment of the present invention combining a distributed database and distributed computing can not only improve data security, but also reduce the capabilities of a single node in a network topology, thereby reducing the cost of designing a single node, and moreover, reducing the cost of designing the entire network topology.

In addition, the task processing method of the exemplary embodiment of the present invention can also be illustrated in the fields of practical application, as described below.

For example, as shown in FIG. 8, the network topology of a large company includes a three-layer network structure. Meanwhile, level 1 includes business lounge nodes 101-106, level 2 includes regional branch nodes 201-202, and level 3 includes provincial company nodes 301. At the same time, each node of the business lounge is connected to various types of device, including facial recognition cameras, people control cameras, a smoke sensor, and an air quality sensor. These devices collect data separately and store it in the corresponding node of the business lounge.

In this case, when the provincial company node needs real-time data, for example, the provincial company node needs to accurately find someone, the provincial company node addresses the layer 2 node, which can provide content related to the task, for example, the provincial company node can send a batch of tasks to the regional branch node at layer 2 (or broadcast about the given task batch), then the regional branch node corresponding to the task will determine if it is its own task, the regional branch node corresponding to the task sends the task batch to the node connected to it of the business lounge, so that the business lounge node receiving the task package parses the data and obtains the operation type as a software update, the data type as a code, and the specific device number, and then executes the task based on the received information and gives feedback after task completion.

In particular, the task package sent by the node to the provincial company includes: real-time reports, data type video, any type of device, and the content is a photo and a tracking number. After receiving the task package, the lounge host will perform the following process.

The lounge node receives the network list file, which is loaded through the jtag (Joint Testing Working Group) interface. When the download is completed, the lounge node download module generates an acknowledgment message and sends it to the lounge node communication module so that the communication module fills in the content in the 1 kB frame format and sends feedback. In this case, after the feedback, the ARM of the lounge node sends instructions to the FPGA of the lounge node, thereby opening the communication module of the FPGA, so that the FPGA directly receives the IP data of the camera connected to the lounge node (among them, the camera connected to the lounge node). the lounge host operates in UDP (User Datagram Protocol) mode). Here, after the lounge node FPGA communication module receives data, decoding through the FPGA video decoding module is completed, and the loaded AI module is responsible for output. The AI module sends back a discovery message, starts the forwarding mode, records the time point, and forwards the last video within 1 minute after DDR (double data rate) to the regional branch office via the communication module. At the same time, regional branches can also conduct further analysis based on real-time reports received from various locations.

At the same time, the FPGA supports the file system through ARM, and the video data is written directly to the hard disk using the FPGA. The FPGA and ARM communicate through a PCIE (PCI Express Peripheral Serial Bus) port. In addition, the business lounge node includes multiple FPGAs, and the FPGA here is only responsible for this scenario task, and other daily monitoring tasks are performed by other FPGAs in parallel.

In addition, the task processing method may also include: periodically reporting the target content by the second node of the two-layer network structure to the first node of the single-layer network structure.

For example, the network topology of a large company shown in FIG. 8 includes a three-layer network structure. When the lounge node periodically reports device data, the specific execution process can be described from the first to fourth steps as follows:

First step: Lounge node daily monitoring data, operation type, device number and data type are stored here. For example, you can store videos related to VIP clients according to VIP business statistics, first face camera, and original video files; or according to VIP business statistics, first face camera, store time, store VIP store time tracking statistics; or according to the statistics of the sellers, the first face camera, the time in the store, store the statistics of the seller's time in the store.

Stage 2: The business lounge node automatically groups packages at the specified time according to the pre-set reporting content (which is mainly statistical content), and the package length is 1 KB. At the same time, the existing network can still be used for data transmission, the address layer is only for content identification, the business lounge node has 10 gateways responsible for encapsulating TCP (Transmission Control Protocol, transmission control protocol) / IP and docking with the existing network, this is the standard TCP/IP protocol, which will not be repeated here.

Step 3: Regional branch nodes, according to the content reported by each lounge node, perform local statistics, generate reports, and use messages (see Table 1 for message format) to continue reporting. At the same time, there is a logical ring between the nodes of regional branches, and each branch has an output for transmitting data to the upper and lower levels.

Step 4: Provincial company node statistics, such as average VIP customer visit time, employee hours, and traffic per store.

Table 1. Message format 8-bit identification country 8-bit identification province 16-bit identification area 16-bit business identification
hall 16-bit identification device 8-bit identification type 16-bit identification data type 8-bit identification type Spare operations devices Data + error correction + validation

Alternatively, the first node of said peer-to-peer network structure sends second corresponding parameters of said target content to said second node according to the address of said second node, including:

Said first node sends said second corresponding parameters to said third node via a route from said first node to said third node in said peer network structure;

Wherein, after said third node receives said second corresponding parameters, said second corresponding parameters are forwarded to the fourth node of said two-layer network structure, so that said fourth node transmits said second corresponding parameters to said second node via a route from said fourth node to said second node. node in said two-layer network structure.

Alternatively, the first node of said peer-to-peer network structure sends second corresponding parameters of said target content to said second node in accordance with the address of said second node.

Said first node sends said second corresponding parameters to said first node via a route to said first control node;

Wherein, after said first control node receives said second corresponding parameters, said second corresponding parameters are sent to the second control node of said two-level network structure, so that said second control node transmits said second corresponding parameters to said second node through the route to said second node.

Alternatively, the first node of said peer-to-peer network structure sends second corresponding parameters of said target content to said second node according to the address of said second node, including:

Said first node sends said second corresponding parameters to said second node via a route from said first node to said second node in said peer network structure.

Alternatively, the first node of said peer-to-peer network structure sends second corresponding parameters of said target content to said second node according to the address of said second node, including:

Said first node sends said second corresponding parameters to said first node via a route to said first control node;

Wherein, after said first control node receives said second corresponding parameters, sends said second corresponding parameters to said second node via a route to said second node.

In this case, the network topology determines the route macroscopically, taking into account any two points (beginning and end). Therefore, as can be seen from the above, when the first node (ie, the start point) sends data to the second node (ie, the end point), it is first determined whether the source and end points are in the same network layer. If they are at the same level, forwarding at the same level is performed. If they are not on the same level, it is sent to the upper or lower levels through a vertical jumper between levels.

Alternatively, the node address of said network topology includes fields corresponding to each layer. For example, when said network topology includes a three-layer network structure, it can be divided into high-level, middle-level, and low-level. When sending a data packet from the first node to the second node, the specific process can be described as follows

If all three of the two addresses are equal, this means that the data packet reaches its destination, and the switch of the second node receives the data packet. And indicates the completion of the exchange function. It then performs subsequent possible storage or computation operations. If both are not equal, further analysis is required: if the high-level addresses of the two are not equal, they are forwarded to the higher-level network structure via a bridge. In addition, if the given node is not directly connected to the high layer, the data packet is forwarded along the layer until the connection node of the high-level network structure is found; if the destination address of the data packet is equal to the address of the high-level local port, but the address of this level does not match, it is sent in the direction of increasing or decreasing the address of this level according to the random label of the address of this level until a local match is found; if the addresses of the upper level and this level of the data packet are equal to the local port, but the address of the lower level does not match, it is sent to the lower level through the jumper; if the given node is not directly connected to the lower layer, it is forwarded along the layer until the node of the connection of the lower layer is found. An embodiment of the present invention also provides a task processor that is applied to a network topology, said network topology includes a plurality of nodes; as shown in FIG. 9, task processor 400 includes:

The first obtaining module 401 used to obtain the target data and the target algorithm necessary to perform the target task;

A second acquisition module 402 used to obtain at least one first type node capable of providing said target data in said network topology and at least one second type node capable of executing said target algorithm.

Selection module 403 used to select a node that provides a set of said target data from at least said one node of the first type as the first target node, and select a node that provides a set of said target algorithm from at least said one node of the second type as the second target node;

A control module 404 used to control said second target node to process said target data at said first target node using said target algorithm.

Alternatively, the selection module 403 includes:

The combination submodule used to obtain the combination formed by the node that provides the set of said target data and the node that provides the set of said target algorithm;

A selection submodule used to select a target combination according to the capability information of each combination and determine a node in said target combination that provides the set of said target data as said first target node, and a node in said target combination that provides the set of said target algorithms as the second target node;

At the same time, the information about the capabilities of the mentioned combination includes the time delay between the node that represents the set of the mentioned target data in the mentioned combination and the node that represents the set of the mentioned target algorithms, the length of time required by the node that provides the set of the mentioned target algorithms to execute the mentioned a target algorithm, and at least one of the cache space sizes of the nodes that provide the set of said target algorithm.

Alternatively, the select submodule is specifically designed to:

selecting a combination corresponding to a minimum time delay value between a node that represents a set of said target data in said combination and a node that represents a set of said target algorithms in said combination as said target combination;

or

selecting a pattern corresponding to a minimum duration value required by the node that provides the set of said target algorithms to execute said target algorithms in said pattern as said target pattern; or

choosing a combination corresponding to the maximum value of the cache

node spaces that provide a set of said target algorithms in said combination, as said target combination; or

selecting a time delay between a node that represents a set of said target data and a node that represents a set of said target algorithms in said combination, and a combination corresponding to the minimum value from the sum of the duration required by the node that represents a set of said target algorithms to execute said target algorithms in combination, as mentioned target combination; or

selecting a combination corresponding to the minimum value in the first ratio of said combination as said target combination, where said first ratio is the time delay ratio between a node that represents a set of said target data and a node that represents a set of said target algorithms in said combination, and the value of the size of the cache space of the nodes, which provide a set of mentioned target algorithm in combination; or

selecting a combination corresponding to the minimum value in the second ratio of said combination as said target combination, where said second ratio is the ratio of the duration required by the node that provides the set of said target algorithms to execute the target algorithm in said combination, and the value of the cache space size nodes that provide the set of said target algorithm in combination.

Alternatively, said task processor also includes:

A partitioning module used to divide a task to be performed into at least one subtask according to information about the capabilities of nodes in said network topology and determine from at least one subtask as said target task.

Alternatively, when said target data includes a video stream or an audio stream, said control module includes:

A first processing submodule used to control said second target node to use said target algorithm to process the first packet of said target data sent by the received said first target node;

At the same time, said first data packet includes a video stream or an audio stream of a given duration.

Alternatively, said task processor also includes:

A processing module used to control said second target node to use said target algorithm to process the second packet of said target data sent by the received said first target node in the process of receiving the processed said first data packet sent by said second target node.

At the same time, said second data packet includes a video stream or an audio stream of a given duration.

Alternatively, said control module includes

myself:

a first transmission submodule used to send the first indication information to said first target node, and said first indication information is used to indicate that said first target node is sending said target data to said second target node; a second transmission submodule used to send second indication information to said second target node, and said second indication information is used to indicate that said second target node uses said target algorithm to process said target data; a receiving submodule used to receive data sent by said second target node after processing said target data using said target algorithm.

As can be seen from the above, the task processor provided by the exemplary embodiment of the present invention obtains the target data and the target algorithm necessary to perform the target task, and determines at least one first type node in the network topology that can provide the target data and, at least one node of the second type that can execute the target algorithm, thereby selecting the first type node and the second type node, and controlling the selected second type node to use the target algorithm to process the target data provided by the selected first type node to achieve the execution of the target tasks. It can be seen from this that in the exemplary embodiment of the present invention, data and algorithms are distributed separately over the network, instead of storing data evenly for the node that needs to perform the target task, and the target algorithm is no longer executed by the node that needs to perform the target task on its own, a is assigned to a node with the processing power of the target algorithm to execute, that is, an embodiment of the present invention combining distributed database and distributed computing, which can reduce the capabilities of one node in the network topology, thereby reducing the cost of designing one node, and moreover, reducing costs for the design of the entire network topology.

An embodiment of the present invention also provides a task processing apparatus that is applied to a network topology, said network topology includes: at least a single layer network structure, each layer of the network structure includes a plurality of nodes connected in a predetermined manner; as shown in FIG. 10, this task processor 800 includes:

The first control unit 801 used to manage the peer network structure to address the second node capable of providing said target content in said peer network structure, in accordance with the first corresponding parameters of the target content required by the first node; wherein said first node is located in said peer network structure, and said second node is said first type node or said second type node;

A second control module 802 used to control said sibling network structure to forward said first relevant parameters to the sibling network structure when said sibling network structure does not address said second node in said sibling network structure;

A third control module 803 used to control the single layer network structure to receive the address of said second node sent by the two layer network structure, wherein the address of said second node is obtained after said two layer network structure addresses said second node in said two layer network structure in in accordance with said first relevant parameters;

The first transmission module 804 used to control the first node of said peer network structure to send the second corresponding parameters of said target content to said second node in accordance with the address of said second node.

Alternatively, the first 801 control module is specifically designed to:

The first node control of said peer network structure translates said first relevant parameters in said peer network structure.

Alternatively, the second control module 802 is specifically designed to:

sending indication information to a third node of said peer network structures when control of the first node of said peer network structure does not receive the address of said second node within a predetermined period of time after broadcasting said first corresponding parameters; controlling a third node of said single-layer network structure forwards said first corresponding parameters to a fourth node of said two-layer network structure in accordance with the indication information; wherein said third node is connected to said fourth node.

Alternatively, the third control module 803 is specifically designed for:

controlling a third node of said single layer network structure to receive an address of said second node sent by a fourth node of said two layer network structure; wherein the address of said second node sent by said fourth node is the address sent by said second node after said fourth node broadcasts said first corresponding parameters in a two-layer network structure; controlling a first node of said peer network structure to receive an address of said second node sent by a third node of said peer network structure; wherein the address of said second node sent by said third node is sent by said fourth node to said third node.

Alternatively, task processor 800 also includes:

A second transmission module used to control the first node of said peer network structure to receive the address of said second node sent by said second node of said peer network structure.

Alternatively, said peer-to-peer network structure is provided with a first control node, and said first control node stores node capability information in said peer-to-peer network structure; said first control module 801 is specifically designed for:

controlling the first node of said peer-to-peer network structure to send said first corresponding parameters to the first control node of said peer-to-peer network structure; controlling a first control node of said peer network structure to address said second node in accordance with said first respective parameters and capability information of nodes in said peer network structure;

Alternatively, said two-layer network structure is provided with a second control node, and said second control node stores node capability information in said two-layer network structure; said second control module 802 is specifically designed to:

forwarding said first corresponding parameters to a two-layer network structure by said one-layer network structure, when the management of said one-layer network structure does not address said second node in said one-layer network structure, including:

controlling the first control node of said peer network structure according to said first relevant parameters and capability information of nodes in said peer network structure, determining that when said second node does not exist in said peer network structure, the first control node of said peer network structure sends said first corresponding parameters to the second control node of the two-layer network structure.

Alternatively, the third control module 803 is specifically designed for:

controlling the first control node of said single layer network structure to receive the address of said second node sent by the second control node of said two layer network structure, wherein said second node address sent by said second control node is determined by the second control node in accordance with said first relevant parameters and information about the capabilities of the node in the mentioned two-level network structure; controlling the first node of said peer network structure to receive the address of said second node sent by the first control node of said peer network structure.

Alternatively, task processor 800 also includes:

A third transmission module used to control the first node of said peer network structure to receive the address of said second node of said peer network structure sent by the first control node of said peer network structure;

Wherein, the address of said second node sent by said first control node is determined by said first control node in accordance with the first corresponding parameters and node capability information in said single-layer network structure.

Alternatively, the first transmission module 804 is specifically configured to:

controlling said first node to send said second corresponding parameters to said third node via a route from said first node to said third node in said peer network structure; wherein, after said third node receives said second respective parameters, said second respective parameters are forwarded to the fourth node of said two-layer network structure, so that said fourth node transmits said second respective parameters to said second node via a route from said fourth node to said second node in said two-layer network structure.

Alternatively, the first transmission module 804 is specifically configured to:

controlling said first node to send said second corresponding parameters to said first node via a route to said first control node; wherein, after said first control node receives said second respective parameters, said second respective parameters are forwarded to the second control node of said two-layer network structure, so that said second control node transmits said second respective parameters to said second node via a route to said second node .

Alternatively, the first transmission module 804 is specifically configured to:

controlling said first node to send said second corresponding parameters to said second node via a route from said first node to said second node in said peer network structure.

Alternatively, the first transmission module 804 is specifically configured to:

controlling said first node to send said second corresponding parameters to said first node via a route to said first control node; wherein after said first control node receives said second respective parameters, sends said second respective parameters to said second node via a route to said second node.

As can be seen from the above, the task processor provided by the embodiment of the present invention, in a single-layer network structure where the first node is located, in accordance with the first corresponding parameters of the target content required by the first node, the content is addressed to the second node, which can provide the target content. content, and when the second node is not addressed in the single-layer network structure, the first corresponding parameters are directly forwarded to the two-layer network structure to be addressed again in the two-layer network structure, until the second node is found, the first node sends the second corresponding parameters of the target content to the second node according to the address of the second node. From this, it can be seen that in the task processing method provided by the embodiment of the present invention, in the process of addressing to the second node, the content addressing is performed only within the layer, and the parameters on which the addressing is based are directly transferred between the layers, thereby reducing the scope of the content addressing, reducing the overhead, and due to the content addressing method, the security requirements are met to a certain extent.

An embodiment of the present invention also provides the electronic device shown in FIG. 11, which includes a processor 501, a communication interface 502, a memory 503, and a communication bus 504, wherein the processor 501, the communication interface 502, and the memory 503 complete communication with each other via the communication bus 504;

Memory 503 for storing program codes;

Processor 501 used to execute a program stored

in memory 503, to perform steps in said task processing method described below; the communication interface 502 is used for communication between the terminal and other devices.

Said task processing method is applied to a network topology, said network topology includes a plurality of nodes; said task processing method includes:

obtaining target data and target algorithm required for

fulfillment of the target task; obtaining at least one first type node capable of providing said target data in said network topology, and at least one second type node capable of executing said target algorithm.

Selection of a node that represents a set of said target data as the first target node from at least one node of the first type, and a node that represents a set of said target algorithms as the second target node from at least one mentioned node of the second type;

Controlling said second target node to use said target algorithm to process said target data at said first target node.

The node that provides the set of said target data is selected from at least said one first type node as the first target node, and the node that provides the set of said target algorithms is selected from at least said one second type node in as the second target node, including:

Receiving a combination formed by a node that provides a set of said target data, and a node that provides a set of said target algorithms;

According to the capability information of each combination, the selection of the target combination, and the node that provides the set of said target data in said target combination is defined as said first target node, and the node that provides the set of said target algorithms in said target combination is defined as second. target node;

At the same time, the information about the capabilities of the mentioned combination includes the time delay between the node that represents the set of the mentioned target data in the mentioned combination and the node that represents the set of the mentioned target algorithms, the length of time required by the node that provides the set of the mentioned target algorithms to execute the mentioned a target algorithm, and at least one of the cache space sizes of the nodes that provide the set of said target algorithm.

Alternatively, said target combination is selected based on information about the capabilities of each combination, including:

selecting a combination corresponding to a minimum time delay value between a node that represents the set of said target data in said combination and a node that represents the set of said target algorithms in said combination as said target combination; or

selecting a pattern corresponding to a minimum duration value required by a node that provides a set of said target algorithms to execute said target algorithms in said pattern as said target pattern; or

selecting a combination corresponding to a maximum cache space value of the nodes that provide the set of said target algorithms in said combination as said target combination; or

selection of the time delay between the node that represents the set of said target data and the node that represents the set of said target algorithms in said combination, and the combination corresponding to the minimum value from the sum of the duration required by the node that represents the set of said target algorithms to execute the mentioned target algorithms in combination, as mentioned target combination; or

selecting a combination corresponding to the minimum value in the first ratio of said combination as said target combination, where said first ratio is the time delay ratio between a node that represents a set of said target data and a node that represents a set of said target algorithms in said combination, and the value of the size of the cache space of the nodes, which provide a set of mentioned target algorithm in combination; or

selecting a combination corresponding to the minimum value in the second ratio of said combination as said target combination, where said second ratio is the ratio of the duration required by the node that provides the set of said target algorithms to execute the target algorithm in said combination and the value of the cache space size nodes that provide the set of said target algorithm in combination.

Alternatively, before obtaining the target data and target algorithm required for said target task to be performed also includes:

According to the capability information of nodes in said network topology, the task to be performed is divided into at least one subtask, and one of at least said subtask is determined to be said target task.

Alternatively, when said target data includes a video or audio stream, said second target control node uses said target algorithm to process said target data at said first target node, including: controlling said second target node uses said target algorithm to process the first packet said target data sent by the received said first target node; wherein said first data packet includes a video stream

or an audio stream of a given duration.

Alternatively, managing said second target node uses said target algorithm, and after processing the first packet of said target data sent by said first target node, also includes:

Controlling the process of receiving said processed first data packet sent by said second target node, said second target node using said target algorithm to process the second packet of said target data sent by said first target node;

At the same time, said second data packet includes a video stream or an audio stream of a given duration.

Alternatively, managing said second target node uses said target algorithm to process said target data at said first target node, including:

sending first indication information to said first target node, said first indication information being used to direct said first target node to send said target data to said second target node; sending second indication information to said second target node, said second indication information being used to direct said second target node to use said target algorithm to process said target data; receiving data sent by said second target node obtained after processing said target data using said target algorithm.

Alternatively, said network topology includes: at least a single layer network structure, each layer of the network structure including a plurality of nodes connected in a predetermined manner; said task processing method also includes the steps of:

The peer network structure addresses a second node capable of providing said target content in said peer network structure in accordance with first corresponding parameters of said target content required by the first node; wherein said first node is located in said single layer network structure, said second node is said first type node or said second type node; when said peer network structure does not address said second node in said peer network structure, said peer network structure forwards said first respective parameters to the bilayer network structure; said single layer network structure receives an address of said second node sent by said two layer network structure, wherein the address of said second node is received by said two layer network structure after addressing said second node in said two layer network structure in accordance with said first respective parameters; said first node of the peer network structure sends second corresponding parameters of said target content to said second node in accordance with the address of said second node.

Alternatively, said peer network structure addresses a second node capable of providing said target content in the peer network structure according to first corresponding parameters of said target content required by the first node, including:

Said first node of the peer-to-peer network structure translates said first corresponding parameters in said peer-to-peer network structure.

Alternatively, when said peer network structure does not address said second node in said peer network structure, said peer network structure forwards said first corresponding parameters to the bilayer network structure, including

When the first node of said peer network structure does not obtain the address of said second node within a predetermined period of time after the translation of said first corresponding parameters, sends indication information to the third node of said peer network structure;

The third node of said single layer network structure forwards said first corresponding parameters to the fourth node of said two layer network structure in accordance with the indication information;

At the same time, said third node is connected to said fourth node.

Alternatively, said single layer network structure receives the address of said second node sent by the two layer network structure, including steps:

a third node of said single layer network structure receives an address of said second node sent by a fourth node of said two layer network structure; wherein the address of said second node sent by said fourth node is the address sent by said second node after said fourth node broadcasts said first corresponding parameters in a two-layer network structure; a first node of said peer network structure receives an address of said second node sent by a third node of said peer network structure; wherein the address of said second node sent by said third node is the address sent by said fourth node to said third node.

Alternatively, the task processing method also includes the step of:

The first node of said peer network structure receives the address of said second node sent by said second node of said peer network structure.

Alternatively, said peer network structure is provided with a first control node, and said first control node stores node capability information in said peer network structure.

Said peer network structure addresses a second node capable of providing target content in the peer network structure according to the first corresponding target content parameters required by the first node, including:

The first node of said peer-to-peer network structure sends said first corresponding parameters to the first control node of said peer-to-peer network structure;

The first control node of said peer-to-peer network structure performs addressing to said second node in accordance with said first respective parameters and node capability information in said peer-to-peer network structure.

Alternatively, said two-layer network structure is provided with a second control node, and said second control node stores node capability information in said two-layer network structure;

When said peer network structure does not address said second node in said peer network structure, said peer network structure forwards said first corresponding parameters to said bilayer network structure, including steps:

When the first control node of said peer network structure determines that said second node does not exist in said peer network structure according to said first relevant parameters and capability information of nodes in said peer network structure, the first control node of said peer network structure sends the first corresponding parameters the second control node of the mentioned two-level network structure.

Alternatively, said single layer network structure receives the address of said second node sent by the two layer network structure, including steps:

The first control node of said single layer network structure receives the address of said second node sent by the second control node of said two layer network structure, wherein the address of said second node sent by said second control node is determined by said second control node in accordance with said first relevant parameters and information about node capabilities in said two-layer network structure;

The first node of said peer network structure receives the address of said second node sent by the first control node of said peer network structure.

Alternatively, the task processing method also includes the steps of:

The first node of said peer network structure receives the address of said second node of said peer network structure sent by the first control node of said peer network structure;

Wherein, the address of said second node sent by said first control node is determined by said first control node in accordance with the first corresponding parameters and node capability information in said single-layer network structure.

Alternatively, the first node of said peer-to-peer network structure sends second corresponding parameters of said target content to said second node according to the address of said second node, including:

Said first node sends said second corresponding parameters to said third node via a route from said first node to said third node in said peer network structure;

Wherein, after said third node receives said second corresponding parameters, said second corresponding parameters are forwarded to the fourth node of said two-layer network structure, so that said fourth node transmits said second corresponding parameters to said second node via a route from said fourth node to said second node. node in said two-layer network structure.

Alternatively, the first node of said peer network structure sends the second corresponding parameters of said target content to said second node in accordance with the address of said second node, including the steps of:

Said first node sends said second corresponding parameters to said first node via a route to said first control node;

Wherein, after said first control node receives said second corresponding parameters, said second corresponding parameters are sent to the second control node of said two-level network structure, so that said second control node transmits said second corresponding parameters to said second node through the route to said second node.

Alternatively, the first node of said peer network structure sends the second corresponding parameters of said target content to said second node in accordance with the address of said second node, including the steps of:

Said first node sends said second corresponding parameters to said second node via a route from said first node to said second node in said peer network structure.

Alternatively, the first node of said peer network structure sends the second corresponding parameters of said target content to said second node in accordance with the address of said second node, including the steps of:

Said first node sends said second corresponding parameters to said first node via a route to said first control node;

Wherein, after said first control node receives said second corresponding parameters, sends said second corresponding parameters to said second node via a route to said second node.

The communication bus referred to in the aforementioned terminal may be a Peripheral Component Interconnection (called - PCI) bus or an Extended Industry Standard Architecture (called -EISA) bus. The communication bus can be divided into address bus, data bus, control bus, etc. For ease of presentation, only one solid line is used in the figure, but this does not mean that there is only one tire or one type of tire.

The memory may include random access memory (Random Access Memory, referred to as - RAM), may also include non-volatile memory (non-volatile memory), such as at least one magnetic disk drive. Alternatively, the memory may also be at least one storage device located away from the aforementioned processor. The memory has a place to store the program code 1031 for performing any of the method steps in the above method. For example, the program code storage space may include each program code 1031 used to implement various steps in the above method, respectively. These program codes may be read from, or written to, one or more computer program products. These computer program products include program code media such as hard disks, compact discs (CDs), memory cards, or floppy disks. Such computer program products are typically portable or fixed storage units, as described with reference to FIG. 12. This memory unit may have a storage segment, storage space, etc., arranged similarly to memory 503 in the electronic device of FIG. 11. The program code may, for example, be compressed in an appropriate form. Typically, the memory unit includes a computer-readable code, that is, a code that can be read, for example, by a processor such as 501, which, when triggered by an electronic device, causes the electronic device to perform various steps in the manner described above.

The above processors may be general purpose processors, including central processing units (Central Processing Unit, called - CPU), network processors (Network Processor, called - NP), etc.; They can also be Digital Signal Processing (DSPs), Field-Programmable Gate Arrays (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

In another exemplary embodiment provided by the present invention, a computer-readable storage medium is also provided. This computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform any of the task processing methods described in the above embodiment.

In the above exemplary embodiment, this may be achieved in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it may be implemented in whole or in part as a computer program product. Said computer program product includes one or more computer instructions. When the instructions of said computer program are loaded and executed on the computer, the process or function described in accordance with the exemplary embodiment of the present invention is generated in whole or in part. Said computer may be a general purpose computer, a dedicated computer, a computer network, or other programmable device. Said computer instructions may be stored on a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, said computer instructions may be transmitted from a website, computer, server, or data center over a wired (eg, coaxial cable, optical fiber) , digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) to another website, computer, server, or data center. Said computer-readable media can be any available media that can be accessed by a computer, or a storage device such as a server, data center, etc. that contains one or more integrated accessible media. Said available media may be magnetic media (eg, floppy disk, hard disk, magnetic tape), optical media (eg, DVD), or semiconductor media (eg, solid state hard disk (SSD)) and the like.

It should be noted that in this article, relative terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that any such actual relationship or sequence exists. between given entities or operations. Moreover, the terms "comprise", "comprise", or any other variations thereof are intended to cover non-exclusive inclusion such that a process, method, product, or apparatus that includes a number of elements includes not only those elements, but also other elements that are clearly not listed, or elements inherent in such a process, method, product or device. Unless further restricted, the elements defined by the statement "including one..." do not preclude the existence of other identical elements in a process, method, product, or device that includes those elements.

Each embodiment in this specification is described accordingly, and the same and similar parts between different embodiments can be referred to each other. Each embodiment focuses on differences from other embodiments. In particular, for the system embodiment, since it is basically the same as the method embodiment, the description is relatively simple, and partial description of the method embodiment can be referred to in appropriate cases.

The foregoing is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modification, equivalent replacement, improvement, etc., made in the spirit and principles of the invention are included within the scope of the present invention.

Claims (36)

1. A task processing method applied to a network topology including a plurality of nodes, including the steps of: obtaining the target data and the target algorithm required to complete the target task; obtaining at least one first type node capable of providing said target data in said network topology, and at least one second type node capable of executing said target algorithm; selection of a node that represents a set of said target data as the first target node from at least said one node of the first type, and a node that represents a set of said target algorithms as a second target node from at least one mentioned node of the second type; controlling said second target node to use said target algorithm to process said target data at said first target node; moreover, the node that provides the set of said target data is selected from at least the mentioned one node of the first type as the first target node, and the node that provides the set of the mentioned target algorithms is selected from at least the mentioned one node of the second type as the second target node; obtaining a combination formed by a node that provides a set of said target data and a node that provides a set of said target algorithms; according to the capability information of each combination, the selection of the target combination, and the node that provides the set of said target data in said target combination is defined as said first target node, and the node that provides the set of said target algorithms in said target combination is defined as second target node; wherein said combination capability information includes a time delay between a node that provides a set of said target data in said combination and a node that provides a set of said target algorithm; or said combination capability information includes the time delay between the node that provides the set of said target data in said combination and the node that provides the set of said target algorithm, as well as the length of time required by the node that provides the set of said target algorithms to execute said target algorithm. algorithm; or said capability information includes the time delay between the node that provides the set of said target data in said combination and the node that provides the set of said target algorithm, as well as the size of the cache space for the node that represents the set of said target algorithms. 2. The method according to claim 1, wherein, in accordance with said information about the possibilities of each combination, a target combination is selected, including the steps of: selecting a combination corresponding to a minimum time delay value between a node that represents the set of said target data in said combination and a node that represents the set of said target algorithms in said combination as said target combination; or selecting a time delay between the node that represents the set of said target data and the node that represents the set of said target algorithms in said combination, and the combination corresponding to the minimum value from the sum of the duration required by the node that represents the set of said target algorithms to execute said target algorithms. algorithms in combination, as mentioned target combination; selecting a combination corresponding to the minimum value in the first ratio of said combination as said target combination, where said first ratio is the time delay ratio between a node that represents a set of said target data and a node that represents a set of said target algorithms in said combination, and values of the cache space size of the nodes that provide the set of said target algorithm in combination. 3. The method according to claim 1, in which, before said obtaining the target data and the target algorithm necessary to perform the target task, in accordance with information about the capabilities of the nodes in the mentioned network topology, the task to be performed is divided into at least one subtask , and one of at least said subtask is defined as said target. 4. The method of claim 1, wherein, in the case where said target data includes a video stream or an audio stream, said controlling said second target node using said target algorithm to process said target data at said first target node includes the steps of: controlling said second target node using said target algorithm to process a first packet of said target data sent by said first target node received; wherein said first data packet includes a video stream or an audio stream of a predetermined duration. 5. The method of claim 4, wherein said controlling said second target node using said target algorithm to process a first packet of target data sent by said first target node received, further comprises the steps of: in the process of receiving said processed first data packet sent by said second target node, controlling said second target node using said target algorithm to process said second packet of said target data sent by said first target node received; wherein said second data packet includes a video stream or an audio stream of a predetermined duration. 6. The method of claim 1, wherein said control of said second target node uses said target algorithm to process said target data at said first target node, comprises the steps of: sending first indication information to said first target node, said first indication information being used to direct said first target node to send said target data to said second target node; sending second indication information to said second target node, said second indication information being used to direct said second target node to use said target algorithm to process said target data; receiving data sent by said second target node obtained after processing said target data using said target algorithm. 7. A task processor applied to a network topology that contains a plurality of nodes, comprising: a first obtaining module used to obtain the target data and the target algorithm necessary to perform the target task; a second acquisition module used to obtain at least one first type node capable of providing said target data in said network topology and at least one second type node capable of executing said target algorithm; a selection module used to select a node that represents the set of said target data as the first target node from at least said one node of the first type, and select a node that represents the set of said target algorithms as the second target node from, at least said one node of the second type; a control module used to control said second target node using said target algorithm to process target data at said first target node; and said selection module contains: a combination submodule used to obtain a combination formed by a node that provides a set of said target data and a node that provides a set of said target algorithm; a selection sub-module used to select a target combination according to the capability information of each combination and to determine a node in said target combination that provides a set of said target data as said first target node, and a node in said target combination that provides a set of said target algorithms as the second target node; wherein said combination capability information includes a time delay between a node that provides a set of said target data in said combination and a node that provides a set of said target algorithm; or said combination capability information includes the time delay between the node that provides the set of said target data in said combination and the node that provides the set of said target algorithm, as well as the length of time required by the node that provides the set of said target algorithms to execute said target algorithm; or said capability information includes the time delay between the node that provides the set of said target data in said combination and the node that provides the set of said target algorithm, as well as the size of the cache space for the node that represents the set of said target algorithms. 8. An electronic device comprising: a processor, a memory and a program code stored in said memory and executed on said processor, implementing a method for processing a task according to any one of paragraphs. 1-6. 9. A computer-readable storage medium that stores program code executed by said processor, implementing a method for processing a task according to any one of paragraphs. 1-6.

Images