Constraint-based method and system for dynamic construction of a mining power network cluster

Abstract

The application provides a kind of computing power network cluster dynamic construction method based on constraint condition, the method includes: the information of registered computing power node and extract the computing task characteristics in task demand;According to the computing task characteristics, the ability of computing power node is evaluated, wherein the ability evaluation is obtained according to the characteristic function of computing task;The computing power node as task initiation node is as root node, and the next computing power node is evaluated and selected based on the reachable routing path of routing table and the constraint condition, wherein the evaluation value of branch computing power node is obtained by evaluation function according to the ability of evaluation node through the ability evaluation Vi;The evaluation value search fitting of branch computing power node that can be routed is carried out by level by level, and the best computing power multistage collaborative computing power node cluster for the computing task is obtained.The application also provides a kind of computing power network cluster dynamic construction system based on constraint condition.

Description

TECHNICAL FIELD

[0001] The application relates to a constraint-based computing power network cluster dynamic construction method, system, device and computer readable storage medium. BACKGROUND

[0002] With the continuous deepening of digital transformation, computing power demand is accelerating iteration: data load is shifting from single load to mixed load, and application scenarios are evolving from simple to complex. Even image training with a relatively small amount of engineering includes data forwarding, compressed storage, encryption and decryption, file systems, data queries, image rendering and other computing modules, involving computing power of multiple chips. Some long-tail and urgent computing scenarios may also require customized dedicated services.

[0003] Currently, there are patents mainly based on network computing power routing and management-based computing power arrangement. For example, computing power routing is based on acquired computing power information to generate a computing power routing table, which records the computing power capability parameters and network capability parameters of local computing power nodes and neighbor computing power nodes. Moreover, a corresponding computing power forwarding table is constructed at each computing power routing node, and the computing power requests received by the computing power routing node are routed and arranged according to the corresponding computing power forwarding table. Currently, the routing and arrangement of computing power networks are mainly based on network capabilities.

[0004] The first step of computing power network construction is to coordinate and connect data centers distributed in various parts of the country through routing and arrangement, and to reasonably allocate and coordinate computing power services through a computing power scheduling system. Currently, the bottleneck is that a large amount of computing power routing infrastructure needs to be added when allocating computing power, and routing and arrangement overheads belong to network settings, increasing network complexity and management difficulty.

[0005] In addition, whether computing power, algorithms, data and application highly coordinated integrated resources can be output to the market based on scenario demand is also a problem that computing power service network service providers must address. SUMMARY

[0006] The application aims at the cooperation of computing power network resources, the scene increases the node demand of the computing power cluster, in order to solve the problem that the computing power network mainly depends on the network capability for arrangement and scheduling, a computing power network cluster dynamic construction method based on constraint conditions is provided, the method comprises the following steps: registering the information of the computing power node and extracting the computing task characteristics in the task demand; the capability of the computing power node is evaluated according to the computing task characteristics, wherein the capability evaluation is obtained according to the characteristic function of the computing task; the computing power node as the task initiation node is taken as the root node, and the next computing power node is evaluated and selected based on the reachable routing path of the routing table and the constraint condition, wherein the evaluation value of the branch computing power node is obtained through the evaluation function of the evaluation node capability according to the capability evaluation Vi; the evaluation value of the routable branch computing power node is searched and fitted through step-by-step, and the best computing power multi-level cooperative computing power node cluster for the computing task is obtained.

[0007] According to the advantageous design scheme of the application, the computing power node information comprises node entrance device information and node self-resource information.

[0008] According to the advantageous design scheme of the application, the node entrance device information refers to node control device MAC, and the node self-resource information refers to computing resource C, multi-level storage resource buf and maximum bandwidth resource B.

[0009] According to the advantageous design scheme of the application, the constraint condition is agreed according to the scene template and the service level agreement, and the constraint condition comprises transmission bandwidth, computing resource and routing table.

[0010] According to the advantageous design scheme of the application, a plurality of constraint conditions form a constraint rule library, and the constraint rule library can continue to add constraint conditions according to the scene, such as computing priority and data priority.

[0011] According to the advantageous design scheme of the application, the capability of the computing power node i comprises computing capability, storage capability and transmission capability.

[0012] According to the advantageous design scheme of the application, the characteristic function of the computing task is as follows:

[0013] V i =p∑C i +n∑Buf i +m∑B i

[0014] Wherein C is the current idle computing resource under the unified computing power metric, Buf is the current idle multi-level storage resource, B is the current allocable bandwidth resource, and pnm is the probability set according to the scene, wherein p+n+m is less than or equal to 1.

[0015] According to the advantageous design scheme of the application, in the scene of data priority, the value of n is set to be greater than p and m.

[0016] According to an advantageous design scheme of the present application, the evaluation function for evaluating the node capability according to the characteristic function of the computing task is:

[0017]

[0018] wherein is the average value of the evaluation results of all branch computing power nodes with the computing power node i as the root node, Ti is the access frequency of the computing power node i, and c is a constant manually set.

[0019] According to an advantageous design scheme of the present application, the evaluation value ei is the capability value evaluation of all branch computing power nodes of the current root node to the contribution degree of the cluster, and the evaluation value ei is dynamic.

[0020] According to an advantageous design scheme of the present application, the selection of the branch computing power node is based on the reachable routing path of the routing table, until the branch computing power node of the cluster satisfying the computing task is searched, the evaluation result is propagated back to the root node currently initiating the task, and then the evaluation values ei along the way are updated, and the backward propagation ensures that the statistical information of each node can reflect the evaluation results of all descendants of the node.

[0021] According to an advantageous design scheme of the present application, the routing table is updated according to the evaluation value ei.

[0022] According to an advantageous design scheme of the present application, the computing power node cluster can be regarded as a computing power node.

[0023] According to an advantageous design scheme of the present application, in addition to the root node, the computing task initiated from the resource-constrained computing power node cooperates with the edge computing power node, the edge center computing power node and the center computing power node to realize local optimization.

[0024] According to an advantageous design scheme of the present application, in addition to the root node, the computing task initiated from the center computing power node at a certain level cooperates with the routable center computing power node, the edge center computing power node and other related routable computing power node resources to realize global optimization.

[0025] According to an advantageous design scheme of the present application, the center computing power node at a certain level is the edge center or the data center.

[0026] According to another aspect of the present application, a computing power network cluster dynamic construction system based on constraint conditions is provided, comprising:

[0027] The registration module is configured to register the information of the computing power node i and extract the computing task characteristics in the task demand.

[0028] An evaluation module is configured to evaluate the capability of the computing power node i according to the characteristics of the computing task, wherein the capability evaluation value Vi is obtained according to a characteristic function of the computing task;

[0029] A node control module is configured to take the computing power node i as the task initiation node as a root node, and evaluate and select the next computing power node based on the reachable routing path of the routing table and the constraint condition, wherein the evaluation value ei of the branch computing power node is obtained by the evaluation function of the node capability according to the capability evaluation value Vi;

[0030] A cluster module is configured to search for the fitting of the evaluation value ei of the branch computing power node through step-by-step routing, and obtain the optimal computing power multi-level collaborative computing power node cluster for the computing task.

[0031] According to another aspect of the present application, an electronic device is provided, comprising:

[0032] at least one processor;

[0033] a memory, wherein at least one program is stored on the memory, and when the at least one program is executed by the at least one processor, the at least one processor implements the above-mentioned constraint condition-based computing power network cluster dynamic construction method.

[0034] According to still another aspect of the present application, a computer readable storage medium is provided, wherein a computer program is stored on the computer readable storage medium, and when the program is executed by a processor, the above-mentioned constraint condition-based computing power network cluster dynamic construction method is implemented.

[0035] In general, the present application is a resource collaboration method and system under the computing power network technology system. The computing power network is a new type of information infrastructure that allocates and flexibly schedules computing resources, storage resources and network resources according to business needs among clouds, networks and edges. It can flexibly meet the computing power needs of different business scenarios, and also changes the traditional service paradigm to provide customers with one-stop services such as computing, application, data, optimization, consultation, operation and maintenance according to demand. The present application can increase the capability of node collaboration under the condition of existing node full load, provide full range of collaboration from cloud to edge to end, and realize "edge scheduling and edge computing" and "edge computing and edge returning results". BRIEF DESCRIPTION OF DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0037] Figure 1 The flowchart of the method for dynamically constructing a constraint-based computing power network cluster is provided.

[0038] Figure 2 The implementation schematic diagram of the method for dynamically constructing a constraint-based computing power network cluster is provided. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the technical solutions of the present application, the technical solutions in the present application will be described clearly and completely below with reference to the drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0040] The present application provides a method and system for computing power network resource cooperation, computing power network node cooperative cluster construction, and adding nodes to the cooperative cluster and dynamically constructing a new cluster.

[0041] The present application will determine the cooperative combination of multi-level computing power nodes in a step-by-step search manner. The definition of "level" is that the nodes requiring one routing are one level. The present application takes the characteristics of computing tasks (such as computing priority, data priority) as the main target, meets the bandwidth, computing resource, routing table and other constraint conditions, takes the computing task characteristic evaluation of the computing power network node to provide the resource guarantee for the computing task as the evaluation function, takes the task initiation node as the root node, and through the step-by-step evaluation value search fitting of the routable nodes, obtains the best computing power multi-level cooperative node cluster for the computing task, which includes but is not limited to the following two cooperative ways, one is that the computing task initiated from the resource-limited computing power node cooperates with the edge computing power node, the edge center computing power node and the center computing power node to realize local optimization, and the other is that the computing task initiated from a certain level center computing power node (which can be an edge center or a data center) cooperates with the routable center computing power node, the edge center computing power node and other related routable computing power nodes to form a cooperative resource and realize global optimization.

[0042] Figure 1The application provides a constraint condition-based computing power network cluster dynamic construction method, which comprises the following steps: registering information of a computing power node and extracting a computing task feature in a task demand; evaluating the capability of the computing power node according to the computing task feature, wherein the capability evaluation value is obtained according to a feature function of the computing task; taking the computing power node as a task initiation node as a root node, and evaluating and selecting a next computing power node based on a reachable routing path of a routing table and the constraint condition, wherein the evaluation value of a branch computing power node is obtained according to an evaluation function of the evaluation node capability through the capability evaluation value Vi; and obtaining the optimal computing power multi-level cooperative computing power node cluster for the computing task through the evaluation value search fitting of the routable branch computing power node in stages.

[0043] According to an advantageous design scheme of the application, the computing power node information comprises node entry device information and node self-resource information.

[0044] According to an advantageous design scheme of the application, the node entry device information refers to node management device MAC, and the node self-resource information refers to computing resource C, multi-level storage resource buf and maximum bandwidth resource B. In addition, the computing power node can be a device or a computing power cluster under a node management device.

[0045] According to an advantageous design scheme of the application, the constraint condition is agreed according to a scene template and a service level agreement, and the constraint condition comprises transmission bandwidth, computing resource and a routing table.

[0046] According to an advantageous design scheme of the application, a plurality of constraint conditions form a constraint rule library, and the constraint rule library can continue to add constraint conditions according to a scene, such as computing priority and data priority.

[0047] According to an advantageous design scheme of the application, the capability of the computing power node i comprises computing capability, storage capability and transmission capability.

[0048] According to an advantageous design scheme of the application, the feature function of the computing task is as follows:

[0049] V i = p∑C i + n∑Buf i + m∑B i

[0050] wherein C is the current idle computing resource under unified computing power measurement, Buf is the current idle multi-level storage resource (including a first-level cache, a second-level cache, SRAM, DRAM and the like in practice), B is the current allocatable bandwidth resource, and p, n and m are probabilities set according to a scene, wherein p + n + m is less than or equal to 1, for example, in a data priority scene, the value of n is set to be greater than p and m.

[0051] According to an advantageous design scheme of the present application, the evaluation function for evaluating the node capability according to the characteristic function of the computing task is:

[0052]

[0053] wherein, is the average value of the evaluation results of all branch computing nodes with node i as the root node.

[0054] The average value reflects the expectation of the return value that node i can provide according to the current search result observation. Ti is the access frequency of node i, reflecting the number of times node i is selected. c is a manually set constant, since the order of the computing nodes is indefinite, c is needed to balance the demand for adoption and the demand for further exploration of the entire algorithm. Starting from the computing task initiation node, for each branch computing node of a non-branch computing node (a non-branch computing node refers to a computing node without branches), the algorithm will calculate an evaluation value ei, and in the same level, according to the ei value, a branch computing node is selected for the next step, until the branch computing node under the c value constraint is reached. The V of the branch computing node (selected once) is a typical characteristic function of the computing task.

[0055] According to an advantageous design scheme of the present application, the evaluation value ei is the capability value evaluation of all branch computing nodes of the current root node for the contribution of the cluster, and the evaluation value ei is dynamic.

[0056] According to an advantageous design scheme of the present application, the selection of the branch computing node is based on the reachable routing path of the routing table, until the branch computing node of the cluster that meets the computing task is searched, and the evaluation result is propagated back to the root node of the current task initiation, and then the evaluation value ei along the way is updated, and the reverse propagation ensures that the statistical information of each node can reflect the evaluation results of all descendants of the node.

[0057] According to an advantageous design scheme of the present application, the routing table is updated according to the evaluation value ei.

[0058] According to an advantageous design scheme of the present application, the computing node cluster can be regarded as a computing node.

[0059] According to an advantageous design scheme of the present application, in addition to the root node, the computing task initiated from the resource-constrained computing node cooperates with the edge computing node, the edge center computing node, and the center computing node to achieve local optimization.

[0060] According to an advantageous design scheme of the present application, in addition to the root node, the computing task initiated from a center computing node at a certain level cooperates with the routable center computing node, the edge center computing node, and other related routable computing node resources to achieve global optimization.

[0061] According to an advantageous design of the present application, the certain first-level center computing power node is an edge center or a data center.

[0062] According to another aspect of the present application, a constraint-based computing power network cluster dynamic construction system is provided, comprising:

[0063] A registration module is configured to register information of a computing power node i and extract computing task features in a task demand;

[0064] An evaluation module is configured to evaluate the capability of the computing power node i according to the computing task features, wherein the capability evaluation value Vi is obtained according to a feature function of the computing task;

[0065] A node control module is configured to take the computing power node i as a root node as a task initiating node, and evaluate and select a next computing power node based on reachable routing paths of a routing table and constraint conditions, wherein the evaluation value ei of a branch computing power node is obtained through the capability evaluation value Vi according to an evaluation function of node capability;

[0066] A cluster module is configured to search and fit the evaluation value ei of the branch computing power node through a level-by-level manner, so as to obtain a computing power node cluster for optimal multi-level cooperation of the computing task.

[0067] According to another aspect of the present application, an electronic device is provided, comprising:

[0068] At least one processor;

[0069] A memory, wherein at least one program is stored on the memory, and when the at least one program is executed by the at least one processor, the at least one processor implements the constraint-based computing power network cluster dynamic construction method described above.

[0070] According to still another aspect of the present application, a computer readable storage medium is provided, wherein a computer program is stored on the computer readable storage medium, and when the program is executed by a processor, the constraint-based computing power network cluster dynamic construction method described above is implemented.

[0071] Figure 2 An implementation schematic diagram of the constraint-based computing power network cluster dynamic construction method provided by the present application is shown.

[0072] The dynamic construction process of the computing power cluster according to the present application is a process of selecting nodes according to an algorithm based on reachable routing paths of a routing table, until a branch computing power node i of a cluster satisfying a computing task is searched, and then an evaluation result is propagated back to a root node currently initiating a task, and then ei values along the way are updated, and the backward propagation ensures that statistical information of each node can reflect evaluation results of all descendants of the node. The statistical information of the node i includes two parts: a total evaluation result ∑i V i and total access times ∑ i T i The total evaluation result represents the sum of all evaluation results of the node according to the task characteristics, and the total access times represent the number of occurrences of the node on the back propagation path. The present application can be dynamically constructed in a cooperative state, and existing cooperative resource nodes can be regarded as a node.

[0073] According to one preferred embodiment of the present application, a computing task is initiated from a resource-limited terminal (computing power node), and the access gateway connected thereto, for example, CPE, analyzes the registration information of the task and constructs a task function, searches nodes at each level according to the reachable route with the CPE as the starting node, the resources of the nodes at each level include base stations BBU, MEC, end office computer rooms, metropolitan computer rooms, etc. connected to the route, until the leaf node of the cluster meeting the computing task is searched, the evaluation result is back propagated to the starting node of the task, and then the ei value along the way is updated. The back propagation ensures that the statistical information of each node can reflect the evaluation results of all descendants of the node. The present application can be dynamically constructed in a cooperative state, and existing cooperative resource nodes can be regarded as a node.

[0074] According to another preferred embodiment of the present application, for some regular tasks (such as protocol analysis, video stream analysis, etc.), a computing task can be initiated by a center computing power node at a certain level (which can be an edge center or a data center), and the center computing power node, edge center computing power node and other related routable computing power node resources of the reachable route form a cooperation to achieve the global optimization for the scene. For example, a certain CDN is specially used for video acceleration, and its function scene is relatively fixed, that is, all nodes reachable by the search route can be searched to form a special video acceleration cluster to provide global optimal video acceleration service for customers.

[0075] It should be noted that there are many types of computing power nodes, and the operation room at each level can be regarded as a computing power node, for example, the operation room of the provincial backbone can be regarded as a large computing power node (i.e. the center computing power node in the text, because almost all the business in the province will pass through this node, and its routing and computing capacity covers the province). However, for a local network, there is also a center node at this level, for example, the street room of a business district can be regarded as a computing power node, and it is also the center computing power node of all devices in the street at this level. In summary, the present application is a resource coordination method and system under the computing power network technology system. The computing power network is a new type of information infrastructure that allocates and flexibly schedules computing resources, storage resources and network resources according to business needs in the cloud, network and edge. It can flexibly meet the computing power needs of different business scenarios, and also change the traditional service paradigm to provide customers with one-stop services such as computing, application, data, optimization, consultation, operation and maintenance. The present application can increase the node coordination capability under the existing node full load, provide full range coordination from cloud to edge to end, realize "edge scheduling and edge computing", and "edge calculation and result return".

[0076] Those of ordinary skill in the art can understand that all or some of the steps in the method disclosed above, the functional modules / units in the system and the device can be implemented as software, firmware, hardware and appropriate combinations thereof. In a hardware implementation, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component can have multiple functions, or one function or step can be performed by several physical components in cooperation. Some or all physical components can be implemented as software executed by a processor, such as a central processor, a digital signal processor or a microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software can be distributed on a computer readable medium, which can include computer storage media (or non-transitory media) and communication media (or transitory media). As known to those of ordinary skill in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. In addition, as known to those of ordinary skill in the art, communication media generally includes computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and can include any information delivery medium.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions described in the foregoing embodiments can still be modified, or some or all of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for dynamically constructing a computing power network cluster based on constraints, comprising: Register the information of computing node i and extract the computing task characteristics from the task requirements; The capability of computing node i is evaluated based on the characteristics of the computing task, wherein the capability estimate Vi is obtained based on the characteristic function of the computing task. The computing power node i, which is the task initiation node, is taken as the root node, and the next computing power node is evaluated and selected based on the reachable route path in the routing table and the constraints. The evaluation value ei of the branch computing power node is obtained by the capability evaluation Vi according to the evaluation function of the evaluation node capability. By progressively searching and fitting the evaluation value ei to the routable branch computing power nodes, an optimal multi-level collaborative computing power node cluster for the computing task is obtained. The branch computing power nodes are selected based on the reachable routing paths in the routing table until all branch computing power nodes in the cluster that satisfy the computing task are found. The evaluation results of all branch computing power nodes are backpropagated back to the root node that initiated the task, and then the evaluation value ei along the path is updated. Backpropagation ensures that the statistical information of each branch computing power node can reflect the evaluation results of all its descendants.

2. The method according to claim 1, characterized in that, The computing node information includes node entry device information and node's own resource information.

3. The method according to claim 2, characterized in that, The node entry device information refers to the node management device MAC, and the node's own resource information refers to computing resources C, multi-level storage resources buf, and maximum bandwidth resources B.

4. The method according to claim 1, characterized in that, The constraints are defined according to the scenario template and service level agreement, and include transmission bandwidth, computing resources, and routing tables.

5. The method according to claim 4, characterized in that, Multiple constraints form a constraint rule base, and the constraint rule base can continue to add constraints according to the scenario. The constraints include computation priority and data priority.

6. The method according to claim 1, characterized in that, The capabilities of computing node i include: computing power, storage capacity, and transmission capacity.

7. The method according to claim 6, characterized in that, The characteristic function of the computational task is: In i =p∑C i +n∑Buf i +m∑B i Where C is the currently idle computing resource under the unified computing power, Buf is the currently idle multi-level storage resource, B is the currently allocable bandwidth resource, and p, n, and m are probabilities set according to the scenario, where p+n+m is less than or equal to 1.

8. The method according to claim 7, characterized in that, In data-first scenarios, the value of n is set to be greater than p and m.

9. The method according to claim 7, characterized in that, The evaluation function for assessing node capabilities based on the characteristic function of the computational task is as follows: in Ti is the average of the evaluation results of all branch computing nodes with computing node i as the root node, Ti is the number of visits to computing node i, and c is a manually set constant.

10. The method according to claim 9, characterized in that, The evaluation value ei is an assessment of the contribution of all branch computing power nodes of the current root node to the cluster, and the evaluation value ei is dynamic.

11. The method according to claim 1, characterized in that, The routing table is updated based on the evaluation value ei.

12. The method according to claim 1, characterized in that, The computing node cluster can be regarded as a single computing node.

13. The method according to claim 1, characterized in that, Apart from the root node, computing tasks initiated from any resource-constrained computing node collaborate with edge computing nodes, edge-center computing nodes, and center computing nodes to achieve local optima.

14. The method according to claim 1, characterized in that, Apart from the root node, computing tasks initiated from a certain level of central computing node collaborate with routable central computing nodes, edge central computing nodes, and other related routable computing node resources to achieve global optimization.

15. The method according to claim 14, characterized in that, The aforementioned central computing node is either an edge center or a data center.

16. A constraint-based dynamic construction system for computing power network clusters, comprising: The registration module sets the information used to register computing node i and extracts the computing task characteristics from the task requirements. An evaluation module is configured to evaluate the capabilities of the computing node i based on the characteristics of the computing task, wherein the capability estimate Vi is obtained based on the characteristic function of the computing task; The node control module is configured to use the computing power node i, which is the task initiation node, as the root node, and to evaluate and select the next computing power node based on the reachable route path and constraints in the routing table. The evaluation value ei of the branch computing power node is obtained by the capability evaluation Vi according to the evaluation function of the node capability. The cluster module is configured to obtain an optimal multi-level collaborative computing node cluster for the computing task by progressively searching and fitting the evaluation value ei to the routable branch computing power nodes. The branch computing power nodes are selected based on reachable routing paths in the routing table until all branch computing power nodes satisfying the computing task are found. The evaluation results of all branch computing power nodes are backpropagated back to the root node that initiated the task, and the evaluation value ei along the path is updated. Backpropagation ensures that the statistical information of each branch computing power node reflects the evaluation results of all its descendants.

17. An electronic device comprising: At least one processor; A memory storing at least one program, which, when executed by the at least one processor, causes the at least one processor to implement the method of any one of claims 1-15.

18. A computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the method of any one of claims 1-15.

Images