Method and device for agile computing

TECHNICAL FIELD [0001] The present invention relates to the field of agile processing, and in particular to the method, nodes and agile computing system.

BACKGROUND ART [0002] In recent years, efforts have been intensified towards the use of computing resources of a number of modules connected by a network to increase the total computing power. Using the resources of many separate computers connected via a network, e.g. the Internet, to solve large scale computational problems is often called grid computing.

[0003] Agile processing also refers to sharing resources, however, agile processing is directed towards a more dynamic and unstable environment with dynamic resource transfer, comparing requirements with network processing. While network processing is intended for large computational resources connected e.g. via a static network, agile processing is in particular intended for use in a much more mobile environment with limited resources, e.g. in the form of intermittent connections between modules, limited power resources, low band connections, etc.

[0004] The article "Agile Computing: Bridging the Gap between Grid Computing and Ad-hoc Peer-to-Peer Resource Sharing", Suri, Niranjan et al., Proceedings of the 3rd iEEE / ACM International Symposium on Cluster Computing and the Grid ( CCGRIDO3), discloses a system of hosts forming groups and adapted for agile processing.

[0005] Nodes in the agile processing group environment are often mobile modules, which leads to the risk of breaking the connection between group nodes.

[0006] US patent application US 2006/0215633 relates to the use of Quality of Service information to monitor a virtual connection to an endpoint for later deciding whether a useful communication session can be established via this connection to the endpoint. With this knowledge, more successful calls can be completed. However, the subject matter of US 2006/0215633 does not take into account factors relevant to agile or distributed processing.

SUMMARY OF THE INVENTION [0007] It has been observed that due to the dynamic environment of the agile processing system, calculations performed in part by a group node may be lost due to the dynamic nature of the system

-2 agile processing. For example, if the connection between the computing node performing a task, e.g. calculations, and the group coordinating node is broken before these calculations are completed, then the calculations already performed will be lost. This means that the compute node has made unnecessary calculations, wasting time and resources, e.g. computing resources and battery power, of both the coordinating node and this computing node.

[0008] Thus, the object of the present invention is to provide a method, nodes and agile processing system that increase the efficiency of the agile processing system.

[0009] Accordingly, a method of operating a group of nodes in a agile processing system is provided, comprising the communication of the coordinating node and the first computing node via the connection, and wherein the first computing node is adapted to act as the coordinating node. This method includes the steps of:

- monitoring of at least one operational parameter indicating the quality of the connection between the coordinating node and the first computing node,

- determining a quality parameter based on this operational parameter, this quality parameter indicating the quality of this connection between the coordinating node and the first computing node, and

- sending information from the first compute node to the coordinating node when the value of this quality parameter falls within the transmission range, this information including information about the part of the task performed by the first compute node.

[0010] An important advantage of the present invention is that it becomes possible to use agile processing also in environments with very high mobility, where nodes often enter and leave groups, which leads to frequent breaking of connections.

[0011] Furthermore, a compute node for agile processing is provided, which compute node is adapted to receive and perform a task from the coordinating node via a connection, comprising:

- a monitoring device adapted to monitor at least one operational parameter indicating the quality of the connection between the coordinating node and that computing node,

- a computing device adapted to determine the quality parameter based on this at least one operational parameter, this quality parameter indicating the quality of this connection between the coordinating node and this calculation node, and

- a transmitting device adapted to send information from this computing node to the coordinating node when the value of this quality parameter is within the transmission range, this information containing information about the part of the task performed by this computing node.

[0012] Furthermore, according to the present invention, a coordinating node for agile processing is provided, said coordinating node being adapted to assign assignments and assign tasks to one or more computing nodes in a group, and is also adapted to receive information about part of the assigned task for the first compute node, e.g. when a quality parameter that is determined on the basis of one or more operational parameters falls within the transmission range.

[0013] An important advantage of the present invention is that the amount of redundancy is reduced

-3 in the agile processing system, which leads to more efficient use of computing resources. [0014] Furthermore, an advantage of the present invention is that the risk of a "blind spot" is reduced, e.g. when the coordinating node is awaiting information from a computing node that has ceased to be available.

BRIEF DESCRIPTION OF THE DRAWINGS [0016] The above and other elements and advantages of the present invention will easily be understood by those skilled in the art, thanks to the following detailed description of examples of its implementation with reference to the figures of the attached drawing, of which:

Fig. 1 illustrates an embodiment of an agile processing system according to the present invention,

Fig. 2 illustrates an embodiment of the agile processing system of the present invention,

Fig. 3 is a block diagram of an example of the method of the present invention,

Fig. 4 is a block diagram of an example of the method of the present invention,

Fig. 5 is a block diagram of an example of the method of the present invention,

Fig. 6 schematically illustrates a computing node according to the present invention,

Fig. 7 schematically illustrates a coordinating node according to the present invention, and

Fig. 8 is a block diagram of an example of the method of the present invention.

DETAILED DESCRIPTION OF THE INVENTION [0016] The figures for clarity are schematic and simplified and only show details that are essential for understanding the present invention, while other details have been omitted. In the figures, the same reference numbers are used for identical or corresponding parts or steps.

In the agile processing system, the coordinating node (s) and the computational node (s) form one or more groups or zones, see e.g. Figs. 1 and 2, A group in the agile processing system may contain any number of coordinating nodes, e.g. one, two, three or more coordinating nodes. In addition, the group may contain any number of compute nodes, e.g., one, two, three or more compute nodes. The present invention relates in particular to a situation in which a node in a group in one way or another loses or is to lose connection with other nodes in a group.

[0018] The coordinating node is responsible for allocating and assigning tasks to other nodes, e.g. one or more computing nodes or other coordinating nodes. The task may include calculations, e.g. data processing and / or measurements of one or more physical quantities. [0019] Figs. 1 and 2 schematically illustrate embodiments of a system adapted for agile processing according to the present invention.

[0020] In Fig. 1, the agile processing system includes a group 2 containing at least one coordinating node, including the first coordinating node COOR1, 10. This first coordinating node 10 is adapted to assign tasks to at least one computing node in group 2, with which at least one compute node includes the first compute node COMP1, 12 and / or the second compute node COMP2, 14. In the illustrated example, the first coordinating node 10 communicates with the first compute node 12 and the second compute node 14 via the corresponding

- first and second connections 16, 18. In the illustrated example, the first connection 16 and the second connection 18 are wireless. In one example, the first connection 16 and / or the second connection 18 may be wired connections.

[0021] Fig. 2 schematically illustrates the agile processing system group 102. This group 102 includes at least one coordinating node, including the first coordinating node COOR1, 10. In Fig. 2, the first coordinating node 10 is adapted to assign tasks to at least one subsequent coordinating node, with the at least one subsequent coordinating node including the second COOR2 coordinating node 20 and / or COOR3 coordinating third node 22, e.g. via connections 24, 26. The second coordinating node 20 allocates tasks from the first coordinating node 10 compute nodes COMP1, 12 and COMP2, 14 via the corresponding link 28, 30, and the third coordinating node 22 allocates tasks from the first coordinating node 10 compute nodes COMP2 and COMP3 via the appropriate connections 32, 34, Optionally, the second coordinating node 20 and the third coordinating node 22 communicate via connection 36. The first coordinating node works as the master coordinating node. In one example, the first coordinating node 10 may also communicate with one or more compute nodes (not shown). In one example, the coordinating node also works as a compute node that performs tasks. Typically, the nodes in the agile processing system are mobile, e.g. mobile terminals or terminals mounted on mobile units, although nodes may also be stationary. Examples of devices that may be adapted to operate as a node in an agile processing system include, but are not limited to; computer, mobile phone, PDA, GPS module, laptop or other mobile devices. In contrast, in principle, any device having computing resources that are not fully utilized can be adapted for agile processing, e.g., devices with free computing resources, or devices adapted to perform the desired task, e.g. in a car.

[0023] Connections between individual nodes in the agile processing system can be wired, wireless, or a combination thereof. A number of protocols can be used for communication between nodes, e.g. Bluetooth, Internet protocol, WIFI, LAN protocols etc. Accordingly, the nodes of the agile processing system are adapted to communicate with other nodes of this system for data, software and other information communication to other nodes of the system. [0024] The agile processing system is described in the publication "Towards an Agile Computing Approach to Dynamie and Adaptive Service-Oriented Architectures", Suri, Niranjan et al., 1st IEEE Workshop on Autonomie Communications and NetWork Management (ACNMO7), page 25- 31, May 2007.

[0025] Furthermore, an advantage of the present invention is that the coordinating node is adapted to allocate tasks more efficiently in terms of both time and computing power.

[0026] In the method of the present invention, the step of monitoring the at least one operational parameter may include measuring the signal strength for the signal between the coordinating node and the computing node in the system. The operating parameter may be the signal strength for any signal from the coordinating node, e.g. signal strength in the pilot channel, signal strength for data signals to the first compute node or other nodes, etc.

[0027] The at least one operational parameter may in one example be the battery condition of the first computing node and / or coordinating node. Additionally or alternatively, the at least one operating parameter may be other parameters, e.g., noise level for calculating the signal-to-noise ratio.

[0028] The quality parameter may be determined or calculated based on one or a number of operational parameters, e.g. one or more sampled values ​​of one or more operational parameters at different times of time. This can be expressed as:

where Q is the quality parameter and O, Is the value of the ith operational parameter in time k

[0029] In one example of the present invention, the step of determining the quality parameter comprises determining this quality parameter based on the signal power to noise ratio for the signal between the coordinating node and the first computing node.

[0030] Preferably, the information transfer step is performed when the quality parameter value falls below the transmission threshold T_trgnsfsr, i.e. when Q <T_trgnsfer.

[0031] Before initiating the task at the computing node, the coordinating node and / or this computing node may be adapted to assess the quality of the connection and decide whether to assign the task based on this assessment. Accordingly, the method may include the step of:

- assigning a task to the first compute node only if the quality parameter reaches or exceeds the initial 7X threshold_aie.

[0032] The task allocation step may be performed before the monitoring step of the at least one operational parameter, i.e. when Q> Τ_ίίΜ.

[0033] Typically, the compute node continues the task to be completed. However, due to limited computing resources or computing node power resources and to avoid wasting power and other resources, it may be desirable to be able to terminate a task in progress at a computing node if the connection between the coordinating node and that computing node is broken. On the other hand, it may be desirable to continue the task to complete this task in the event that the computing node will be able to send the result to the coordinating node via an alternative or the same connection at a later time.

[0034] After the information transfer stage and / or if the calculation node determines that the information transfer stage has not been completed, e.g. due to a connection failure, and / or if the connection between the coordinating node and this calculation node has ceased to exist, a decision can be made, whether this compute node should stop or continue the rest of the task. Accordingly, the method may include the step of:

- decide whether the first compute node should stop or continue the task after the stage of sending information about the completed part of the task.

[0035] By enabling the computing node to interrupt the task of the method according to the present invention,

Thus, it ensures better utilization of resources in the agile processing system, by avoiding unnecessary calculations in the computing node, thus saving limited resources of this node.

In a preferred embodiment of the invention, the value of the quality parameter depends on the quality of the connection between the coordinating node and the computing node. If the connection between the coordinating node and the computing node is weak, it may be desirable to provide alternative routes of communication between them. Accordingly, the method may further comprise the step of:

- attempting to establish an alternative connection between the coordinating node and the first compute node, e.g. when the quality parameter value is within the configuration range.

[0037] The step of attempting to establish an alternative connection may be performed when the value of the quality parameter is below the configuration threshold T_sehjp, i.e. when Q ** fsetup [0038] The transmission range and configuration range can be set and / or changed by the coordinating node and / or the first computing node.

[0039] An alternative connection between the coordinating node and the first computing node may be a connection via the next computing node and / or the next coordinating node in the group.

[0040] When the coordinating node receives information from the first compute node about the part of the task performed by the first compute node, the remainder of this task can be reallocated to ensure that the task is completed.

[0041] Accordingly, the method may comprise the steps of:

- the coordinating node assigning the remainder of the task of the first compute node based on information from that first compute node, and

- the coordinating node assigning the remainder of the task of the first compute node to at least one compute node in the group.

[0042] The remainder of the task may be reallocated to the first compute node if the connection has been restored or still exists and / or other compute nodes in the group. The reallocation of the remainder of the task to the first compute node can be determined based on the value of the quality parameter, e.g. the remainder of the task can be reallocated to the first coordinating node if Q £ Τ_{/ ηΛΛ}/_β.

[0043] Preferably, monitoring of at least one operational parameter is performed by the first computing node, however also the coordinating node may additionally or alternatively monitor at least one operational parameter.

[0044] The information transmitted in the information transmission step may include information regarding the value of the quality parameter and / or information that the value of this quality parameter has fallen in the transmission.

[0045] Fig. 3 illustrates an example of a method according to the present invention. This method includes the step 108 of assigning a task to the first compute node. The method further includes the step 110 of monitoring at least one operational parameter. In step 110, the signal strength for the signal from the coordinating node is measured to calculate the quality parameter in step 112. If this quality parameter is in the transmission range, the method proceeds to step 114, in which information about the part of the task performed by the first node is sent from the first computing node to the coordinating node

-Ί calculation to at least try to avoid losing an already completed part of the task. Otherwise, the method returns to monitoring the at least one operational parameter, the information transmitted in step 114 may further include information about the value of the quality parameter. After the transfer in step 114, the method proceeds to step 116, in which the coordinating node evaluates the information sent from the first compute node and on this basis reallocates the remainder of the task to be completed by the compute nodes in the group. This remaining part of the task can be reassigned to the first compute node or other compute nodes in the group. Preferably, steps 110 and 112 are carried out at the first computing node. In one example of the method of the present invention, steps 110 and 112 are performed at the coordinating node, thereby requiring that coordinating node to send a request for information transfer at the first computing node at step 114.

[0046] Fig. 4 illustrates another example of a method according to the present invention. In the illustrated example, this method proceeds to step 118 after step 114, in which information about the completed part of the task is sent. Step 118 includes the step of deciding whether the first compute node should stop or continue with the task. This decision can be based on the value of the quality parameter. Alternatively or additionally, this decision may be based on whether the transmission in step 114 has been completed or discontinued. If the task continues, the method optionally proceeds to step 120, in which, if the decision in step 118 is made by the first compute node, before returning to step 110, the coordinating node is informed that the first compute node continues the task.

[0047] In one example of the present invention, the step of deciding whether to interrupt or continue the task after the step of sending information about the completed part of the task includes selecting a calculation scheme from a number of calculation schemes, e.g. based on the current quality parameter value. [0048] Fig. 5 illustrates another example of a method according to the present invention. In addition to the stages previously discussed with reference to Fig. 4, the method further includes the step 124 of attempting to establish an alternative connection between the coordinating node and the first compute node, e.g., the first compute node 12 sends a request to the coordinating node 10. The step 124 of attempting to establish an alternative connection may depend on whether the quality parameter value Q is within the configuration range, as illustrated by optional step 126. Additionally or alternatively, the step 124 of attempting to establish an alternative connection may depend on the current capacity and capabilities of the compute node, e.g. battery level, load of other tasks, memory status, etc. At step 120, the first compute node continues the task and optionally informs the node coordinating.

[0049] In the illustrated example, step 124 of attempting to establish an alternative connection between the coordinating node and the first computing node includes attempting to establish a new connection between that coordinating node and the first computing node, if the quality parameter indicates that the connection between that coordinating node and the computing node , e.g. the first connection 16 has been broken.

[0050] An alternative connection between the coordinating node and the first computing node may

- be connected via one or more other nodes in the system, e.g. a second compute node in the same group and / or a second coordinating node in the same or different group.

[0051] The monitoring device in the computing node according to the present invention may be adapted to measure the signal strength for one or more signals between the coordinating node and the computing node. In one example, the monitoring device is adapted to measure the signal strength for a signal from a coordinating node, which signal indicates the presence of a coordinating node.

[0052] The quality parameter Q may be determined by a computing node or, using a more centralized approach, the quality parameter may be determined by a coordinator. The quality parameter may be determined or calculated on the basis of one or more operational parameters. In one example of a computing node, the computing device may be adapted to determine a quality parameter based on the value of the signal power to noise ratio of the signal between the coordinating node and the computing node.

[0053] The transmitting node of the computing node may be adapted to send information to the coordinating node when the value of the quality parameter falls below the transmission threshold, e.g. when Q <^ Transfer- The transmitting device may be a transceiver.

[0054] In one example of a computing node, the quality parameter is calculated as the value of the signal power to noise ratio of the signal from the coordinating node.

[0055] The compute node may also be adapted to decide whether to interrupt or continue the task after sending information about the completed part of the task. This decision may be based on the value of the quality parameter, which may indicate that no connection is available between the calculation node and the coordinating node, e.g. by setting the value to zero.

[0056] Furthermore, the computing node may be adapted to attempt to establish an alternative connection between the coordinating node and the computing node when the quality parameter value is within the configuration range, e.g., when the quality parameter Q is equal to or less than the configuration threshold value.

[0057] The computing node may be adapted to send information to the coordinating node, the information containing information indicating that the value of the quality parameter was within the transmission range, e.g. that the signal power to noise ratio for the connection between that computing node and this node coordinate, fell below the transfer threshold Ttransfer [0058] Fig. 6 illustrates schematically an example of a computing node for agile processing according to the present invention, e.g. a first computing node for use in the method according to the present invention. Compute node 12 is adapted to receive and perform a task from a coordinating node, e.g., coordinating node 10, via a connection, e.g., first connection 16. Computing node 12 includes a monitoring device 202 adapted to monitor an operational parameter, e.g., signal strength for a signal from a coordinating node that indicates the quality of connection 16, as well as a calculating device 204 adapted to determine the signal-to-noise ratio as a quality parameter based on measured signal strength. In addition, the computing node 12 includes a transmitting device 206 adapted to transfer information from this computing node 12 to the node

-Coordinating 10 when the value of this signal-to-noise ratio falls below the transmission threshold. The transmitted information contains information about the part of the task performed by this computing node 12. The devices 202, 204, 206 are connected to each other via wires 208.

[0059] Furthermore, the computing node 12 is adapted to decide whether to interrupt or continue the task after sending the information. This decision is based on the value of the quality parameter, which is updated frequently, e.g. every 2 seconds.

[0060] In addition, the compute node 12 is adapted to attempt to establish an alternative connection between the coordinating node 10 and the compute node 12, via another node in the group, e.g., the second compute node 14, when the quality parameter value falls below the configuration threshold value.

[0061] The coordinating node according to the present invention may be adapted to determine the allocation of the remainder of the task of the first compute node based on information from the first compute node. In addition, the coordinating node may be adapted to reallocate the remainder of this task or the entire task of the first compute node to at least one compute node in the group, e.g. to the first compute node if the nadai connection is of sufficient quality or has been restored, or to a second compute node with similar capabilities.

[0062] Fig. 7 schematically illustrates a coordinating node 10 according to the present invention. This coordinating node 10 includes a processor 302 and a transceiver 304 connected to it by a wire 306. The processor 302 is adapted to assign assignments and assign tasks to one or more computing nodes in a group. In addition, the coordinating node 10 includes a transceiver 304 that is adapted to receive information from the first compute node 12 regarding a portion of the task assigned to the first compute node when a quality parameter that is determined based on one or more operational parameters falls within transmission range.

[0063] Furthermore, the coordinating node 10 is adapted to determine the allocation of the remainder of the task of the first compute node, based on information from this first compute node, and by means of the transceiver 304 is adapted to allocate this remaining part of the task of the first compute node, at least one compute node in the group.

In one embodiment of the present invention, the first compute node continues to perform the task for a predetermined time before attempting to reconnect the coordinating node and the first compute node directly or through one or more other nodes. Such an embodiment is illustrated in Fig. 8, wherein the method in addition to the steps of the exemplary method of Fig. 5, further includes a step 130 of continuing the task for a predetermined time, e.g., more than 10 seconds, e.g., from about 10 seconds to about 10 minutes. This predetermined time may have a predetermined length, e.g. about 30 seconds, 1 minute, 3 minutes or 5 minutes. The length of this time may depend on the value of the quality parameter.

[0065] It should be noted that in addition to the embodiments of the present invention, shown in the figures of the attached drawing, the present invention may be implemented in various forms and should not be

- taken as limited to those embodiments shown herein. Instead, these embodiments are ensured that the present disclosure is thorough and complete and fully conveys the concept of the present invention to those skilled in the art.