Dynamically dispatching special tasks in a distributed network

Abstract

A method for operating a network, in which nodes distribute information to each other in the network and a proper subset of nodes performs at least one special task that affects the overall operation of the network, comprising the steps of: assigning weights to each node in the network, each node may: as a transferor node, conditionally or unconditionally delegate its own weight share to other receiver nodes; based on the weight of each node plus the share delegated to it as a receiver node and minus the share delegated to other receiver nodes as a transferor node, a proper subset of nodes assigned with special tasks is selected according to predetermined criteria; wherein the transferor node delegates the share: conditionally to at least two receiver nodes under a release condition, the release condition being that after the selection, the corresponding receiver node does not belong to the proper subset; and having an additional restriction: when accepting one of the receiver nodes into the proper subset, at most the same share that is conditionally delegated to multiple receiver nodes can be considered.

Description

Technical Field

[0001] The invention relates to a distributed network of computer-implemented nodes in which special tasks are performed by a proper subset of the nodes, wherein the special tasks are related to the operation of the network as a whole. Background Art

[0002] A blockchain is a data storage device organized as a chain of data blocks. Before being added to the blockchain, each new data block is checked for compliance with a predetermined set of rules and additionally contains the cryptographic hash value of the predecessor block. Thus, every subsequent change of the information in a data block results in at least one cryptographic hash value in the chain no longer matching the content of the block in question, unless the hash values ​​of all existing subsequent blocks are also adapted.

[0003] The management of the blockchain by means of a distributed network of computer-implemented nodes actually hinders such an adaptation. In the network, all nodes exchange information that is already in the blockchain and that is to be newly recorded, not only with each other but also with the requesting client stations. Tasks for managing and continuing the blockchain, such as adding new blocks to the blockchain, are special tasks that can be assigned to a subset of nodes within the network according to an automated consensus mechanism. Subsequent changes to the data in the blockchain therefore require the consensus of the majority of the nodes in the network.

[0004] In this case, the probability that a given node allows the next block to be added to the blockchain can depend, for example, on the computing effort that the node has exerted for the network (“Proof of Work”). Alternatively, the probability can depend, for example, on the amount of collateral, with which the corresponding node guarantees compliance with the rules (“Proof of Stake”). In this case, in particular, the node can be left to contribute to the collateral amounts of other nodes from a predefined list on its own terms (“Delegation Proof of Stake”) in order to support the node in terms of its “candidacy” for the addition of the next block. Summary of the invention

[0005] In the context of the present invention, a method for operating a network consisting of a plurality of computer-implemented nodes is developed. The nodes distribute information among each other in the network and exchange information with requesting client stations. If the network manages a blockchain, for example, the client stations can use their own requests to read out, for example, already existing information from the blockchain and / or register new information for storage in the blockchain.

[0006] The proper subset of nodes performs at least one special task that influences the operation of the network as a whole. If the network manages a blockchain, for example, the special task can, in particular, include at least one partial contribution to the creation of a new block to be added to the blockchain and / or the checking of the correctness of the information contained in the block with respect to the rules predefined for the blockchain and / or the addition of the block to the blockchain in the event of a positive outcome of the check.

[0007] For example, the information in a new block can include transactions. The check for correctness can then be extended, for example, based on whether the transactions are each signed by an authorized person and whether one and the same balance is not spent multiple times in an impermissible manner. The creation of a block can include, for example:

[0008] ● Collect transaction candidates from other nodes and / or client sites;

[0009] ●Select a subset of valid transactions from this;

[0010] ● forming a coordinated sequence of valid transactions, where the sequence is prioritized, for example, according to the fees sent with the transaction; and

[0011] ● From this sequence a subset of possible members of the block is formed, which subset takes into account, for example, the space requirements of the individual transactions and the maximum available space in the block to be created.

[0012] Alternatively or also in combination, a block in a blockchain can include at least one fact to be authenticated. Examples of such facts are event protocols (log files) of a computer system, contractual provisions, timestamps or also measured values ​​that must be collected in a demonstrable form.

[0013] In the network, each node is assigned a weight. In this case, each node has the possibility to delegate its share of the weight as a transferor node conditionally or unconditionally to other recipient nodes. In this case, an unconditional delegation is always immediately effective, while the validity of a conditional delegation depends on a deferral condition or a release condition and can also be changed subsequently.

[0014] Based on the weight of each node plus its delegated share as a receiver node and minus its delegated share to other receiver nodes as a lender node, a proper subset of nodes assigned at least one special task is selected according to at least one predefined criterion.

[0015] This principle works similarly to the previously mentioned "delegated proof of stake". However, it is modified as follows: at least one share is conditionally delegated from the transferor node to at least two recipient nodes. The delegation is therefore subject to the release condition that the corresponding recipient node does not belong to the true subset after the selection.

[0016] Therefore, if a transferor node wants to support multiple different receiver nodes to become "candidates" for admission into a proper subset of nodes that undertake a particular task in the network, the transferor node does not have to choose only one of these receiver nodes. Instead, the transferor node can fully provide its support to all desired receiver nodes at the same time.

[0017] In this case, only the following restriction applies: when one of the receiver nodes is included in the proper subset, at most one and the same share that is conditionally delegated to multiple receiver nodes can be considered. The conditional delegation therefore does not allow for unlimited replication of contributions to the final result, i.e. the consensus on which nodes in the network perform a particular task.

[0018] In the illustrative example, it is assumed that there is a node A with a weight of 5, a node B with a weight of 10, and a node C with a weight of 11. Node A can now conditionally delegate, for example, a weight share of 4 to the two nodes B and C. Then, the following four situations may occur with respect to nodes B and C:

[0019] Neither nodes B nor C can enter the proper subset of nodes implementing a special task. The conditional assignment of weight shares 4 by the transferor A to the receiver nodes B and C will therefore become invalid, and A will enter the selection of the proper subset with its original weight 5.

[0020] Node B is admitted to the proper subset, but not node C. Node B increases the conditionally delegated weight share of 4; it therefore enters the selection of the proper subset with a total weight of 14. In contrast, the conditional assignment of the same weight share of 4 to node C becomes ineffective, so that node C remains at its original weight of 11. Node A drops to weight 1 in the selection.

[0021] Node C is admitted to the proper subset, but not node B. Here, the situation is reversed: Node C increases its weight share by 4, and is at a total weight of 15. Node B remains at its original weight of 10, and node A drops to a weight of 1.

[0022] Nodes B and C are both admitted to the proper subset. Based on the restrictions, in this case, it is only possible to either increase node B to a weight of 14 and keep node C at 11, or increase node C to a weight of 15 and keep node B at a weight of 10. In both cases, node A drops to a weight of 1.

[0023] This example shows that the calculation of the final selection and the final weights becomes significantly more complex compared to Delegated Proof of Stake. If on the one hand which nodes are admitted to the proper subset depends on the weights, and on the other hand this result subsequently changes the weights on which the result was based, this can be understood as a circular argument in the broadest sense.

[0024] However, it has been found that in exchange for this complexity, the tendency of the network to converge with respect to statistical and predictable results when the proper subsets are repeatedly reselected is significantly reduced. The permanent centralization of special tasks at a small number of nodes is prevented. This is very advantageous for the reliability and security of the network, especially in connection with blockchain management.

[0025] The trend towards centralization can be most easily understood using the example of the “proof of work” consensus.

[0026] Centralization trends may also emerge in the context of a Delegated Proof of Stake consensus if a sender node is promised a reward for delegating a share of its weight to a receiver node that results in that receiver node being admitted to the proper subset. The promise of such a reward creates an incentive to prioritize weight shares to receiver nodes that have the highest probability of actually being admitted to the proper subset and should be rewarded.

[0027] The possibility of conditionally delegating one and the same weight share to multiple recipient nodes arises in this context: The transferor node conditionally delegates the weight share to, for example:

[0028] ●It can not only be delegated to known established receiver nodes that are accepted into the true subset with a relatively high probability;

[0029] • It can also delegate to newer receiver nodes, which although have the prospect of being admitted into the proper subset are not certain.

[0030] The transferor node can only increase its probability of receiving a reward, because when the established receiver node is not accepted into the true subset against expectations, the use of the new receiver node opens up additional opportunities for its acceptance. On the other hand, if the established receiver node is accepted into the true subset as expected, conditional delegation to the new receiver node does not lead to a "loss" of weight share. The transferor node then also receives its reward from the established receiver node.

[0031] As a result, the barriers to the admission of new nodes into the proper subset are significantly reduced overall. As a result, with a higher probability, the composition of the proper subset of nodes that undertake special tasks in the network is a well-balanced mix and changes from time to time, both with regard to the operators of these nodes and also with regard to their geographical location. In this way, it is more difficult to predict which nodes will undertake special tasks at what time, which makes targeted manipulation attempts more difficult.

[0032] Thus, if it is difficult to predict which node will add the next block to the blockchain, this helps, for example, to ensure the integrity of the information stored in the blockchain. An attacker who wants to cause a node to add a block that does not comply with the code through targeted manipulation does not know which node he must attack for this purpose.

[0033] Typically, when operating a distributed network, any type of incentive is a driving force for agreeing on nodes to take over special tasks. For example, when performing a specific special task, an increased consumption of computing power, network bandwidth or energy may occur. The operator of the node can then, for example, make the degree of readiness for providing such resources dependent on the costs incurred being at least partially reimbursed. As explained above, the tendency of the reward system and false incentives to develop in the direction of network centralization is reduced by conditional delegation. Therefore, it is not necessary to avoid the use of the reward system when operating the network in order to avoid the self-reinforcing centralization of the network.

[0034] In a particularly advantageous embodiment, the transferor node selects the recipient node from a subset that includes the nodes that are being selected for the special task and / or have been selected for this purpose. In this way, it is avoided that expenditures in the form of time and / or computing power are incurred in selecting exactly those nodes that are not able, not intended or not allowed to take on the special task into the true subset of nodes that are responsible for the special task.

[0035] In a particularly advantageous embodiment, the selection of the proper subset includes nominating at least one candidate subset of nodes. It is ensured that the candidate subset is consistent with the utilization of the delegated share, wherein at most one and the same delegated share is considered for a receiving node in the candidate subset. This can be checked based on a predefined candidate subset. If the candidate subset does not pass this check, it is discarded. However, the candidate subset can also be generated at the outset with the following condition: at most one and the same delegated share is considered for one receiving node.

[0036] The candidate subsets are evaluated with evaluation scores according to a freely selectable optimal criterion, and the candidate subsets are selected as true subsets based on the evaluation scores.

[0037] For example, the candidate subset with the best evaluation score can be selected as the true subset. It is also possible to list candidate subsets with similar evaluation scores within a predetermined range, for example, in a candidate list. Each candidate subset from such a candidate list can then be selected as the final true subset, for example, with a probability weighted accordingly to its own evaluation score.

[0038] Therefore, from the possible candidate subsets, those that do not satisfy the following restriction are first excluded: ultimately only conditionally delegated weight shares can be considered for the receiving node that enters the true subset. Those candidate subsets that meet this condition are self-consistent at least with respect to the circular argument between being admitted to the true subset and subsequently clearing the conditionally delegated shares that do not work. For the in-depth evaluation of these candidate subsets, any optimal criteria can be used, wherein the specificity of the conditional delegation no longer needs to be taken into account.

[0039] The tasks to be performed within the scope of selecting a proper subset can be performed, for example, by a central authority in the network, but can also be distributed to the nodes in the network, for example, in an arbitrary manner. For example, for some applications of the network it can make sense for all nodes to accept a candidate subset. On the other hand, for other applications it may make sense for only the nodes that currently belong to the proper subset to accept a candidate subset. The calculation of the evaluation score and the selection of the candidate subset with the best evaluation score can be performed, for example, by other nodes and / or combinations of nodes.

[0040] In order to nominate candidate subsets based on the evaluation score, for example with the best evaluation score, and to select the candidate subset, any algorithm can be used in a situation-adaptive manner. In this way, for example, edge conditions that are predefined by the specific application of the network can be implemented. Thus, for example, a so-called "Minimal Agency Scheme" can be specified for safety-related applications in order to limit the maximum influence possibility of each individual node.

[0041] The program routines utilized in the process of selecting a proper subset can be implemented in particular in the form of so-called "Smart Contracts". These are programs that are executed, for example, on nodes of a blockchain (for example, nodes of the public Ethereum blockchain) and that, after their one-time installation, are publicly visible there but cannot be modified anymore. Even the creator of the "smart contract" cannot subsequently modify or delete it if this possibility is not already provided in the original program code of the "smart contract".

[0042] Alternatively or also in combination therewith, the program routine for the selection of a proper subset can be configured in such a way that at least deviations of the selection process from previously defined guidelines are detected. For example, one and the same task can be assigned to a plurality of nodes, and the rationality of the results can be mutually verified by these nodes or also by other nodes.

[0043] In a particularly advantageous embodiment, at least one candidate subset is determined based on the candidate distribution of the utilization of the delegated shares by applying a predetermined criterion for selecting a proper subset. If it is determined which transferors delegate which weighted shares unconditionally or conditionally to which receiving nodes, then for each weighted share conditionally delegated to multiple receiving nodes there is still the following degree of freedom: which receiving node ultimately benefits from this weighted share. The decision on all weighted shares for multiple delegations in this regard can form a candidate distribution for the utilization of the delegated shares.

[0044] The optimality criterion may include, for example, the sum of all weights under the assumption that the proper subset is selected corresponding to the candidate subset. As the proper subset, the candidate subset is then preferably selected which has the greatest support by all nodes in the network overall.

[0045] This optimal criterion may alternatively or in combination include, for example, the ratio of the sum of the shares unconditionally delegated to all nodes in the candidate subset to the sum of the shares unconditionally delegated to all nodes of the network. For example, a threshold value or other rated value may be specified for this ratio. In this way, it may be considered, for example, that in the unconditional delegation of weight shares from the transferor node to the recipient node, there is a stronger "willingness to support" of the transferor node than in the conditional delegation. In the case of conditional delegation, the transferor node still retains an alternative; whereas in the unconditional delegation, from the perspective of the transferor node, the recipient node has no alternative.

[0046] In another particularly advantageous embodiment, the optimality criterion comprises a comparison of the distribution of a predetermined characteristic for all nodes of the network with the distribution of said characteristic for all nodes of the candidate subset. In this case, the distribution can relate to any characteristic that can be meaningfully assigned to an individual node. For example, the characteristic can relate to the type of entity (e.g. a person, a company or a public institution) that operates the node. If, for example, for all nodes in the network there is a diversity in which nodes operated by individuals, companies and public institutions complement each other in a certain percentage of the mixture ratio, then this mixture ratio can also be transferred to the proper subset to be formed of nodes that undertake special tasks.

[0047] In another particularly advantageous configuration, a new candidate subset of nominating nodes is nominated until the evaluation score of the candidate subset meets a predetermined criterion or the predetermined limit of the cost for selecting a true subset is exhausted, depending on which of the following situations occurs first. The cost can be predetermined or measured, for example, in units of time and / or computing steps, or can also be predetermined or measured in units of any other quantitative measure. Because the candidate subset is first tested in view of the self-consistency and is only further evaluated in terms of content if it passes the test, it is difficult to foresee in advance how quickly the search for the best true subset will converge overall. By limiting the cost, it can be ensured that at the desired time point there is always a true subset of nodes that undertake special tasks available. Therefore, for example, when running a blockchain, there is often a pre-determination: at what period should new blocks be added to the blockchain, so as to ensure sufficient throughput, for example, for transactions to be stored in the blockchain.

[0048] In another particularly advantageous embodiment, the weight assigned to each node in the network corresponds to a quantitative share of at least one resource. The node has control over the quantitative share and guarantees the flawless execution of a specific task with the quantitative share. In this way, a transferor node that can conditionally or unconditionally delegate a weight share to a recipient node can, in particular, thereby demonstrate the trust it brings to the recipient node.

[0049] The resource can be managed by the network itself, for example. However, the resource can also be, for example, quite generally a resource in which the network can verify and / or control the node's right to control the resource. The network can control the resource, for example, with a "smart contract", wherein the network places the smart contract on the other blockchain, for example the Ethereum blockchain.

[0050] The guarantee of flawless execution can, for example, mean in particular: in response to determining that a node in the proper subset has not executed or has executed its own special task part defectively, completely or partially taking away the node and / or entrusting the node with a share to the transferor of the node with the right to use the following quantitative share of the resource, which corresponds to the weight of the node including the share entrusted to it as a receiving node.

[0051] In other words, if the conditions for securing the quantity share as a deposit are not met, the deposit should be paid. To secure the deposit, in particular, the following quantity share of the resource can be frozen, for example, corresponding to the weight of the node in the proper subset, including its delegated share as a receiving node.

[0052] In this connection, the resource used to store the weight share is preferably a resource in which each node itself has a strong interest. If, for example, transactions are stored in the blockchain, the resource can, for example, correspond to an amount.

[0053] The expression of the way in which a node makes the decision to accept a reward or follow an interest is only used for the purpose of better understanding based on the simulation (Analoga) in the human decision-making process. However, it should not be understood as follows: the nodes must be at least partially controlled by humans. The nodes in the network work independently, and the corresponding operators of these nodes can make general pre-determinations for them. It is precisely the independent and automated work of the nodes that leads to the self-reinforcing centralization effect that the method described here overcomes. This is very similar to stocks that are automatically traded at a high frequency in securities exchanges, where the joint action of nominally independent brokers leads to a self-reinforcing effect involving the system as a whole.

[0054] In another particularly advantageous embodiment, the special task includes at least a partial contribution to the following: implementing and / or coordinating changes to the following protocol, wherein the protocol is used for information exchange within the network and / or for information exchange with requesting client stations and / or for performing tasks. In this case, it also involves the following task, which is advantageously in the control of a restricted organization (Gremium) of nodes corresponding to a proper subset, because the simultaneous participation of all existing nodes is more hindering. The code can, for example, contain a plurality of principles on the "meta-level": how changes to the protocol in the network should be coordinated. The protocol can then, for example, include the following description: how these blocks of the blockchain look. The coordination process within the network can then, for example, be aimed at increasing the maximum block size, because the total amount of information (e.g., transactions) that should be stored in the block is high and repeatedly reaches a bottleneck with the latest block size. If coordination within the network can be performed, a possible “hard fork”, ie a split of a blockchain into two blockchains, can be avoided because there are no inconsistencies to be resolved between a certain group of participants.

[0055] In another particularly advantageous embodiment, the special task includes at least a partial contribution to the reselection of a proper subset of nodes. As explained above, it is advantageous in terms of the security and availability of the network to regularly change the nodes in the network that are responsible for the special task. This change therefore does not mean that the trust in the nodes that have previously been responsible for the special task is taken away. On the contrary, the new selection is made by precisely those nodes for which trust has previously been explicitly expressed.

[0056] Since the nodes are implemented as computers and the network is a collection of nodes, the method can in particular also be embodied as software running on the nodes. Therefore, the invention also relates to a computer program with machine-readable instructions which, when executed on one or more computers, cause the one or more computers to perform one of the described methods. In this sense, control devices for vehicles and embedded systems for technical devices which are also capable of executing machine-readable instructions are also to be regarded as computers.

[0057] The invention also relates to a machine-readable data carrier with a computer program and / or a download product. A download product is a digital product that can be transmitted via a data network, i.e. can be downloaded by a user of the data network, and can be sold for immediate downloading, for example in an online store.

[0058] Furthermore, a computer can be equipped with the computer program, the machine-readable data carrier or the download product. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Further measures for improving the invention are shown in more detail below together with the description of preferred exemplary embodiments of the invention based on the drawings.

[0060] Figure 1 An embodiment of a method 100 is shown;

[0061] Figure 2 A greatly simplified example of a network 1 is shown in which the method 100 can be used. DETAILED DESCRIPTION

[0062] Figure 1 is a flow chart of an embodiment of a method 100. The method 100 starts as follows: There is a network 1 having a large number of nodes 2. A proper subset 4 of these nodes 2 performs at least one special task that affects the operation of the network 1 as a whole.

[0063] In step 110, a weight 2c is assigned to each node 2. In step 120, each node 2 is provided with the possibility of conditionally or unconditionally delegating a share 2c' of its own weight 2c to a receiving node 2b as a transferor node 2a. In step 130, based on the weight 2c of each node 2 plus its share 2c' delegated as a receiving node 2b and minus its share 2c' delegated to other receiving nodes 2b as a transferor node 2a, a proper subset (4) of nodes 2 to which at least one special task is assigned is selected according to a predetermined criterion 4a.

[0064] Within this box 120 is depicted how the conditional delegation is performed. According to block 121, at least one share 2c' is conditionally delegated by the transferor node 2a to at least two receiver nodes 2b under a release condition, wherein the release condition is that after the selection 130, the corresponding receiver node 2b does not belong to the proper subset 4. In this case, an additional restriction 121a applies: in the case of the admission of one of these receiver nodes 2b into the proper subset 4, at most one and the same share 2c' that can be conditionally delegated to multiple receiver nodes 2b can be considered.

[0065] Within this box 130 , exemplary criteria 4 a are shown according to which a proper subset 4 can be selected based on the weights 2 c and the delegated shares 2 c ′.

[0066] According to block 131, at least one candidate subset 4b of the node 2 is nominated. According to block 132, it has been ensured when nominating the candidate subset or by subsequent verification that the candidate subset 4b is consistent with the utilization of the delegated share 2c', wherein at most one and the same share 2c' is considered for the receiving node 2b in the candidate subset 4b. If it is determined during the verification that this is not the case (true value 0), the restriction 121a is not met and the candidate subset 4b cannot be used. Branch back to the nomination 131 of the new candidate subset 4b. On the contrary, if the restriction 121a is met (true value 1 in block 132), the candidate subset 4b is evaluated with the evaluation score according to the best criterion 133a.

[0067] If the evaluation score 133b satisfies the criterion 135 (true value 1), the candidate subset 4b is determined as the true subset 4 of the nodes 2 that perform the special task in the network 1 according to block 134. On the contrary, if the evaluation score 133b does not satisfy the criterion 135 (true value 0), the branch returns to the nomination 131 of the new candidate subset 4b.

[0068] However, a check is performed here beforehand according to block 136 to determine whether the predefined limit for the expenditure for selecting the proper subset 4 has been exhausted. If this is the case (true value 1), a branch is made forward to block 134 and the candidate subset 4b with the best evaluation score 133b up to that point in time is determined as the proper subset 4. The check according to block 136 is also performed when branching back from block 132 to block 131.

[0069] According to step 140, a quantitative share of resources of the network 1 can be kept frozen and thus secured as a deposit for performing at least one special task in accordance with regulations, wherein the quantitative share of resources corresponds to the weight 2c of the nodes 2 in the proper subset 4, including the shares 2c' delegated to these nodes 2 as receiving nodes 2b.

[0070] According to step 150, in response to determining 150 that there is a node 2 in the true subset 4 that has not performed or has defectively performed its own special task part, the quantitative share of the following resources can be invalidated, wherein the quantitative share of the resources corresponds to the weight 2c of the node 2, which includes the share 2c' delegated to the node 2 as the receiving node 2b.

[0071] Figure 2 The diagram shows a greatly simplified example of a network 1 in which the method 100 can be used. The nodes 2 exchange information with each other and with the client stations 3 along a bidirectional path drawn with dashed lines. The network 1 as a whole provides at least one service for the client stations 3 .

[0072] A weight 2c is assigned to each of these nodes 2. Some nodes 2 act as a transferor node 2a and delegate a share 2c' of their weight 2c to a receiver node 2b, wherein one and the same share 2c' can be delegated by the transferor node 2a to multiple receiver nodes 2b. Based on the weight 2c including the delegated share 2c', a proper subset 4 of nodes 2 is selected for performing a special task in the network 1.

[0073] The recipient node 2b may for example be selected from a subset 5 of nodes 2 of the network that are specifically selected to perform a special task.

Claims

1. A method (100) for operating a network (1) consisting of a plurality of computer-implemented nodes (2), wherein the nodes distribute information among each other in the network (1) and exchange information with requesting client stations (3), and wherein a proper subset (4) of the nodes (2) performs at least one special task, which affects the operation of the network (1) as a whole, wherein the method comprises the steps of: =In the network (1), each node (2) is assigned (110) a weight (2c), wherein each node (2) has the possibility of, as a transferor node (2a), conditionally or unconditionally delegating (120) a share (2c') of its own weight (2c) to other receiver nodes (2b); ● based on the weight (2c) of each node (2) plus the share (2c') entrusted to the node as a receiver node (2b) and minus the share (2c') entrusted to other receiver nodes (2b) by the node as a lender node (2a), selecting (130) the proper subset (4) of the nodes (2) to which at least one special task is assigned according to at least one predetermined criterion (4a); The transferor node (2a) transfers at least one share (2c') ● conditionally delegating (121) to at least two receiver nodes (2b) under a release condition, wherein the release condition is that the corresponding receiver node (2b) does not belong to the proper subset (4) after the selection (130), It has the additional restriction (121a): when admitting one of the receiver nodes (2b) into the proper subset (4), at most one and the same share (2c') conditionally delegated to one of the multiple receiver nodes (2b) can be taken into account.

2. The method (100) according to claim 1, wherein the transferor node (2a) selects the receiver node (2b) from a subset (5) comprising nodes (2) that are nominated for a special task and / or nodes (2) that have been nominated for this task.

3. The method (100) of claim 1, wherein the selecting (130) of the proper subset (4) comprises: ● nominating (131) at least one candidate subset (4b) of the node (2); = ensuring (132) that the candidate subset (4b) is consistent with the utilization of the delegated shares (2c'), wherein at most one and the same delegated share (2c') is considered for the receiver nodes (2b) in the candidate subset (4b); = evaluating (133) the candidate subset (4b) according to the optimality criterion (133a) with an evaluation score (133b); and - Based on the evaluation scores (133b) selecting (134) the candidate subset (4b) as the proper subset (4).

4. The method (100) according to claim 3, wherein the optimal criterion (133a) includes: a prerequisite for selecting the candidate subset (4b) as the proper subset (4) is based on the sum of all weights (2c).

5. A method (100) according to claim 3, wherein the optimal criterion (133a) comprises: the ratio of the sum of the shares (2c') unconditionally delegated to all nodes (2) in the candidate subset (4b) to the sum of the shares (2c') unconditionally delegated to all nodes (2) in the network (1).

6. A method (100) according to claim 3, wherein the optimal criterion (133a) comprises: a comparison of the distribution (6a) of a predetermined characteristic (6) of all nodes (2) of the network (1) with the distribution (6b) of the predetermined characteristic (6) of all nodes (2) of the candidate subset (4b).

7. A method (100) according to claim 3, wherein a new candidate subset (4b) of the node (2) is nominated (131), depending on which of the following two situations occurs first: until the evaluation score (133b) of the candidate subset meets a predetermined criterion (4a) or until a predetermined limit (136) on the cost measured in time, computing steps or other quantitative metrics for selecting the true subset (4) is exhausted.

8. The method (100) according to claim 3, wherein at least one candidate subset (4b) is determined based on the candidate distribution using the delegated shares (2c') by applying the predefined criterion (4a).

9. A method (100) according to any one of claims 1 to 8, wherein the weight (2c) assigned (110) to each node (2) in the network (1) corresponds to the following quantitative share of at least one resource, the node (2) has control over the quantitative share and the node (2) guarantees the defect-free execution of the special task with the quantitative share.

10. The method (100) according to claim 9, wherein the following quantitative share of the resource is kept frozen (140), which corresponds to the weight (2c) of the node (2) in the proper subset (4), including the share (2c') entrusted to the node (2) as a receiving node (2b).

11. The method (100) according to claim 9, wherein in response to determining (150) that there is a node (2) in the true subset (4) that has not executed or has defectively executed its own special task part, the right to use the following quantitative share of the resource by the node (2) and / or the transferor node (2a) that has delegated the share to the node (2, 2b) is completely or partially taken away (160), wherein the quantitative share corresponds to the weight (2c) of the node (2) including the share (2c') delegated to the node (2) as the receiving node (2b).

12. A method (100) according to any one of claims 1 to 8, wherein the special task comprises at least a partial contribution to: creating a new block to be added to the blockchain managed by the network (1); and / or verifying the correctness of the information contained in such a block with respect to the rules predetermined for the blockchain, and / or adding the block to the blockchain in the event of a positive outcome of the verification.

13. The method (100) of claim 12, wherein the information in the new block to be added to the blockchain comprises: trade; and / or at least one fact to be authenticated.

14. A method (100) according to any one of claims 1 to 8, wherein the special task comprises at least one partial contribution to: implementing and / or coordinating at least one change of a protocol and / or at least one change of other rules predetermined for the network (1), wherein the protocol is used for information exchange within the network (1) and / or for information exchange with a requesting client station (3) and / or for performing a task.

15. The method (100) according to any one of claims 1 to 8, wherein the special task comprises at least one partial contribution to reselecting the proper subset (4) of the nodes (2).

16. A computer program product comprising machine-readable instructions which, when executed on one or more computers, cause the one or more computers to perform the method (100) according to any one of claims 1 to 15.

17. Machine-readable data carrier and / or download product, wherein the machine-readable data carrier and / or download product comprises a computer program product according to claim 16.

18. Computer equipped with a computer program product according to claim 16 and / or a machine-readable data carrier and / or a download product according to claim 17.

Images