An IOTA node system architecture and cost optimization algorithm

Abstract

The present invention relates to the field of Internet of Things blockchain technology, and discloses an IOTA node system architecture and cost optimization algorithm. The node system architecture includes a client and at least one IOTA node, and an IOTA node includes a Tip selection module, a PoW module with a built-in CO-IOTA cost optimization module, and a local Tangle module; in time slot t, the IOTA node transmits the updated local Tangle to the CO-IOTA cost optimization module and then executes the cost optimization algorithm to obtain the optimal message confirmation rate of the next time slot. The present invention decouples multi-objective optimization problems with low computational complexity, evaluates key factors using relevant optimization theories such as Lyapunov optimization theory, embeds queues and constraints related to messages into the Lyapunov optimization framework, and can determine the threshold of the IOTA node message confirmation rate when any information is unknown, and timely adjusts the difficulty value and message confirmation rate in the difficulty of the PoW module to effectively reduce the IOTA node cost, so as to solve the optimization problem of long-term complex cost.

Description

Technical Field

[0001] The present invention relates to the field of Internet of Things blockchain technology, and in particular to an IOTA node system architecture and a cost optimization algorithm. Background Art

[0002] As the pace of digital transformation and intelligent upgrading of the economy and society accelerates, the Internet of Things (IoT) has become an important part of the country's new infrastructure, but its centralized services are extremely vulnerable to malicious attacks, which can seriously cause a large amount of information theft and damage. In addition, the generation of massive IoT information makes tracking more difficult, and the emergence of blockchain technology has brought new opportunities for the further development of the Internet of Things.

[0003] Blockchain technology uses a distributed accounting method and has the advantages of decentralization, tamper-proofing, and traceability. Since the Internet of Things has high requirements for concurrency, scalability, and message fees, blockchain based on a single-chain architecture can no longer meet its actual needs. Blockchain technology based on a directed acyclic graph (DAG) structure has gradually attracted attention from all parties and has become an effective alternative to single-chain architecture blockchain in the Internet of Things.

[0004] As blockchain technology based on DAG structure continues to mature, the more successful ones include IOTA, Obyte (formerly known as Byteball), Nano (formerly known as RaiBlocks) and Conflux. Compared with the above three, IOTA is a revolutionary new public chain system designed specifically for the Internet of Things. The distributed ledger is called Tangle, which supports high concurrency, high scalability and zero transaction fees; the basic unit of Tangle is message, and one message constitutes a block; the unconfirmed terminal message in Tangle is called Tip.

[0005] Compared with blockchain technology based on a single-chain architecture, IOTA's work efficiency has been greatly improved. IOTA abandons the consensus mechanism and incentive mechanism of traditional blockchains, and IOTA nodes (i.e., IOTA full nodes) have equal status. Since the IOTA system does not charge message fees, the large amount of costs incurred in managing and maintaining the system need to be borne by all IOTA nodes.

[0006] However, as the scale of the Internet of Things continues to grow, the number of IOTA nodes has increased accordingly, and its cost optimization problem has become increasingly prominent; therefore, how to effectively reduce the cost of IOTA nodes has become a problem that needs to be solved urgently. Summary of the invention

[0007] The purpose of this invention is to provide an IOTA node system architecture and cost optimization algorithm to solve the following technical problems:

[0008] How to effectively reduce IOTA node costs.

[0009] The purpose of the present invention can be achieved through the following technical solutions:

[0010] An IOTA node system architecture, the node system architecture includes a client and at least one IOTA node, for one of the IOTA nodes, the IOTA node includes:

[0011] A Tip selection module, configured to select a matching Tip result and send it back to the client after receiving a Tip request sent by the client;

[0012] A PoW module, used for performing PoW calculation after receiving the puzzle request sent by the client, and sending the obtained Nonce value back to the client; the PoW module has a built-in CO-IOTA cost optimization module;

[0013] The local Tangle module is used to receive the packaged message sent by the client and perform validity check; if the message is valid, the local Tangle is updated and broadcast to the entire network;

[0014] For one of the IOTA nodes, at time slot t, the IOTA node transmits the updated local Tangle to the CO-IOTA cost optimization module and then executes the cost optimization algorithm to obtain the optimal message confirmation rate for the next time slot.

[0015] As a further solution of the present invention: an IOTA node system cost optimization algorithm, the cost optimization algorithm includes:

[0016] In each time slot t among the d time slots, for the i-th IOTA node among the n IOTA nodes, obtain the corresponding number of confirmation messages and number of messages ;

[0017] in, , the number of messages represents the number of new messages arriving at the i-th IOTA node at time slot t, the number of confirmed messages represents the number of messages confirmed by the i-th IOTA node at time slot t;

[0018] Calculate the message queue backlog based on the queue backlog evolution equation ;

[0019] Calculate the cost consumed by the i-th IOTA node in time slot t according to the consumption cost calculation formula ;

[0020] Calculate the message cumulative queue backlog based on the backlog formula and message delays accumulate backlog ;

[0021] Calculate the Lyapunov penalty drift according to the penalty drift calculation formula ;

[0022] Determine the message confirmation rate threshold ;

[0023] The message confirmation rate threshold Store in an array, and update the optimal message confirmation rate of the i-th IOTA node according to the array.

[0024] As a further solution of the present invention: the queue backlog evolution equation includes:

[0025]

[0026] in, represents the backlog queue of the t-time slot cache in the i-th IOTA node, , .

[0027] As a further solution of the present invention: the consumption cost calculation formula includes:

[0028]

[0029]

[0030]

[0031] in, is the average cost of confirming two Tips, b is the average cost of verifying a message, c is the average cost of updating the local Tangle once, and f is the average cost of calculating a Nonce value. The average cost of broadcasting a Tangle.

[0032] As a further solution of the present invention: the backlog formula includes:

[0033]

[0034] in, , , represents the average time it takes for a message to reach the i-th IOTA node and be confirmed, represents the cumulative backlog of messages that cannot meet the latency requirement of the i-th IOTA node in time slot t.

[0035] As a further solution of the present invention: the penalty drift calculation formula includes:

[0036]

[0037] in, , is a non-negative controllable parameter describing the relative weight between queue stability and minimization of IOTA node cost, is the queue backlog vector for time slot t.

[0038] As a further solution of the present invention: the message confirmation rate threshold is determined The methods include:

[0039]

[0040] in, The maximum delay in confirming each message.

[0041] Beneficial effects of the present invention: The present invention decouples multi-objective optimization problems with low computational complexity, evaluates key factors using relevant optimization theories such as Lyapunov optimization theory, makes decisions when any information is unknown, and determines the threshold of the IOTA node message confirmation rate by embedding message-related queues and constraints into the Lyapunov optimization framework. The difficulty value and message confirmation rate in the difficulty of the PoW module can be adjusted in time to effectively reduce the IOTA node cost, so as to solve the long-term optimization problem of complex cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will be further described below in conjunction with the accompanying drawings.

[0043] Figure 1 This is an IOTA node system architecture diagram including the CO-IOTA cost optimization mechanism of the present invention;

[0044] Figure 2 The changes of the message backlog queue of the IOTA node before and after optimization in the present invention;

[0045] Figure 3 This is the change of the node cost of the IOTA node in the present invention before and after optimization. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] See also Figure 1 As shown, the present invention is an IOTA node system architecture, and the node system architecture includes a client and at least one IOTA node.

[0048] In the IOTA node system, the client is mainly responsible for creating and packaging messages, while the IOTA node is responsible for verifying and publishing messages. For an IOTA node, it includes a Tip selection module, a PoW module with a built-in CO-IOTA cost optimization module, and a local Tangle module;

[0049] The workflow of the IOTA node system is as follows:

[0050] After receiving a message request, the client starts to create the message;

[0051] The client sends the Tip request to the Tip selection module. After receiving the request, the Tip selection module runs the Tip selection algorithm, selects the appropriate Tip (usually 2, but it can also be determined according to the specific situation), and then sends the Tip result back to the client.

[0052] The client sends a request to the IOTA node again to solve the difficulty value calculation problem. The core is to perform a lightweight proof of work (PoW) operation through the PoW module, find a random number (Nonce value) and return it to the client;

[0053] Finally, the client submits the packaged message to the IOTA node for validity check; if the message is valid, the IOTA node updates the local Tangle and broadcasts it to the entire network. After other nodes receive the Tangle with the new message, the consensus mechanism is used to keep the ledger of the entire network unified.

[0054] The local Tangle module is used to receive the packaged message sent by the client and perform validity check; if the message is valid, the local Tangle module updates the local Tangle and broadcasts it to the entire network;

[0055] As part of the IOTA consensus mechanism, CO-IOTA must run on each IOTA node. Initialize the IOTA node system. Assuming that the system is stable at time slot t, each IOTA node starts the CO-IOTA mechanism for the first time.

[0056] Taking an IOTA node in the above IOTA node system as an example, the operating principle of the CO-IOTA mechanism is explained:

[0057] At time slot t, the IOTA node transmits the updated local Tangle to the CO-IOTA cost optimization module and executes the cost optimization algorithm. The message confirmation rate threshold is obtained by using the cost optimization algorithm, and then it is transmitted to the PoW module for difficulty adjustment. The optimal message confirmation rate for the next time slot is obtained and adjusted in the next time slot.

[0058] In this embodiment of the present invention, an IOTA node system cost optimization algorithm is disclosed, and the cost optimization algorithm is explained as follows:

[0059] In the IOTA node system, the rates of IOTA node publishing and confirming messages follow the Poisson distribution. In this implementation, there are n IOTA nodes in the IOTA node system, represented by the number i. . IOTA nodes in time intervals The time interval is divided into d time slots of equal length, each time slot is represented by t. .

[0060] For each IOTA node, we need to focus on the difference between the number of newly arrived messages in each time slot t and the number of confirmed messages of this node. At the beginning of the cost optimization algorithm, for the i-th IOTA node among the n IOTA nodes, we need to obtain the corresponding number of confirmed messages. and number of messages ; and All satisfy independent and identical distribution; the number of messages Indicates the number of new messages arriving at the i-th IOTA node in time slot t. After the new message arrives, it is first cached in the backlog queue of the i-th IOTA node, waiting to be confirmed; the number of confirmed messages Represents the number of messages confirmed by the i-th IOTA node at time slot t;.

[0061] Calculate the message queue backlog of the i-th IOTA node according to the queue backlog evolution equation ;

[0062]

[0063] in, represents the number of messages in the backlog queue cached in the ith IOTA node at time slot t, , .

[0064] The time factor needs to be considered in the queuing process to meet the delay requirements. According to Little's law, the average confirmation delay constraint can be expressed as:

[0065]

[0066] in, The maximum delay confirmation time for each message, which is determined by the IOTA Committee (in the IOTA system, the introduction of the IOTA Committee can strengthen the supervision and management of the IOTA system, reduce the overall IOTA node calculation amount, and reduce unnecessary resource consumption) according to the system situation. Without considering the hardware cost and the loss of each node itself, it can be defined as is the average cost of confirming two Tips, b is the average cost of verifying a message, c is the average cost of updating the local Tangle once, and f is the average cost of calculating a Nonce value. is the average cost of broadcasting a Tangle, then the cost consumed by the i-th IOTA node in time slot t can be calculated according to the consumption cost calculation formula ;

[0067]

[0068]

[0069]

[0070] Due to the particularity of the IOTA structure and network delay, the message confirmations in each time slot are entangled with each other, so the technical problem of the present invention cannot be solved by simple combinatorial optimization. Considering that the stability of the entire IOTA system is conducive to the steady growth of Tangle, the technical problem can be well solved by Lyapunov optimization theory.

[0071] According to the prior art, Non-negative and satisfies It has strong stability and can calculate the message accumulation queue backlog according to the backlog formula and message delays accumulate backlog ;

[0072] The backlog formula includes:

[0073]

[0074] in, , , represents the average time it takes for a message to reach the i-th IOTA node and be confirmed, represents the cumulative backlog of messages that cannot meet the latency requirement of the i-th IOTA node in time slot t.

[0075] and Same, Satisfy its delay requirement constraint formula , and is stable. Then the evolution of the queue is analyzed through the Lyapunov optimization framework.

[0076] set up The Lyapunov function in the CO-IOTA mechanism is expressed as:

[0077]

[0078] in, is the queue backlog vector for time slot t, is the Lyapunov function.

[0079] Note that if you By taking a sufficiently small value, the backlog of each queue can be kept at a low level to meet the stability of the queue. Therefore, it is necessary to introduce the Lyapunov drift under the condition of a single time slot to achieve this goal. The increase in the backlog of all queues from time slot t to time slot t+1 can be expressed by the following formula:

[0080]

[0081] This drift can be understood as the expected change of the Lyapunov function in a time slot. Since new messages may be lost or invalid during network transmission, by introducing and adjusting The value is used to minimize drift and cost, and find a trade-off between queue backlog and node cost. In this way, it can be explicitly derived from a mathematical point of view , and finally determine the message confirmation rate threshold ;

[0082] Calculate the Lyapunov penalty drift according to the penalty drift calculation formula ;

[0083]

[0084] in, , is a non-negative controllable parameter that describes the relative weight between queue stability and minimization of IOTA node cost, ensuring the stability of the actual and virtual queues while minimizing the total cost.

[0085]

[0086]

[0087]

[0088]

[0089] Lyapunov's drift penalty satisfy:

[0090]

[0091] in, .

[0092] Finally, we can get:

[0093]

[0094] Afterwards, the message confirmation rate threshold Store in an array, and update the optimal message confirmation rate of the i-th IOTA node according to the array. In this embodiment, the optimal message confirmation rate should be close to or equal to Only in this case can the difficulty value in PoW be adjusted in time and rate control be performed to effectively reduce the cost of IOTA nodes.

[0095] In this embodiment of the present invention, this paper uses a simulation method consistent with the current mainstream method to verify the effectiveness of the proposed mechanism in optimizing IOTA node costs. In MATLAB R2022a, the change trend of the message backlog queue in the node before and after optimization is first compared and analyzed, and then the change of node cost before and after optimization is compared.

[0096] Assume there are n IOTA nodes in the IOTA network. The parameters in the experiment are set as follows: , , , , , . Not optimized Queues and Queue, optimized Queues and Queue Figure 2 As shown in , the optimized queue will tend to be stable over time, always fluctuating around a fixed value. Figure 2 As shown in (d), the fluctuation range of the accumulated backlog of new messages is significantly reduced after optimization.

[0097] Under the same conditions as above, three IOTA nodes were randomly selected for cost analysis. Figure 3 As shown, the optimized cost of the nodes (solid line) is lower than the unoptimized cost (dashed line).

[0098] In summary, it can be clearly seen from the above experimental verification that the cost optimization algorithm in the CO-IOTA mechanism of the present invention reduces the computing cost of the IOTA node, can play a role in collecting IoT information, and effectively saves costs.

[0099] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An IOTA node system architecture, comprising a client and at least one IOTA node, characterized in that: For one of the IOTA nodes, the IOTA node includes: A Tip selection module, configured to select a matching Tip result and send it back to the client after receiving a Tip request sent by the client; A PoW module, used for performing PoW calculation after receiving the puzzle request sent by the client, and sending the obtained Nonce value back to the client; the PoW module has a built-in CO-IOTA cost optimization module; The local Tangle module is used to receive the packaged message sent by the client and perform validity check; if the message is valid, the local Tangle is updated and broadcast to the entire network; For one of the IOTA nodes, at time slot t, the IOTA node transmits the updated local Tangle to the CO-IOTA cost optimization module and then executes the cost optimization algorithm to obtain the optimal message confirmation rate for the next time slot.

2. An IOTA node system cost optimization algorithm, applied to an IOTA node system architecture as claimed in claim 1, characterized in that: The cost optimization algorithm includes: In each time slot t among the d time slots, for the i-th IOTA node among the n IOTA nodes, obtain the corresponding number of confirmation messages c i (t) and the number of messages r i (t); Where, t∈{1,2,...,d}, the number of messages r i (t) represents the number of new messages arriving at the i-th IOTA node at time slot t, and the number of confirmed messages c i (t) represents the number of messages confirmed by the i-th IOTA node at time slot t; Calculate the message queue backlog ε according to the queue backlog evolution equation i (t+1); The cost consumed by the i-th IOTA node in time slot t is calculated according to the consumption cost calculation formula i (t); Calculate the message queue backlog F according to the backlog formula i (t) and the message delay accumulated backlog F i (t+1); Calculate the Lyapunov penalty drift DD(I(t)) according to the penalty drift calculation formula; Determine the message confirmation rate threshold The message confirmation rate threshold Store in an array, and update the optimal message confirmation rate of the i-th IOTA node according to the array.

3. The IOTA node system cost optimization algorithm according to claim 2, characterized in that: The queue backlog evolution equation includes: Among them, ε i (t) represents the backlog queue cached in the ith IOTA node at time slot t, 4. The IOTA node system cost optimization algorithm according to claim 3, characterized in that: The consumption cost calculation formula includes: Cost i (t)=(a+c+f+g)*c i (t)+b*y i (t) y i (t)=c(t)-c i (t) Among them, a is the average cost of confirming two Tips, b is the average cost of verifying a message, c is the average cost of updating a local Tangle, f is the average cost of calculating a Nonce value, and g is the average cost of broadcasting a Tangle.

5. The IOTA node system cost optimization algorithm according to claim 2, characterized in that: The backlog formula includes: in, Time i (t) represents the average time it takes for a message to reach the i-th IOTA node and be confirmed, F i (t) represents the accumulated queue backlog of messages that cannot meet the delay requirement of the i-th IOTA node in time slot t, ε i (t) represents the backlog queue cached in the ith IOTA node at time slot t, ε i (0)=0.

6. The IOTA node system cost optimization algorithm according to claim 4, characterized in that: The penalty drift calculation formula includes: Among them, V>0 is a non-negative controllable parameter used to describe the relative weight between queue stability and IOTA node cost minimization, and I(t) is the queue backlog vector at time slot t.

7. The IOTA node system cost optimization algorithm according to claim 6, characterized in that: Determining the message confirmation rate threshold The methods include: Where H is the maximum delayed confirmation time of each message.

Images