Adaptive Sharded Blockchain Network Management via Deep Q-Network

Solution Overview

Problem

Current blockchain systems face scalability issues due to low transaction processing speed (TPS) and energy consumption, particularly in IoT networks, where the Proof of Work (PoW) algorithm is inefficient, and sharding techniques compromise security for improved TPS.

Innovation Solution

An apparatus and method using deep reinforcement learning to manage a sharded blockchain network by optimizing block size, block interval, and number of shards based on real-time network conditions, ensuring security and maximizing TPS while adhering to latency and security constraints.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If PoW algorithm is used for consensus, then security is maintained, but transaction processing speed (TPS) is low and energy consumption is high

Engineering Contradiction:
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent divides the blockchain network into multiple shards (segments), allowing parallel transaction processing across different shard groups. This segmentation enables the system to maintain security through distributed consensus while significantly improving TPS by processing transactions in parallel rather than sequentially across the entire network.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic adjustment of network parameters including block size, block interval, and number of shards based on real-time network conditions. This dynamic approach allows the system to optimize between security and TPS by adapting the consensus parameters and shard configuration according to current transaction loads and network state.

Inventive Principle:
Principle #15Dynamics

2Productivity

If sharding technique is used to improve TPS, then transaction processing speed increases, but security level is reduced

Engineering Contradiction:
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent segments the network into multiple shard groups where each group processes transactions independently in parallel. This segmentation improves TPS by distributing the transaction load across multiple groups while maintaining security through the distributed nature of the segmentation, preventing any single point of failure or malicious control.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent dynamically changes key parameters including the number of shards, block size, and block interval based on network conditions and security requirements. By adjusting these parameters, the system can optimize TPS while maintaining security thresholds, resolving the contradiction between sharding benefits and security concerns.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If block size is increased to improve TPS, then transaction processing capacity increases, but latency increases

Engineering Contradiction:
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent implements dynamic adjustment of block size based on real-time network conditions and transaction demands. Rather than using a fixed large block size that would increase latency, the system adapts block size dynamically to optimize between processing capacity and confirmation time, reducing latency while maintaining high TPS.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent uses periodic block generation with adjustable intervals to balance processing capacity and latency. By optimizing the block interval parameter dynamically, the system ensures that blocks are generated frequently enough to maintain low latency while containing sufficient transactions to achieve high processing capacity.

Inventive Principle:
Principle #19Periodic action

Data Source

PatentUS11687562B2Apparatus and method for adaptively managing sharded blockchain network based on deep Q network
Publication Date: 2023.06.27
  • US11687562B2 patent drawing
  • US11687562B2 patent drawing
  • US11687562B2 patent drawing

AI summary

An apparatus for managing sharded blockchain network obtains state information including an inter-node data transmission rate and computing resources of each node in the current epoch of a shard-based blockchain network in which N nodes are distributed to at least one shard, a consensus process record of each node in each shard and a ratio of malicious nodes, estimates at least one action each including block size, block interval and number of shards corresponding to the state information by using an artificial neural network in which a pattern estimation method has been previously learned, selects one action among the estimated at least one action and applies the one action to the shard-based blockchain network.