TBDD: A New Trust-based, DRL-driven Framework for Blockchain Sharding in IoT

TL;DR

TbDd tackles the challenge of secure, scalable blockchain sharding in IoT by integrating a trust-driven evaluation mechanism with a DRL-based shard-allocation framework. It introduces a Block Verification Table and Local/Global Trust Tables to compute dynamic trust, and a Shard Risk Evaluator to trigger resharding under collusion risk. The DRL agent (the TbDd Committee) optimizes shard allocations to maximize a multi-component reward, balancing shard load, CSTs, and trust variance while respecting intra-shard fault tolerance. Experimental results show TbDd achieving higher throughput and lower shard corruption than random, community, and traditional trust-based sharding under simulated collusion, indicating practical benefits for secure IoT deployments.

Abstract

Integrating sharded blockchain with IoT presents a solution for trust issues and optimized data flow. Sharding boosts blockchain scalability by dividing its nodes into parallel shards, yet it's vulnerable to the $1\%$ attacks where dishonest nodes target a shard to corrupt the entire blockchain. Balancing security with scalability is pivotal for such systems. Deep Reinforcement Learning (DRL) adeptly handles dynamic, complex systems and multi-dimensional optimization. This paper introduces a Trust-based and DRL-driven (\textsc{TbDd}) framework, crafted to counter shard collusion risks and dynamically adjust node allocation, enhancing throughput while maintaining network security. With a comprehensive trust evaluation mechanism, \textsc{TbDd} discerns node types and performs targeted resharding against potential threats. The model maximizes tolerance for dishonest nodes, optimizes node movement frequency, ensures even node distribution in shards, and balances sharding risks. Rigorous evaluations prove \textsc{TbDd}'s superiority over conventional random-, community-, and trust-based sharding methods in shard risk equilibrium and reducing cross-shard transactions.

Figures (9)

• Figure 1: System model.
• Figure 2: The proposed blockchain sharding system architecture - TbDd.
• Figure 3: The proposed blockchain sharding system flowchart - TbDd, including four steps. ① Trust table updated: update the Block Verification Table (BVT) and Global Trust Table (GTT), checking whether triggering Algo. \ref{['alg:1']}. ② Train the DRL algorithm: use the DRL-based model to virtually resharding through several epochs and output a new node allocation result. ③ Update shard allocation: allocate nodes to the shards according to the DRL-based training result. ④ Monitor network performance: monitor and step into retraining.
• Figure 4: The flow diagram of the proposed computing trust score system comprises BVT, LTT and GTT.
• Figure 5: Fig. \ref{['fig.5a']} represents the comparison among Random-based, Community-based, Trust-based, TBDD-DQN and TBDD-PPO across $100$ epochs. Figs. \ref{['fig.5b']} and \ref{['fig.5c']} represent the reward performance between TBDD-DQN and TBDD-PPO across varying node numbers in the environment settings across $30$ epochs, respectively.