Abstract
Blockchain technology is becoming the cornerstone for the development and deployment of other technologies like Federated Learning (FL) and the Internet of Things (IoT), as it plays a critical role in data sharing and incentives. Blockchains supports decentralization, data-privacy protection, security, and reliability. Assuring secure data sharing in mobile computing and FL is challenging because of untrustworthy participants and unknown data quality. Blockchain provides trust in decentralized environments without requiring trusted third parties. By using smart contracts, blockchain has been able to supporting rich decentralized applications. However, the scalability of blockchain is a challenge that prevents its wide adoption by high-performance applications. To address the blockchain scalability issue, various blockchain sharding technologies and off-chain solutions have been proposed. To improve the network throughput, blockchain sharding divides the entire network into several smaller parallel groups and exploits fast consensus algorithms in blockchain shards. Off-chain solutions, such as payment channel networks (PCNs), transfer the slow on-chain transactions to the off-chain environment, in which transactions can be accelerated. Without consensus and on-chain expensive operations, off-chain scalable solutions significantly reduce transaction costs and increase transaction throughput. This special issue aims to provide a forum for the presentation of state-of-the-art research approaches that advance the construction of intelligent blockchain systems. A total of 27 articles were accepted after a two-round rigorous review process. Based on their topics, we have grouped the accepted articles into four categories: blockchain-based federated learning systems, blockchain and the IoT, blockchain scalability, and high-performance blockchains. In what follows, we introduce these articles and their contributions.
Introduction
This special issue of the IEEE Journal on Selected Areas in Communications presents a comprehensive collection of 27 research articles that explore the intersection of blockchain technology and future communication networks. As the digital landscape evolves toward hyper-connectivity, blockchain is emerging as a "cornerstone" for the deployment of Federated Learning (FL) and the Internet of Things (IoT). By providing a decentralized framework for trust, security, and data privacy, blockchain addresses the inherent challenges of untrustworthy participants and variable data quality in distributed systems.
Addressing the Trust and Scalability Gap
The transition to decentralized machine learning and massive IoT deployments faces a significant "trust gap." In FL, verifying the quality of data provided by participants without a central authority is a major hurdle. In IoT, managing secure data sharing and task offloading across heterogeneous edge devices is equally complex.
While blockchain offers a solution through smart contracts and decentralization, it is often hampered by the "scalability trilemma." Traditional blockchain architectures can be too slow and resource-intensive for the high-performance requirements of modern communication networks. This special issue focuses on "Intelligent Blockchain"—the dual pursuit of using blockchain to secure communications and applying intelligent algorithms to enhance the performance and scalability of the blockchain itself.
Thematic Research Pillars
The research presented in this issue is categorized into four primary research areas:
1. Blockchain-based Federated Learning Systems
Research in this area explores the integration of blockchain to secure FL processes. Key innovations include "energy-recycling" consensus mechanisms, where the computational effort typically wasted on Proof-of-Work (PoW) puzzles is redirected to perform actual machine learning training tasks. Additionally, multi-layer Stackelberg game models are proposed to optimize the balance between system overhead and economic incentives for participants.
2. Blockchain and the Internet of Things (IoT)
This pillar focuses on secure data collection and resource management. Contributions include service-oriented edge resource allocation frameworks and the use of Unmanned Aerial Vehicles (UAVs) as mobile blockchain nodes. These technologies enable secure, decentralized data sharing in remote or highly dynamic environments, supported by efficient encryption schemes that guarantee flexible data authorization.
3. Blockchain Scalability
To address throughput limitations, researchers propose advanced sharding and off-chain solutions. Sharding involves partitioning the network into parallel groups to process transactions concurrently. Off-chain solutions, such as Payment Channel Networks (PCNs), move transactions away from the main chain to reduce latency and costs. Notable examples include the "Meepo" sharded consortium system and priority-aware transaction scheduling algorithms.
4. High-Performance Blockchains
This section investigates the use of Deep Reinforcement Learning (DRL) and unsupervised learning to optimize network topology and resource management. Topics include reducing latency in consensus protocols and ensuring "storage sustainability" for miners through mechanisms like Fee and Transaction Expiration Time (FTET), which prevent the ledger from becoming an unmanageable storage burden.
Key Insights and Results
The research presented in this issue demonstrates significant quantitative and qualitative improvements:
- Enhanced Efficiency: Proposed FL frameworks reduce communication traffic and training time while maintaining high model accuracy.
- Increased Throughput: Sharding and off-chain technologies drastically increase transaction throughput and reduce costs compared to traditional on-chain operations.
- Robust Security: New anti-collusion smart contracts and encryption methods improve the reliability of decentralized systems. Specifically, anti-collusion smart contracts were shown to reduce false-positive rates in "watchtower" mechanisms designed to protect off-chain transactions.
- Adaptive Protocols: The move toward self-adjusting blockchain protocols—where parameters like block size adapt to real-time network conditions—marks a shift toward more resilient and responsive decentralized systems.
Conclusion and Future Outlook
While this special issue highlights significant progress, several challenges remain. The computational overhead of intelligent optimization algorithms and the need for better adaptation to dynamic mobile environments are ongoing areas of research. Future work will likely focus on further improving storage sustainability and developing "meta-regulation" protocols that allow blockchains to autonomously adapt to the evolving needs of future communication networks.
Summary updated