본문 바로가기
  • Home

Lightweight Blockchain Framework with GAN-based Adaptive Raft Consensus Mechanism in FANET Environment

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2026, 31(8), pp.87~95
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : July 13, 2026
  • Accepted : August 5, 2026
  • Published : August 31, 2026

Dana Yang ORD ID 1

1한국성서대학교

Accredited

ABSTRACT

This paper proposes a lightweight blockchain framework that combines GAN-based data restoration with an adaptive Cell-based Raft consensus for FANET environments, where high mobility and frequent link changes cause data loss and consensus delays. In the proposed framework, image data corrupted during UAV-to-UAV transmission is restored using ESRGAN, and blockchain transactions are generated based on the hash values of the restored data. In addition, the framework dynamically organizes the leader and consensus-participating nodes by comprehensively evaluating the trust level and data quality of each UAV, and aggregates the generated transactions into cells for batch processing, thereby reducing repetitive AppendEntries RPC message exchanges and the communication burden on the leader node. NS-3-based consensus simulations show that the proposed system reduces the consensus latency by approximately 15.1% and the number of RPC messages by approximately 19.3% on average. Separate image restoration experiments show that ESRGAN maintains an SSIM of approximately 0.67 at a packet loss rate of 20%.

Citation status

* References for papers published after 2025 are currently being built.