QB-AutoIDS: A Blockchain-Based Decentralized Autonomous Cyberattack Detection System for Consumer Electronics Using Hybrid Double Q-Learning and Bi-LSTM

Blockchain technology holds promise for securing decentralized interactions in consumer electronics, especially within IoT ecosystems such as smart home devices and wearable health monitors. However, user transactions remain vulnerable to cyberattacks, including DDoS and spoofing, which threaten data confidentiality and system availability. To address this, we propose QB-AutoIDS, a decentralized autonomous intrusion detection system that combines clipped double Q-learning (CDQL) with bidirectional long short-term memory (Bi-LSTM) networks for real-time cyberattack detection in blockchain-enabled consumer IoT environments. QB-AutoIDS operates directly on blockchain nodes and optimizes detection performance through CDQL-enhanced Bi-LSTM modeling. We evaluate our method on four benchmark datasets—DDoS-LFA, CSE-CIC-IDS, BoT-IoT, and AWS—achieving 99.105%, 99.5%, 99.99%, and 99.4% accuracy, respectively. On DDoS-LFA, QB-AutoIDS improves over Collaborative Learning (CoL), Centralized Learning (CeL), and Independent Learning (IL) by 1.76% in accuracy, 3.11% in precision, and 3.33% in recall. Compared to baseline models (LSTM, GMDH, and CNN), QB-AutoIDS further increases accuracy by 4.6%, 11.1%, and 20.9%, respectively, underscoring its effectiveness for scalable and accurate cyber threat detection in resource-constrained consumer IoT applications.