Blockchain-Enabled Cross-Device Authentication and Efficient Task Allocation Using Fuzzy Grey Wolf Optimization for IoT Networks

Cross-device authentication and task allocation across IoT devices, nodes, sensors, and gateways are difficult due to limited resource constraints and the heterogeneous nature of IoT devices. Traditionally, IoT devices rely on centralized servers, which pose single points of failure, data breaches, and security risks. To overcome these challenges, this paper proposes a framework for cross-device authentication and task allocation in an IoT network. To enable secure, faster cross-device communication, the paper proposed the BCHAuth algorithm, which integrates blockchain and ChaCha20-Poly1305 cryptography. Additionally, for optimal task allocation, FGWO is proposed coupled with fuzzy logic and grey wolf optimization algorithm. Initially, each IoT device is assigned a unique, decentralized ID and transaction records are stored in the blockchain's on-chain metadata via a smart contract. Smart contracts define the registration process and authentication protocols to ensure that only authorized devices with valid credentials can participate in the IoT network. To address the complex task-allocation problem, the proposed FGWO adapts conventional GWO to the discrete and heterogeneous IoT environment. The fuzzy inference system (FIS) evaluates each authenticated and eligible device using its processing power ${\pi }_i$, current workload ${\lambda }_i$, available bandwidth ${\beta }_i$, residual energy $E_i$ and trust score $T_i$ to generate a suitability score $S_i$. The suitability score guides the initialization of discrete task-to-device allocation candidates, while the conventional GWO leader-based mechanism performs exploration and exploitation. Following the position update, fuzzy suitability information is further used during candidate refinement to redirect selected tasks toward higher-suitability feasible devices. To evaluate the performance of the proposed BCHAuth and FGWO algorithms, they are compared with several state-of-the-art algorithms. The experimental results demonstrate that the proposed BCHAuth and FGWO achieve maximum authentication success rate (95.81%), minimum energy consumption (2.87 mJ), network bandwidth usage (10.52 kbps), average task allocation time (77.564 ms), CPU and memory utilization (28.85% and 32.62%), task latency and task waiting time (32.13 and 13.45ms), and the highest packet delivery ratio (PDR) (87.73%).