Abstract

In multi-hop narrowband networks, historical data recovery is essential for maintaining data consistency, diagnosing network issues, and ensuring long-term reliability. However, these networks face significant challenges, including packet loss, delayed acknowledgments, limited node storage, and energy constraints in battery-powered nodes. Traditional retransmission-based recovery mechanisms, such as Automatic Repeat reQuest (ARQ) and Forward Error Correction (FEC), often introduce excessive network overhead and are inefficient in low-power, multi-hop communication scenarios. This paper proposes an efficient historical data recovery mechanism that leverages a line-powered gateway, which continuously records network topology changes, node status updates, and event logs. The gateway facilitates historical data retrieval via both wired (UART) and wireless interfaces, optimizing recovery based on latency constraints and network conditions. When direct gateway retrieval is not feasible, the system employs a multi-hop redundancy-based recovery mechanism, enabling neighboring nodes to reconstruct missing data. If neither the gateway nor neighboring nodes contain the required information, the system applies an interpolation-based data reconstruction approach, utilizing statistical models such as linear interpolation and Kalman filtering to estimate missing values. To further enhance efficiency, the system implements energy-aware retrieval strategies, including batch-based data requests, compressed storage, and adaptive selection of recovery methods to minimize power consumption in battery-operated nodes.
The proposed approach is evaluated through simulations in a multi-hop narrowband mesh network, analyzing key performance metrics such as data recovery success rate, network overhead, and latency impact. Results demonstrate that the method reduces data recovery latency by up to 40%, improves the recovery success rate by 25% compared to ARQ-based methods, and maintains energy efficiency in battery-powered nodes. These findings highlight the scalability and reliability of the proposed historical data recovery framework, making it a viable solution for resource-constrained multi-hop narrowband networks.