Abstract:

Neural networks inspired by biological systems mimic the adaptive, resilient, and self-organizing properties of the human brain, making them ideal for dynamic and unpredictable environments. These features are particularly beneficial in disaster zones where communication systems are often disrupted and require rapid self-repair and reconfiguration. Conventional communication frameworks rely on static infrastructure and centralized control, making them vulnerable to collapse during natural or human-made disasters. These methods often fail to adapt in real-time to changing environmental conditions and node failures. To overcome these limitations, we propose the Bio-Inspired Adaptive Self-Healing Communication Framework (BASHComm). This model utilizes biologically-inspired neural network algorithms, including Hebbian learning for adaptive node connectivity, spiking neural networks (SNNs) for real-time signal processing, and decentralized reinforcement learning to dynamically reroute communication paths. The framework emulates neuroplasticity and redundancy, enabling autonomous fault detection and repair. The proposed BASHComm system is designed for deployment in ad-hoc mobile networks using low-power edge devices. It self-organizes in response to node failures or environmental changes and reconfigures communication routes without central coordination. This allows for robust, scalable, and energy-efficient operation under harsh and unstable conditions common in disaster-struck areas. Simulation and prototype testing demonstrate that BASHComm significantly improves system uptime, reduces message loss, and shortens reconfiguration time compared to traditional methods. The findings suggest that biologically-inspired neural network frameworks offer a promising direction for building intelligent, self-healing communication systems capable of sustaining critical information flow during emergency response scenarios.