A novel optimized fuzzy neural network for enhanced topology control in k-connected mobile adhoc networks

Mobile Ad hoc Networks are self-organizing networks, dynamic in nature and consist of mobile nodes independent of the infrastructure. However, they are susceptible to the topology changes, communication link variations and node failures which make them highly susceptible to faults and disruptions. To maintain reliable communication under these unpredictable and dynamic scenarios is a major challenge. Fault-tolerant topology control is one of the main features of MANETs to maintain effective and secure connectivity and communication under these adverse conditions. To address the challenges a new Optimized Fuzzy Neural Network (OFNN) scheme is developed to establish efficient fault-tolerant topology control. In the first stage, an Improved Rabbit Optimization Algorithm is proposed to strategically select Cluster Heads, aiming to optimize the overall clustering efficiency. Next the input parameters of CHs such as neighbor node distance (NND), path stability (PS), and link expire time (LET) are given into OFNN to predict path reliability. Finally, data transmission is made by selecting the path with the maximum computed neuron value, which ensures path reliability and the overall network performance. To further refine the network's fault tolerance and overall efficiency, the performance of the OFNN is enhanced by using Osprey Optimization Algorithm.