LCSMIBR: Design of a Low complexity Sleep-scheduling Model with Iterative Bioinspired learning for improved Routing efficiency under heterogeneous network condition

Abstract

Abstract One of the best methods for increasing energy efficiency in large-scale wireless networks is sleep scheduling. The scalability of existing sleep scheduling algorithms is limited in real-time applications due to their higher complexity or lack of coordination. Furthermore, these models' routing performance needs to be adjusted for various use scenarios. This essay suggests designing a low complexity sleep-scheduling model with iterative bioinspired learning for better routing efficiency in the context of diverse networks in order to integrate these qualities. At first, the suggested model gathers progressively larger data sets on node location, energy usage, fault frequency, and additional quality of service measures. These parameter sets are used to train a set of scalar nodes (nodes with low efficiency levels), that decide initial sleep scheduling cycles for high-efficiency multimedia nodes. The sleep schedules are co-ordinated by multimedia nodes during their inter-node communications, which assists the scalar nodes to iteratively improve sleep-and-wake cycles. These cycles are decided by a Bacterial Foraging Optimization (BFO) process, which assists in improving energy efficiency without compromising on QoS performance levels. The BFO process is cascaded with a fault-tolerant Grey Wolf Particle Swarm Optimizer (GWPSO), that evaluates high-efficiency routing paths that can provide low delay, low energy consumption, with lower probability of faults. Due to these operations, the proposed model is able to reduce communication delay by 4.9%, and energy needed for communication by 8.5%, while improving throughput by 3.4%, and consistency of routing-performance by 1.9% when compared with existing routing techniques. This performance was observed to be consistent even under node-level faults for heterogenous network scenarios.