Meta-Learning Based Device-to-Device Task Allocation for Improved Performance in Resource-Constrained Environments

D2D communication tasks are difficult to assign in resource-constrained environments like mobile edge computing and the Internet of Things. System performance, energy efficiency, and user happiness improve when task distribution is optimized. This paper presents a meta-learning technique for allocating Device-to-Device (D2D) tasks based on network node processing, resource, and communication overhead. The suggested technique evaluates the allocation strategy using a deep learning model. The approach is updated repeatedly using a Q-learning algorithm. We compare meta-learning and random allocation in a simulated network with five devices and ten tasks. The assessment of allocation techniques uses throughput and resource use statistics. The meta-learning technique outperforms random allocation in device task allocation stability and effectiveness, according to studies. Meta-learning, an Artificial Intelligence (AI) method, improves work allocation and system efficiency by dispersing resource use among various devices. Compare meta-learning approach performance to networked devices and workloads. Meta-learning scales better than random assignment to more devices and tasks without sacrificing performance.