A Deep Multi-Task Learning Approach to Impulsive Noise Parameter Estimation

Abstract

Impulsive noise poses a significant challenge to the reliability of wireless communication systems, necessitating accurate estimation of its statistical parameters for effective mitigation. This paper introduces a multitask learning (MTL) framework based on a CNN-LSTM architecture enhanced with an attention mechanism for the joint estimation of impulsive noise parameters. The proposed model leverages a unified weighted-loss function to enable simultaneous learning of multiple parameters within a shared representation space, improving learning efficiency and generalization across related tasks. Experimental results show that the proposed MTL framework achieves stable convergence, faster training, and enhanced scalability with modest computational overhead. Benchmarking against conventional single-task learning (STL) models confirms its favorable complexity-performance trade-off and significant memory savings, indicating the effectiveness of the MTL approach for real-time impulsive noise parameter estimation in wireless systems.