As an important component of self-healing technology, outage detection is of great significance for the smooth operation of subsequent fault diagnosis and outage compensation operations. In this paper, we propose an outage detection algorithm that combines a hybrid Generative Adversarial Network (GAN) with an overlap-sensitive Artificial Neural Network (ANN) to solve the data imbalance problem as well as the overlap problem between data classes in outage detection. The proposed algorithm first synthesizes the outage data and adjusts the data distribution by hybrid GAN outage data distribution features. Then, based on the distribution of the samples in the feature space, the K-nearest neighbor algorithm is used to calculate the degree of overlap between the classes of the samples, and the samples are assigned weights accordingly. Finally, the resulting weight set is combined with the calibrated dataset and weighted to train the artificial neural network classifier. Simulation results show that when compared with the existing outage detection algorithm, the proposed algorithm improves the outage detection performance significantly, and can accurately detect multiple classes of outage cells.
A Hybrid GAN-Based Outage Detection Algorithm for Wireless Networks
2024-06-24
844821 byte
Conference paper
Electronic Resource
English
Outage Probability Analysis in Soft Handover for 3G Wireless Networks Path
British Library Conference Proceedings | 2005
|Automotive Lamp Outage Detection
SAE Technical Papers | 1972
|Hybrid Encryption Algorithm Based on Wireless Sensor Networks
British Library Conference Proceedings | 2019
|An outage compensation algorithm for WLANs
IEEE | 2013
|First-Order PMD Outage Prediction Based on Outage Maps
British Library Online Contents | 2010
|