RESUMEN
This work presents a method for estimating key quality indicators (KQIs) from measurements gathered at the nodes of a wireless network. The procedure employs multivariate adaptive filtering and a clustering algorithm to produce a KQI time-series suitable for post-processing by the network management system. The framework design, aimed to be applied to 5G and 6G systems, can cope with a nonstationary environment, allow fast and online training, and provide flexibility for its implementation. The concept's feasibility was evaluated using measurements collected from a live heterogeneous network, and initial results were compared to other linear regression techniques. Suggestions for modifications in the algorithms are also described, as well as directions for future research.
RESUMEN
We develop an unsupervised approach to condition monitoring of cellular networks using competitive neural algorithms. Training is carried out with state vectors representing the normal functioning of a simulated CDMA2000 network. Once training is completed, global and local normality profiles (NPs) are built from the distribution of quantization errors of the training state vectors and their components, respectively. The global NP is used to evaluate the overall condition of the cellular system. If abnormal behavior is detected, local NPs are used in a component-wise fashion to find abnormal state variables. Anomaly detection tests are performed via percentile-based confidence intervals computed over the global and local NPs. We compared the performance of four competitive algorithms [winner-take-all (WTA), frequency-sensitive competitive learning (FSCL), self-organizing map (SOM), and neural-gas algorithm (NGA)] and the results suggest that the joint use of global and local NPs is more efficient and more robust than current single-threshold methods.