A Novel Energy Efficiency Metric for Model-Based Fault Diagnosis of Telecommunication Central Offices

Abstract A novel energy metric is presented, to be adopted for monitoring and diagnosis of telecommunication (TLC) central offices (COFs). Such an activity is motivated by the TLC players need to substantially reduce their energy demand, both to increase their market competitiveness and meet the stringent green-house gas (GHG) emission regulations. The proposed metric, the utilization factor (UF), was thus defined according to the energy break-down of TLC-COFs. Then, suitable data-processing techniques were applied to develop a diagnosis-oriented UF model. Model accuracy, found to be always capable of guaranteeing UF estimation errors safely below 15 % for all non-faulty COFs, was proven adequate to perform model-based fault detection and isolation of relevant malfunctioning, such as abnormal data acquisition and non-optimal energy management.