The Celtic Gandalf Framework

Knowledge acquisition for diagnosis model in wireless networks

“…those that self-configure, selfprotect, self-heal and self-optimize) are the cost-effective solution to tackle the high complexity inherent to these networks. This paper is focused on self-healing (also called automatic troubleshooting) in the RAN of wireless systems, which, up to now, has received little attention in the existing literature, despite the huge interest of network operators and manufacturers of equipment (Altman et al, 2006).…”

Learning of model parameters for fault diagnosis in wireless networks

“…Diagnosis is currently a manual process accomplished by experts dedicated to daily analysis of the main Key Performance Indicators (KPIs) and the alarms of the network, aiming to determine the cause of the problems. Despite the huge interest of network operators and manufacturers of equipment in the automation of the troubleshooting procedures [1], this topic has received almost no attention from the academic community. The benefits of automating the diagnosis process are numerous: highly experienced staff can be released from mundane aspects of daily troubleshooting and can focus on the more complex cases; knowledge from different experts can be combined into a common diagnosis system, therefore making this knowledge available to the whole business at all times; diagnosis is faster and, therefore, the problems can be solved in shorter time; troubleshooting tasks can be deskilled; etc.…”