Biblio

In cellular network systems, it is commonly found that many errors or failures are caused by non-functioning components or human errors. Most failures are detected by a centralized Operation and Maintenance (OAM) software which will trigger an alarm as a form of warning. In fact, there are conditions when a failure or error occurs, but it cannot be detected by OAM software, which in turn will result in many complaints coming from customers. An event like this is called a sleeping cell, which is a condition where the network has a poor performance but does not generate alarm notifications in the Operation and Maintenance Center. In this paper, sleeping cell analysis was carried out on the LTE network using a self-healing approach to speed up the cell outage detection process. The process of sleeping cell analysis was based on the database of cell performance daily for all eNodeB located in West Java, referring the uplink and downlink values as the main parameters. The acquired database would then be processed and analyzed by the measurement method based on inference statistics, where this method would process a portion of the research data (sample), to draw the conclusions regarding the characteristics of the overall data population. Furthermore, data analysis was performed with signaling ladder diagram (SLD) approach to observe the signaling flow on the network, specifically in the uplink and downlink process, which is the initial indication of a sleeping cell.