Estimating the effect of network element events in a wireless network

摘要: Network events, like outages, are costly events for communication service providers (CSPs) not only because they represent lost revenue but also of adverse effects suffered by the CSP’s customers. Quantifying effect negative on certain key performance indicators allows CSP to measure network resources impacted, provide data a more robust assurance process, and assign appropriate severity events. These additional insights may help optimize resource allocation, ticketing, troubleshooting response times. This paper presents novel heuristic algorithm that takes advantage daily patterns observed in most wireless stability differences between original time series lagged version. The proposed uses those previous actual values make accurate predictions time-series traffic volume estimated event. is compared with state-of-the-art autoregressive, integrated, moving average (ARIMA) model results reported. has reduced standard deviation error percentage 4.8 points, no bias, executes 97% faster than ARIMA model. provides an methodology online or batch event impact estimation could potentially be implemented traditional relational database management systems (SQL) Big Data environments.