A Multivariate Approach to Predicting Quantity of Failures in Broadband Networks Based on a Recurrent Neural Network

Abstract

In this paper, we present a multivariate recurrent neural network model for short-time prediction of the number of failures that are expected to be reported by users of a broadband telecommunication network. An accurate prediction of the expected number of reported failures is becoming increasingly important to service providers. It enables proactive actions and improves the decision-making process, operational network maintenance, and workforce allocation. Our previous studies have shown that the recursive neural network is flexible enough to approximate the dynamics of the failure reporting process. Development of the model is based on long-term monitoring of failure-reporting processes and experience gained through fault management related to the network of one of the leading Croatian telecom providers (T-HT). Many factors, both in the network and outside the network, influence the time series representing failure reporting. The model encompasses the most important predictor variables and their logical and temporal dependencies. Predictor variables represent internal factors such as profiles of past and current quantities of failures as well as external factors like weather forecasts or announced activities (scheduled maintenance) in the network. External factors have a strong effect on fault occurrence, which finally results in failures reported by users. These factors are quantified and included as input variables to our model. The model is fitted to the data from different sources like an error-logging database, a troubleticket archive, announced settings logs and a meteo-data archive. the model is examined on simulation tests varying the prediction horizons. Assessment of the model's accuracy is made by comparing results obtained by prediction and the actual data. This research represents a real-world case study from telecom operations. The developed prediction model is scalable and adaptable so that other relevant input factors can be added as needed. Hence, the proposed prediction approach based on the model can be efficiently implemented as a functionality in real fault-management processes where a variety of available input data of different volumes exist.

Predicting Network Faults using Random Forest and C5.0

“…There are several research conducted that focused on network fault prediction using different data and techniques. Current research work treats fault related data in two ways: (i) using the quantity of trouble tickets created and time series predictive model to predict the quantity of network fault [1,2] and (ii) using system logs generated by specific network components to predict the likelihood of the components to be faulty [3,4] In this paper, 7-day sliding window is used to aggregate the sessions of the Internet usage data and the network quality data. The customer trouble ticket archive is used to label the aggregated data as fault or no-fault.…”

“…There was a focus in Indonesia's telco [1] on predicting the quantity of network fault related to customer on-premise equipment in a monthly interval using Autoregressive Integrated Moving Average (ARIMA). The multivariate recurrent neural network model can also be used to forecast the network fault in both short-term (hourly and daily interval) and long-term (weekly and monthly interval) [2]. Given enough training data, the number of trouble tickets that will be created can be forecasted in advance allowing the telco to manage their workforce allocation.…”

A Predictive Approach for Managing Network Port Resources of Service Providers

“…A communication system is built using any number or types of network infrastructure equipment. Those equipment types can be customer premise equipments (CPEs), copper twisted pair, fiber optic cable, Digital Subscriber Line Access Multiplexer (DSLAM) port, distribution point, main distribution frame and splitter [1]. Service Providers are deploying broadband IP networks nationwide to provide subscribers better and scalable Internet services.…”

“…There are different studies on predicting the network resource availability in different regions of the world [1], [2], [3], [4], [5]. The authors in [2] are studying availability of Internet service and propose a method that predicts the availability of IP-VPN end-to-end service in Taiwan.…”

“…The authors in [2] are studying availability of Internet service and propose a method that predicts the availability of IP-VPN end-to-end service in Taiwan. The authors in [1] have proposed a multivariate recurrent neural network model to predict the number of failures in broadband networks of a leading Croatian telecom provider. For example, the authors in [6] have demonstrated up to 30% energy savings at the Points of Presences (PoPs) of a nationwide Italian Internet service provider.…”