Inmaculada Serrano

Data mining for fuzzy diagnosis systems in LTE networks

A Knowledge Acquisition learning algorithm is proposed for troubleshooting in LTE.A sensitivity analysis is performed on the proposed algorithm.The algorithm is tested with a live network scenario.The performance has been compared with a Bayesian Network based algorithm. The recent developments in cellular networks, along with the increase in services, users and the demand of high quality have raised the Operational Expenditure (OPEX). Self-Organizing Networks (SON) are the solution to reduce these costs. Within SON, self-healing is the functionality that aims to automatically solve problems in the radio access network, at the same time reducing the downtime and the impact on the user experience. Self-healing comprises four main functions: fault detection, root cause analysis, fault compensation and recovery. To perform the root cause analysis (also known as diagnosis), Knowledge-Based Systems (KBS) are commonly used, such as fuzzy logic. In this paper, a novel method for extracting the Knowledge Base for a KBS from solved troubleshooting cases is proposed. This method is based on data mining techniques as opposed to the manual techniques currently used. The data mining problem of extracting knowledge out of LTE troubleshooting information can be considered a Big Data problem. Therefore, the proposed method has been designed so it can be easily scaled up to process a large volume of data with relatively low resources, as opposed to other existing algorithms. Tests show the feasibility and good results obtained by the diagnosis system created by the proposed methodology in LTE networks.

Automatic Root Cause Analysis for LTE Networks Based on Unsupervised Techniques

The increase in the size and complexity of current cellular networks is complicating their operation and maintenance tasks. While the end-to-end user experience in terms of throughput and latency has been significantly improved, cellular networks have also become more prone to failures. In this context, mobile operators start to concentrate their efforts on creating self-healing networks, i.e., those networks capable of troubleshooting in an automatic way, making the network more reliable and reducing costs. In this paper, an automatic diagnosis system based on unsupervised techniques for Long-Term Evolution (LTE) networks is proposed. In particular, this system is built through an iterative process, using self-organizing maps (SOMs) and Wards hierarchical method, to guarantee the quality of the solution. Furthermore, to obtain a number of relevant clusters and label them properly from a technical point of view, an approach based on the analysis of the statistical behavior of each cluster is proposed. Moreover, with the aim of increasing the accuracy of the system, a novel adjustment process is presented. It intends to refine the diagnosis solution provided by the traditional SOM according to the so-called silhouette index and the most similar cause on the basis of the minimum Xth percentile of all distances. The effectiveness of the developed diagnosis system is validated using real and simulated LTE data by analyzing its performance and comparing it with reference mechanisms.

Diagnosis Based on Genetic Fuzzy Algorithms for LTE Self-Healing

Self-organizing network (SON) mechanisms reduce operational expenditure in cellular networks while enhancing the offered quality of service. Within a SON, self-healing aims to autonomously solve problems in the radio access network and to minimize their impact on the user. Self-healing comprises automatic fault detection, root cause analysis, fault compensation, and recovery. This paper presents a root cause analysis system based on fuzzy logic. A genetic algorithm is proposed for learning the rule base. The proposed method is adapted to the way of reasoning of troubleshooting experts, which ease knowledge acquisition and system output interpretation. Results show that the obtained results are comparable or even better than those obtained when the troubleshooting experts define the rules, with the clear benefit of not requiring the experts to define the system. In addition, the system is robust, since fine tuning of its parameters is not mandatory.

LTE performance data reduction for knowledge acquisition

In the last years, mobile networks have seen a great increase in complexity, as the data traffic, the demand for quality and the variety of offered services have grown. The management costs of modern networks are growing, at the same time as operators compete to offer shorter downtime and less impact of network issues on the user experience. Self-Organizing Networks (SON) offer a solution to these problems. Among these SON functionalities in cellular networks, self-healing automates the resolution of problems in the radio access network. To perform the task of diagnosis (or root cause analysis), Knowledge-Based Systems (KBS) are often used. These systems need a previous process of training (or learning), in which they are fed instances of real problems. In this paper, an algorithm for extracting the key information for these vectors is proposed. The inputs of the algorithm are big matrices of time-dependent network performance data, and the outputs are simple one-dimensional vectors ready to be used in learning algorithms.

Correlation-Based Time-Series Analysis for Cell Degradation Detection in SON

The increasing amount of network elements in the current deployments of cellular networks is leading to an enormous complexity of the operation and maintenance. Self-organizing networks (SONs) is a good solution for operators to save operational expenditures by automating network management. One of the key challenges in this context is the automatic identification of degraded cells. In this letter, a method to detect degraded cells through the analysis of the time evolution of metrics is proposed. Results show that cell faulty patterns can be effectively detected by comparing them with a set of fictitious degraded patterns.

Knowledge Acquisition for Fault Management in LTE Networks

Self-healing is one of the main functionalities of Self-Organizing-Networks. Among self-healing functions, diagnosis or root cause analysis, consisting of identifying the fault cause in problematic cells, is one of the most complex tasks. Expert systems, such as Fuzzy Logic Controllers or Bayesian Networks, have been previously proposed to implement automatic diagnosis systems in the radio access segment of mobile communication networks. In order to achieve accurate results, these diagnosis systems should contain the knowledge of experienced LTE troubleshooting experts. However, these experts normally have neither the time nor the expertise in artificial intelligence to define the expert system. In this work, we propose a novel knowledge acquisition system that obtains this knowledge in the least possible intrusive way. The proposed method collects the Performance Indicators data from the relevant time intervals together with the expert’s diagnosis and uses them as inputs for a Data Mining algorithm to extract diagnosis rules.

Data Analytics for Diagnosing the RF Condition in Self-Organizing Networks

The current trend in the management of mobile communication networks is to increase the level of automation in order to enhance network performance while reducing Operational Expenditure (OPEX). In this context, the 3rd Generation Partnership Project (3GPP) has presented different solutions. On the one hand, Self-Organizing Networks (SON) include self-healing capabilities, which allow operators to automate their troubleshooting tasks in order to identify and solve the problems of the network. On the other hand, the use of mobile traces or Minimization of Drive Tests (MDT) are proposed to automate the collection of users measurements and signalling messages. This paper proposes to combine both solutions, SON and traces, with the purpose of quickly detecting and solving issues related to the radio interface. That is, the user information gathered by the cell traces function is used to perform an automatic diagnosis of the RF condition of each cell. In addition, the proposed approach allows to precisely locate RF problems based on the assessment of the RF condition. Mobile traces constitute large sets of data, whose analysis requires the application of big-data analytics techniques. The proposed system has been evaluated in two different live LTE networks, demonstrating its validity and utility.

Unsupervised System for Diagnosis in LTE Networks Using Bayesian Networks

Nowadays, the size and complexity of mobile networks are growing ceaselessly. Therefore, the management of mobile networks is a significant, expensive and demanding task to perform. In order to simplify this task, Self-Organizing Networks (SON) appear as a unified solution to autonomously manage a mobile network. One of the fundamental functions of SON is self-healing. Within self- healing, the objective of fault diagnosis or root cause analysis is the identification of problem causes in faulty cells. With that aim, in this paper, an unsupervised diagnosis system for LTE (Long Term Evolution) based on Bayesian networks is presented. In particular, the system is divided in two separate steps. First of all, the discretization of the input data is done. Then, the system provides an identification of the cell status. Depending on the discretization method, the performance of the system is different, so, in this paper, different methods have been evaluated. Results have proven the high success rate achieved with the proposed system, particularly when the Expectation-Maximization (EM) algorithm is used for the discretization.

Modelling LTE Solved Troubleshooting Cases

Abstract Self-Organizing Networks (SON) add automation to the Operation and Maintenance of mobile networks. Self-healing is the SON function that performs automated troubleshooting. Among other functions, self-healing performs automatic diagnosis (or root cause analysis), that is the task of identifying the most probable fault causes in problematic cells. For training the automatic diagnosis functionality based on support-decision systems, supervised learning algorithms usually extract the knowledge from a training set made up from solved troubleshooting cases. However, the lack of these sets of real solved cases is the bottleneck in the design of realistic diagnosis systems. In this paper, the properties of such troubleshooting cases and training sets are studied. Subsequently, a method based on model fitting is proposed to extract a statistical model that can be used to generate vectors that emulate the network behavior in the presence of faults. These emulated vectors can then be used to evaluate novel diagnosis systems. In order to evaluate the feasibility of the proposed approach, an LTE fault dataset has been modeled, based on both the analysis of real cases collected over two months and a network simulator. In addition, the obtained baseline model can be very useful for the research community in the area of automatic diagnosis.

Improving Cell Outage Management Through Data Analysis

Self-organizing networks (SONs) are an important feature for network management automation in the new generation of mobile communications. While SONs have been considered to be part of the recent 3GPP standards such as LTE, it is expected that future 5G mobile networks will present new challenges for SON solutions. One of the most important use cases in self-healing is cell outage compensation (COC). This article proposes an important improvement in COC functionality based on analyzing large real data sets from live networks in order to adapt compensation to the real neighboring context. First, we propose two methods (offline and online) to classify outages depending on the degraded metrics in neighboring cells, and we present results for a set of cell outages occurring in a live LTE network. Second, we propose a method for estimating lost traffic due to cell outages in order to quantify the load that cannot be intrinsically absorbed by neighboring cells. Finally, a novel COC methodology is proposed by taking into account the results obtained in the two previous studies.