Spectral Efficiency Self-Optimization through Dynamic User Clustering and Beam Steering

This paper presents a novel scheme for spectral efficiency (SE) optimization through clustering of users. By clustering users with respect to their geographical concentration we propose a solution for dynamic steering of antenna beam, i.e., antenna azimuth and tilt optimization with respect to the most focal point in a cell that would maximize overall SE in the system. The proposed framework thus introduces the notion of elastic cells that can be potential component of 5G networks. The proposed scheme decomposes large-scale system-wide optimization problem into small-scale local sub-problems and thus provides a low complexity solution for dynamic system wide optimization. Every sub-problem involves clustering of users to determine focal point of the cell for given user distribution in time and space, and determining new values of azimuth and tilt that would optimize the overall system SE performance. To this end, we propose three user clustering algorithms to transform a given user distribution into the focal points that can be used in optimization; the first is based on received signal to interference ratio (SIR) at the user; the second is based on received signal level (RSL) at the user; the third and final one is based on relative distances of users from the base stations. We also formulate and solve an optimization problem to determine optimal radii of clusters. The performances of proposed algorithms are evaluated through system level simulations. Performance comparison against benchmark where no elastic cell deployed, shows that a gain in spectral efficiency of up to 25% is possible depending upon user distribution in a cell.

Anomaly Detection in Wireless Networks

This chapter is an introduction to multi-cluster based anomaly detection analysis. Various anomalies present different behaviors in wireless networks. Not all anomalies are known to networks. Unsupervised algorithms are desirable to automatically characterize the nature of traffic behavior and detect anomalies from normal behaviors. Essentially all anomaly detection systems first learn a model of the normal patterns in training data set, and then determine the anomaly score of a given testing data point based on the deviations from the learned patterns. The initial step of learning a good model is the most crucial part in anomaly detection. Multi-cluster based analysis are valuable because they can obtain the insights of human behaviors and learn similar patterns in temporal traffic data. The anomaly threshold can be determined by quantitative analysis based on the trained model. A novel quantitative “Donut” algorithm of anomaly detection on the basis of model log-likelihood is proposed in this chapter.

Mining of Leaders in Mobile Telecom Social Networks

In the age of Big Data, the social network data collected by telecom operators are growing exponentially. How to exploit these data and mine value from them is an important issue. In this article, an accurate marketing strategy based on social network is proposed. The strategy intends to help telecom operators to improve their marketing efficiency. This method is based on mutual peers' influence in social network, by identifying the influential users (leaders). These users can promote the information diffusion prominently. A precise marketing is realized by taking advantage of the user's influence. Data were collected from China Mobile and analyzed. For the massive datasets, the Apache Spark was chosen for its good scalability, effectiveness and efficiency. The result shows a great increase of the telecom traffic, compared with the result without leader identification.

Continuous-Time Markov Chain-Based Reliability Analysis for Future Cellular Networks

It is anticipated that the future cellular networks will consist of an ultra-dense deployment of complex heterogeneous Base Stations (BSs). Consequently, Self-Organizing Networks (SON) features are considered to be inevitable for efficient and reliable management of such a complex network. Given their unfathomable complexity, cellular networks are inherently prone to partial or complete cell outages due to hardware and/or software failures and parameter misconfiguration caused by human error, multivendor incompatibility or operational drift. Forthcoming cellular networks, vis-a-vis 5G are susceptible to even higher cell outage rates due to their higher parametric complexity and also due to potential conflicts among multiple SON functions. These realities pose a major challenge for reliable operation of future ultra-dense cellular networks in cost effective manner. In this paper, we present a stochastic analytical model to analyze the effects of arrival of faults in a cellular network. We exploit Continuous Time Markov Chain (CTMC) with exponential distribution for failures and recovery times to model the reliability behavior of a BS. We leverage the developed model and subsequent analysis to propose an adaptive fault predictive framework. The proposed fault prediction framework can adapt the CTMC model by dynamically learning from past database of failures, and hence can reduce network recovery time thereby improving its reliability. Numerical results from three case studies, representing different types of network, are evaluated to demonstrate the applicability of the proposed analytical model.

Identifying Dissatisfied 4G Customers from Network Indicators

Feedback data directly collected from users are a great source of information for telecom operators. They are usually retrieved as complaints and survey data. For the mobile telecoms sector, one purpose is to manage those data to identify network problems leading to customer dissatisfaction. In this paper, a quantitative methodology is used to predict dissatisfied users. It focuses on extraction and selection of predictive features, followed by a classification model. Two sets of data are used for experiments: one is related to complaints, the other to survey data. Since the methodology is similar for those two sets, prediction efficiency and influence of features are compared. Specific influence of user loyalty in survey data is highlighted. Thus, the methodology presented in this article provides a reference for the mobile operators to improve procedures for collecting feedback answers.

Detecting Abnormal Traffic in Wireless Networks Using Unsupervised Models

Development of high-speed LTE connections has induced an increasingly quantity of traffic data over the network. Detection of abnormal traffic from this continuous stream of data is crucial to identify technical problem or fraudulent intrusion. Unsupervised learning methods can automatically describe structure of the data and deduce patterns of the network. There are useful to identify unexpected behaviors and to promptly detect new type of anomalies. In this article, traffic in wireless network is analyzed through different unsupervised models. Emphasis is given on models combining traffic data with time stamps information. A model called Gaussian Probabilistic Latent Semantic Analysis (GPLSA) is introduced and compared with other methods such as time-dependent Gaussian Mixture Models (time-GMM). Efficiency and robustness of those models are compared, using both sampled and LTE traffic data. Those experimental results suggest that GPLSA can provide a robust and early detection of anomalies, in a fully automatic, data-driven solution.

Network-Based Targeting

This chapter focuses on two kinds of targeting in mobile industry: to target churning customers and to target potential customers. These two targeting strategies are very important goals in Customer Relationship Management (CRM). In the first part of the chapter, the author reviews churn prediction models and its applications. In the second part of the chapter, traditional innovation diffusion models are reviewed and agent-based models are explained in detail. Customers in telecom industry are usually connected by large and complex networks. To understand how network effects and consumer behaviors – such as churning and adopting – interplays with each other is of great significance. Therefore, detailed examples are given to network-based targeting analysis.

Predicting 4G Adoption with Apache Spark

Companies have long realized the value of targeting the right customer with the right product. However, this request has never been so inevitable as in the era of big data. Thanks to the tractability of the customers' behavior, the preference information for each individual is collected and updated by the firm in a timely fashion. In this study, we developed a targeting strategy for telecommunication companies to facilitate the adoption of 4G technology. Utilizing the most up to date machine learning technique and the information about individual's local network, we set up a prediction model of consumer adoption behavior. We then applied the model to the real world and conduct field experiment. We worked with the largest telecommunication company in China and used Apache Spark to analyze the data from the complete customer based of a 2nd tie city in eastern China. In the experiment group, we asked the company to use the list we generated as the targets and in the control group, the company used the existing targeting strategy. The results demonstrated the effectiveness of the proposed approach comparing to existing models.

Modeling for Time Generating Network

Modeling co-occurrence data generated by more than one processes in network is a fundamental problem in anomaly detection. Co-occurrence data are joint occurrences of pairs of elementary observations from two sets: traffic data in one set are associated with the generating entities (Time) in the other set. Clustering algorithms are valuable because they can obtain the insights from the varied distribution associated with generating entities. This chapter leverages co-occurrence data that combine traffic data with time, and compares Gaussian probabilistic latent semantic analysis (GPLSA) model to a Gaussian Mixture Model (GMM) using temporal network data. Experimental results support that GPLSA holds better promise in early detection and low false alarm rate with low complexity of implementation in a fully automatic, data-driven solution.

Evaluating Wireless Network Accessibility Performance via Clustering-Based Model

Using the large amount of data collected by mobile operators to evaluate network performance and capacity is a promising approach developed in the recent last years. One of the challenge is to study network accessibility, based on statistical models and analytics. In particular, one aim is to identify when mobile network becomes congested, reducing accessibility performance for users. In this paper, a new analytic methodology to evaluate wireless network accessibility performance through traffic measurements is provided. The procedure is based on ensemble clustering of network cells and on regression models. It leads to identification of zones where the accessibility remains high. Numerical results show efficiency and relevance of the suggested methodology.