Clustering Enabled Wireless Channel Modeling Using Big Data Algorithms

Recently, rapid growth in data services has ushered in the so-called big data era, and data mining and analysis techniques have been widely adopted to extract value from data for different applications. Channel modeling also benefits in this era, in particular by exploiting algorithmic techniques developed for big data applications. In this article, the challenges and opportunities in clustering-enabled wireless channel modeling are discussed in this context. First, some well known clustering techniques, which are potentially capable of enabling clustered channel modeling, are presented. Next, the motivation of cluster-based channel modeling is presented. The typical concepts of clusters used in channel models are summarized, and the state-of-the-art clustering and tracking algorithms are reviewed and compared. Finally, several promising research problems for channel clustering are highlighted.