Resource allocation optimization in LTE-A/5G networks using big data analytics

The next generation wireless cellular communication networks are going to be completely internet protocol (IP) based. This leads to evolution of cellular networks into software controlled hierarchical abstract networks. Software controlling enables extraction of large volumes of networkdata usage and transactions related information in the form of log files, configuration files, database entries etc. This paper proposes a mechanism for efficient exploitation of this big volume of data for addressing the ever existing problem of optimum radio resource allocation to users. This work is based on the map-reduce data processing technique, by implementing non-orthogonal binary singular value decomposition (SVD) of binary data matrices for fuzzy pattern recognition and grouping. Patterns identified by proposed method can be interpreted efficiently for resource allocation in wireless networks.