Resource Allocation for Network Slicing in RAN Using Case-Based Reasoning

As a key technology of 5G, network slicing can meet the diverse needs users. In this research, we study resource allocation in radio access networks (RAN) by case-based reasoning (CBR). We treat user distribution scenario as case and stored massive number cases library. CBR is used to match new with library find similar determine best slice bandwidth ratio based on these cases. matching process, k-nearest neighbors (KNN) algorithm retrieve cases, nearest k being determined considering sparsity reduction locality-preserving projections. Although only an initial study, results confirm that proposed architecture capable allocating resources efficiently terms prediction error computational cost.