An Association Rules Based Conflict-Graph Construction Approach for Ultra-Dense Networks

Ultra-dense networks (UDNs) have been widely recognized as a promising technology in meeting the higher throughput requirement of the fifth generation (5G) network. However, the dense deployment of femtocells brings an unprecedented challenge of introducing severe co-channel interference (CCI), which puts a great limitation on the network performance. Therefore, the interference mitigation problem in UDNs is of great importance. The conflict-graph is the representation of underlying interference constraints of the network and the basis of interference mitigation. Most prior studies construct conflict-graphs based on accurate geographical distance information although it is usually unavailable by the network operators in actual. A more practical association rules based down-link conflict-graph construction approach, which utilizes negative acknowledgement (NACK) data, new data indicator (NDI) data and resource block (RB) allocation data, is proposed in this paper. The relative interference intensities of different interfering sources are determined by the calculated confidence degrees and taken as the edge weights of the constructed conflict-graph. Furthermore, the proposed conflict-graph construction approach is extremely accurate especially for severe interference, which has been verified by simulation results.