User association with jointly maximising downlink sum rate and minimising uplink sum power for heterogeneous cellular networks

In heterogeneous cellular networks (HCNs), the user association is a challenging topic since some different base stations coexist. Moreover, because of the asymmetric uplink and downlink in HCNs, a joint uplink and the downlink association algorithm should be designed to improve the system performance. For the practical implementation, the authors need to ensure that the algorithm is highly effective. Thus, they try to design an association strategy that jointly maximises downlink sum rate and minimises uplink sum power, and formulate it as a sum-utility maximisation problem. To solve this problem, they design a centralised association algorithm via a gradient descent method, and develop a distributed association algorithm via dual decomposition. Simulation results show that, compared with the signal strength-based association, range expansion association and the method proposed by Ye, their scheme has some significant advantages in the mass.