Abstract: This paper studies joint optimization of user access and power control in downlink cooperative heterogeneous networks. Given coordinated transmission among base stations (BSs) to suppress the inter-cell interference, the user access and power control policies are jointly optimized, aimed at minimizing the total transmit power of all BSs, subject to per-user rate requirement, per-BS power constraint, and the restriction on the maximal number of BSs accessed by each user. This problem is a combinatorial optimization problem involving both 0/1 variables and continuous variables, where exhaustive searching is required to obtain globally optimal solution at the expense of very high computation complexity. To reduce the complexity, an efficient joint optimization method for user access and power control based on geometric programming is proposed, which first approximates the involved 0/1 variables as continuous functions, then transforms the resultant optimization problem into a standard geometric programming problem that can be solved efficiently. Simulation results show that, compared to the existing accessing nearest BS policy, accessing receive-power-maximal BS policy, and accessing biased-receive-power-maximal BS policy, the proposed method can effectively balance the traffic load of cells, and thus can support much higher data rate requirements of users and can also largely reduce the total transmit power under the same users’ data rate requirement.