Energy-Efficient Cell Association and Load Analysis in Heterogeneous Cellular Networks

This paper exploits the fundamental connection between cell association and load in the sense of bits-per-joule energy efficiency. Users adopt a proposed energy- efficiency-based association reward function to associate with their base stations in a heterogeneous cellular network consisting of K independent Poisson point processes (PPPs) of base stations (BSs). We first investigate the cell association probability and the statistics of the kth-tier BSs for a universal association reward function and they can be tailored for any cell association schemes that satisfy the constraints on the association reward function. Under the proposed energy-efficient cell association (EECA) scheme, the statistical properties of an associated BS, coverage probability, average energy efficiency and cell load are all derived and analyzed. We show that EECA can achieve a (much) higher coverage probability, average energy efficiency, and better cell load balancing if compared with maximum received power and nearest BS association schemes.