Multichannel Analysis of Cell Range Expansion and Resource Partitioning in Two-Tier Heterogeneous Cellular Networks

Abstract

Cellular heterogeneous networks (HetNets) can improve capacity by offloading users from congested macro cells to lightly loaded small cells through biased association known as cell range expansion (CRE). However, the offloaded (range-expanded) users must be protected from macro interference through time/frequency resource partitioning. In this paper, we develop an analytical framework to evaluate the performance gain due to CRE further supported by resource partitioning in two-tier (macro-pico) networks with multichannel downlinks, e.g., those based on orthogonal frequency division multiple access (OFDMA). By exploiting the flexibility in subchannel allocation offered by OFDMA, frequency-domain resource partitioning is proposed in which the macro tier is muted on a fraction of total subchannels, which are allocated exclusively to range-expanded pico users. The load perceived by a base-station is a key factor in determining its interference contribution over the network and is directly affected by user offloading and resource partitioning. Thus, the analysis of such systems must incorporate cell load. While previous studies mostly rely on full-load assumption, in this paper, we properly characterize cell load as the function of user density, association bias and resource partitioning fraction. We then, evaluate the performance in terms of average user data rate over the entire network, and also investigate the optimal choice of association bias and resource partitioning fraction.