A Throughput-Enhanced Transmission Scheme for Decoupled Heterogeneous Networks

Abstract

We propose a throughput-enhanced transmission scheme for decoupled heterogeneous networks to address the problems of less precise estimated channel and low downlink (DL) throughput of decoupled user equipments (UEs). We develop a least square (LS) data-aided (DA) channel estimator to improve DL channels of decoupled UEs using decoded uplink (UL) data along with training sequences to jointly estimate channels. The normalized mean square error (NMSE) of LS DA method is derived with the bit error ratio (BER) of decoded UL data sequences. Further, an interference-nulling (IN) zero-forcing (ZF) precoding is designed to cancel DL interference from UL base stations of decoupled UEs by leveraging estimated channels of decoupled UEs in UL. Simulations show DA methods outperform conventional estimators in NMSE and promote the rate performance of both decoupled and coupled UEs. IN ZF precoding can further boost the rate of decoupled UEs and bridge the rate gap.