Pilot Enrichment Methods for Improving Quality of Received Signal in Underwater Acoustic OFDM Systems

Abstract

In OFDM systems, pilots are usually sent along with the data carriers to calculate channel impulse response (CIR) for channel estimation. However, transmitted signals are normally distorted by many environmental factors when propagating through underwater communication channels. Thus, the original pilots seem to be not enough to recover “good” CIR for channel estimation in underwater acoustic OFDM (UWA-OFDM) systems. Recent studies have developed data pilot-aided (DPA) strategies, known as STA and CDP, for time-variant channels like vehicle-to-vehicle (V2V) communications. These methods work on the assumption of high cross-correlation of CR between two adjacent data symbols and the support of two long training symbols in the IEEE 802.11p packet. These conditions are not provided in the UWA-OFDM systems, which leads to be infeasible to apply existing DPA strategies. Therefore, this paper introduces a novel pilot enrichment (PE) method for improving the performance of the channel estimation in UWA-OFDM systems. The core idea is that some data subcarriers are extracted under a condition related to constellation mapping for creating supplementary “pilots”. The “pilots” are used together with the original pilots for channel estimation. Our method is evaluated in comparison to the MMSE estimation on various modulation schemes. The experimental results illustrate that the PE estimator can perform better than the MMSE estimator in terms of bit error rate (BER), especially on medium and high SNR values.