Environmental equalizer for underwater communications

Abstract

This paper presents an environmental-based equalization algorithm for underwater communications. This algorithm is based on the passive time-reversal (pTR) and waveguide invariant properties of ocean channels. Passive time-reversal allows for the implementation of a simple communications system, but it loses performance in the presence of geometric mismatch between the probe-signal and the actual data symbols transmission. The waveguide invariant properties state that geometric mismatches, both in depth and range, can be partially compensated by applying an appropriate frequency shift in the passive time-reversal operator. Results with binary PSK data at a carrier frequency of 3.6 kHz, collected during the MREA'04 sea trial, show that the Mean Square Error (MSE) between the transmitted and the received data symbols remains stable at least to a range mismatch of about 37.5 m in the presence of source depth and an array depth oscillations of approximately 0.7 m. In such conditions, when comparing the proposed pTR equalizer with plain pTR, an overall gain of approximately 4.11 dB in output MSE is achieved. I. INTRODUCTIONIn recent years Time Reversal (TR) has received particular attention from the scientic community. After practical demonstration of its spatial-temporal focusing capabilities in the ocean [1] several applications of active TR (aTR), from tomography to communications, were suggested [1], [2]. Passive TR uses a receive-only array, and a probe-signal is transmitted ahead of the data for channel Impulse Response (IR) estimation. The IR estimate is then used as a synthetic channel for temporal focusing of the data signal, which is equivalent to the deconvolution of the multipath generated by the real channel.When applied to underwater coherent communications the achieved TR focus is not perfect due to errors on the IR estimate and the time variability of the channel, resulting in uncompensated intersymbol interference (ISI) [3]. That problem is even more relevant in communications with a moving source and/or receiver. In that case it is intuitive that a rapid degradation of passive TR temporal focusing will occur due to the increasing mismatch between the assumed and actual channels. In order to guarantee a longer stability of the focal spot, three solutions are usually proposed: one is to transmit probe-signals more frequently; another is to use an adaptive algorithm to track the IR from the initial probe signal IR estimation; and nally a third alternative is to use a low-complexity equalizer with only one coefcient per channel. A performance comparison between those adaptive pTR variants is presented in [4]. The major inconvenient of the rst method is that frequent transmission of probe-signals reduces the overall transmission rate. In the second the channel estimates are represented by a large number of coefcients that have to be adapted. The third case seems to be the best compromise between complexity and efciency. In this paper a different approach is proposed by considering that the ...

Probe timing optimization for time-reversal underwater communications

“…For high SN R in ≈ −10 dB the residual ISI of the TR operator becomes dominant and leads to saturation of SN R mse . For high residual ISI, in the HF case, Figure 4(b) shows that for a low SN R in ≈ −25 dB good agreement in shape is obtained between the ideal pTR and MSE curves a better agreement would be obtained if instead of the SN R out given by (17) the SN R out given by (18) was used, nevertheless the agreement in shape its enough for the optimum time window length prediction. For high SN R in ≈ −10 dB the residual ISI of the TR operator becomes dominant and leads to saturation of SN R mse .…”

“…Future developments should address the problem of using the proposed time-window optimization with adaptive pTR-based equalizers [17]. Up to now the proposed probe timing optimization method as been successfully applied with the pTR based environment equalizers presented in [18].…”

Path specific Doppler compensation in time-reversal communications

“…8 In current Doppler compensation techniques, Doppler distortion is compensated with a single value that fails to give maximum compensation. 9,10 In Silva et al, 11 an improved version of pTR frequency shift pTR (FSpTR) was proposed that includes appropriate frequency shifts in the pTR impulse response estimates so as to maximize the output power at each time instant during the interval between successive probes. It was proved with both simulated and real data that FSpTR could provide an MSE gain of several decibels and in Vilaipornsawai et al, 12 a DFE was integrated with FSpTR to improve the performance of the communication system.…”

Underwater communications for moving source using geometry-adapted time reversal and DFE: UAN10 data

“…This system is considered as a Single Input Multiple Output (SIMa) communication system. To address the time-varying channel effect, caused by a moving source/receiver, the geometry adapted pTR technique was proposed in [2], [3]. It was shown that by employing a frequency shifted version of the estimated channel impulse response in the pTR processing, geometric changes can be partially compensated.…”

“…It was shown that by employing a frequency shifted version of the estimated channel impulse response in the pTR processing, geometric changes can be partially compensated. Hence, the technique of [3] is referred to as Frequency Shift pTR 978-1-61284-4577-0088-0/11/$26.00 ©2011 IEEE (FSpTR). Although a pTR-based technique can mitigate the lSI problem, an equalizer is required to eliminate the residual lSI as shown theoretically in [4].…”