Mitigating the effect of multipath using cross-correlation: application to underwater network cardinality estimation

Abstract

Number of node cardinality is an important parameter for a networks proper operation and this parameter can be determined using conventional protocol-based techniques, but these techniques have several limitations in underwater environment due to large propagation latency, high error rate, node mobility, non-negligible capture effect and high path loss. To develop a suitable cardinality estimation technique in presence of unique characteristics of underwater network, a statistical signal processing approach based on cross-correlation of Gaussian signals received at multiple sensors has been proposed in our previous works. In the initial formulation of this approach, a simplified model has been investigated without considering the multipath propagation effect. One of the common challenges for underwater wireless communication is multipath propagation of signals which result in inter-symbol interference ISI. An important feature of cross-correlation is its ability to mitigate the multipath effect which is demonstrated in this paper and applied to our previously proposed estimation scheme for multipath compensation. It is shown that, due to high dispersion coefficient of the medium, multipath effects on estimation results are negligible and a generalised process of multipath compensation for any dispersion factor is obtained to improve the robustness of this scheme.