Design of networking protocols for shallow water peer-to-peer acoustic networks

Communication between a set of underwater systems such as remote sensors, autonomous underwater vehicles and control vessels would enhance the effective use of such systems tremendously. As electromagnetic waves do not propagate well underwater, acoustics plays a key role in underwater communication. Although point-to-point acoustic links can be established via numerous modulation schemes, an acoustic communication network demands multiuser communication. In such an environment, orthogonal modulation schemes could provide a solution for multiple simultaneous acoustic links. As an alternative to orthogonal schemes, random shared access technology has proven successful in many wireless networks. Through numerical simulations, we compare the performance of orthogonal and random shared access for underwater networking. Over the past few decades, numerous networking protocols have been developed for use in wired and wireless networks. Due to significant differences in the characteristics of electromagnetic and acoustic channels, these networking protocols require modifications to perform well in underwater networks. As sound waves are much slower than electromagnetic waves, the latency in communication is typically much higher. Due to the multipath propagation and ambient noise, the effective data rates are lower and packet loss is much greater. In this paper, we simulate variants of some popular protocols for underwater use, focusing on the Physical and Datalink layers of the OSI protocol stack. The aim is to select an appropriate Physical Layer and Datalink Layer model for a small underwater network to be implemented.

I. Introduction

As the underwater acoustic channel is significantly different from wired and wireless electromagnetic channels, it is expected that network protocols may perform very differently in underwater acoustic networks. ln this paper, we present results from simulation of mainly the Datalink and Physical layers of an underwater network. Since the primary function of the Datalink layer in this simulation is Medium Access Control (MAC), we use the term MAC layer to mean the same. Other aspects of Datalink layer such as error correction were not simulated. A discrete-event simulation software was used to implement the models. The models of the Physical and MAC layers were adapted from the popular network protocols [1] A simple underwater channel model was used to test network performance. The primary performance criterion used is network throughput. The aim is to compare the relative performance of various networking schemes for use in underwater networks.