Unified simulation and implementation software framework for underwater MAC protocol development

MAC protocol development for underwater acoustic modem has been an active research area for many years. The primary mode of investigation is through network simulations. Subsequent implementation in modems for testing in sea trials typically involves porting of simulation code into the appropriate programming language and software environment in a modem. Correct porting is critical since, if the modem implementation differs from the simulation code, comparison of sea trial results with simulation results become misleading and the performance observed in simulations might be compromised. One of the challenges associated with the above porting process is to maintain exact algorithmic match in the modem implementation and the simulation code as in many protocols, minor variations in the logic can have significant effects on the protocol behavior. Apart from programming errors in porting, this task can be made difficult if the modem hardware have additional complexities which make one-to-one translation difficult. We have developed a software framework to address this. In this structured framework, identical C code for the MAC protocol runs in both the simulator and the modem. The simulator captures the essential behavior of the modem and uses the same software interfaces as the modem. Since the same code is used with no change, performance comparisons between simulations and sea trials become meaningful. We present a case study using the ARL OFDM modem operating in the 31.25 kHz centre frequency, where one MAC protocol is simulated and also operated with no code changes in the modem. The authors aim to make this framework publically available, whereby other researchers can develop MAC protocols that will run in an underwater modem with no modifications. This framework makes a modem trial an easy step after simulation study. The aim of this paper is to provide an overview of the key aspects of this framework.

I. Introduction

There are many commercial and university based acoustic modems in use around the world, and they use many forms of MAC protocols - both scheduled (such as TDMA) and random access. Improved underwater MAC protocols for varying deployment scenarios are constantly being investigated in many research institutes and commercial entities. The primary mode of investigation is through simulations using network simulators such as NS2 etc, with proprietary modifications to model the underwater channel. Simulations are used since the design/debug/validate cycle is less costly. Some of the promising protocols can then be implemented in modems for sea trials. This typically involves porting of simulation code into the appropriate programming language and software environment in a modem. There are many challenges associated with such a porting process. One of them is to maintain exact algorithmic match in the modem implementation and the simulation code since minor variations in the protocol logic can have significant effects on the protocol behavior. If coded separately based on simulation code from another environment, there could be significant costs in terms of time and effort to develop it.