Underwater Acoustic Sensor Networks (UASNs) are an enabling technology for many underwater applications. A typical UASN scenario is depicted in the adjacent figure. Acoustic waves are the preferred choice for underwater communication, since radio and optical waves suffer underwater from severe attenuation and pointing precision difficulties respectively. Use of acoustic waves underwater place extreme constraints to the functionality of Medium Access Control (MAC) protocols. The underwater environment features limited and noise-dependent bandwidth availability and excessive variable propagation delay. This poses challenges to the design of MAC protocols including attempts to achieve low end-to-end delay, high channel utilisation, fairness and low complexity. The functionality of MAC protocols should be stable despite synchronisation difficulties, space unfairness and bursty short-packet traffic. The most common way to achieve synchronisation underwater is the use of a global scheduler, guard intervals and exchange of timing signals. To this end, single-hop topologies suit UASN applications very well. The performance of a MAC protocol is determined by the ability to adapt to different underwater scenarios (including distinct characteristics with respect to water motion, signal attenuation, background noise, interference etc) and non-environmental factors such as (the network topology, data rate, hop length, network size, packet duration, etc).
My PhD research is concerned with the issues surrounding the use of existing MAC protocols in representative underwater scenarios, based on a number of underwater sound propagation models. New approaches have been developed which are capable of adapting to the time-varying channel conditions in order to provide high throughput capability and controlled delay. Riverbed Modeller is a network protocol design and simulation tool which has been used in this research to create a model of the underwater acoustic channel. A number of its pipeline stages have been modified to reflect several underwater propagation mechanisms. It has been found that most traditional protocols suffer in performance due to the peculiarity of the underwater environment. The design of an effective underwater MAC protocol should in particular deal with the bandwidth utilisation versus end-to-end delay trade-off. Combined free/demand-assignment multiple- access (CFDAMA) is of particular interest as it exhibits significantly enhanced performance with respect to minimising end-to-end delay and maximising channel utilisation underwater with excellent adaption to changes in data traffic conditions. My research provides new insights into the use of the CFDAMA schemes underwater, including its underlying individual assignment strategies and with different data traffic types. It proposes new schemes based on CFDAMA that suit UASNs.
Members
- Wael Gorma
- Paul Mitchell
- Yuriy Zakharov
Funding
- Libyan Embassy - London
Dates
- June 2015 - May 2019
Research