Improving the performance of low-bandwidth communications networks for disaster zones and remote areas

Author: Ghassan Al-Nuaimi

Al-Nuaimi, Ghassan, 2022 Improving the performance of low-bandwidth communications networks for disaster zones and remote areas, Flinders University, College of Science and Engineering

Terms of Use: This electronic version is (or will be) made publicly available by Flinders University in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. You may use this material for uses permitted under the Copyright Act 1968. If you are the owner of any included third party copyright material and/or you believe that any material has been made available without permission of the copyright owner please contact with the details.


Telecommunications forms a vital part of many aspects of modern life. However, telecommunications access is not available to all. In particular, maintaining access to effective telecommunications following disasters is particularly problematic, as the infrastructure on which it demands may be damaged, destroyed or overwhelmed. Similarly, people who live in small isolated communities are often not able to access telecommunications, because it is not cost-effective to extend cellular coverage to these communities.

This is particularly true in logistically difficult environments, such as in Pacific Island Nations, where communities may live on islands that are many kilometres from the next nearest island or community, and where the islands may be mountainous. Add the difficulties of frequent cyclones, the risk of tsunami, the corrosive tropical maritime climate and the relatively

small economies of Pacific Island Nations, and it is no surprise that many people in the Pacific do not have reliable access to telecommunications, even before a disaster strikes. The problem, of course, becomes much worse when a disaster does strike, and may contribute to the loss of life.

One particularly troubling aspect of this is when the lack of access to telecommunications services means that coastal communities in the Indo-Pacific do not have access to a reliable tsunami early warning system. The lack of effective early warning to tsunamis is a frequent contributor to the loss of many lives when tsunamis do occur.

This thesis seeks to mitigate these problems by enhancing the Serval Mesh, in particular, the Low-Bandwidth Asynchronous Rhizome Delivery (LBARD) component, to support the automatic relaying of mobile-phone originated messages via long-range HF radios, as well as overseeing many internal improvements to LBARD, that improve its efficiency, such as the inclusion of the new TreeSync bundle list synchronisations protocol.

Further, in the case of the need for tsunami early warning systems a design is proposed for a low-cost, small and easily maintainable tsunami early warning system based on a combination of the Serval Mesh and an innovative satellite receiver. A proof-of-concept for this system is presented, where a ground-station is used to feed low-latency messages for broadcast delivery using a geostationary satellite, which can be received using a small dish-free and easy to aim receiver. This concept is then further developed to envisage a complete low-cost multi-hazard early warning system that is not dependent on any fixed terrestrial infrastructure, and in particular, is composed entirely of small, low-cost devices that are much easier to work with under the logistical challenges that exist in Pacific Island Nations.

Collectively, these innovations pave the way towards the provision of free and resilient basic telecommunications services – including both text messaging and early warning of disasters like tsunamis – using only low-cost systems, and building on the past pedigree of the Serval Mesh, including past field trials in the Pacific.

Keywords: Serval, telecommunications in remote areas, low-bandwidth networks, Early Warning Systems

Subject: Computer Science thesis

Thesis type: Doctor of Philosophy
Completed: 2022
School: College of Science and Engineering
Supervisor: Paul Gardner-Stephen