IPv6
The world has
run out of unallocated IPv4 addresses. Yet the
transition to IPv6 is still slow. While IPv6 deployment has significantly increased over the last few years,
the growth has been very uneven around the world. In some countries, like the USA or Germany, the
fraction of end systems that use IPv6 is well over 15% (Sep 2015), but in many other countries the fraction is very small
according to
earlier work we showed that
a large part of the allocated IPv4 space appears to be unused. Another factor may be an increased deployment of Network Address
Translation (NAT) -- something we will investigate in the near future. This project also aims to measure the IPv6 readiness
of Internet hosts, for example through
IPv6 readiness measurements,
and identify barriers for IPv6 deployment. Another goal of the project is to identify issues and develop improvements for
the IPv6 protocol suite. Work is underway, in a new research project, which will
Survey the State of IPv6 Deployment in Australia and China
Covert Channels
Network steganography is the art of
hiding information in network protocols.
The purpose of network steganography is to create hidden communication channels -- covert channels. These channels can be used
by criminals or hackers to hide their communication or ex-filtrate data, but they can also be used by dissidents or whistle
blowers to circumvent detection/censorship and avoid prosecution. More recently malware, such as botnets, also use covert channels
to hide the traffic between bots and control-and-command servers. Covert channels can also be used for authentication, attacks on
anonymization techniques or network traffic watermarking. The project has a number of goals. First, we are working on a
characterization and classification of existing covert channels, for example in previous research we proposed to
group information hiding techniques into patterns. Secondly, we carry
out research to identify previously unknown covert channels in existing and future networks, protocols and applications. Thirdly,
this project looks into how covert channels could be used for privacy protection. Lastly,
we are researching effective countermeasures for eliminating, limiting or detecting covert channels.
Internet Measurement
Network Meaurement
Timely and accurate studies on the composition and nature of the Internet are crucial for continued research and innovation. In the past we have captured anonymised, and analysed Internet traffic entering and exiting a university network. A range of Network and Transport layer statistics, such as the mean and distribution of packet sizes DCSP, SACK, ECN, MSS usage, Window Scaling and Timestamps were also investigated. Active measurement techniques have been used to determine Round Trip Times (RTTs) and the number of lost and misordered packets. The limitation of this past
study is that it provides a brief snapshot of network parameters in time. In the future we plan to perform the same type of network analysis and measurement in a continuous and automated manner, providing the opportunity to track changes over time. This page will track this project and may be used to release summary results.
WiFi Measurement
Over half of the transmission time in WiFi networks is dedicated to ensuring that errors are corrected or detected. An increased understanding of why frames are lost is a pragmatic approach to improving real world 802.11 throughput.
We
studied the factors which correlate with packet loss and found that about a third of WiFi frames are lost and require retransmission at the link layer.
The results suggest that frames, which spend less time in the air, were less likely to be lost but at proportionally lower rates than simple channel error models suggest. Frames transmitted at high data rates, were also more likely to be successful than those transmitted at low data rates. The size of the packet, in bytes, did not significantly effect the probability of loss. There are implications for a number of research areas.
Reducing the airtime by 10 times, on average, reduces the loss rate by approximately 10%. As the results suggest that the frame size has little effect on the loss rate, larger wireless frames, may have a highly beneficial effect on overall throughputs.
Future research may attempt to see the effect of 9000 byte and larger frames. Other ideas are to investigate the extent to which the packet size effects the loss rate in 3G and 4G networks. Ultimateny throughput benefits, when using a larger frame size over the internet would be
significant.