Insert Photo Here
Senior Research Scientist
CyberPhysical Systems Research Program
phone: +61 2 9490 5611
Todays most applications rely on pervasive service residing in various location of the networks. This raises several conflicting requirements including security, privacy concern and overall performance. Making sure that these applications maintain an acceptable Service Level Agreement while providing these requirement is generally regarded as an open problem for distributing computing. During my career I have tackled this problem by not only developing new distributed algorithms but also by incorporating new functions inside the network in order to better support these applications.
During my career, I have pursued this research excellence through rigorous experimentation, measurements, and implementation of distributed computing systems capable of operating at large scale over a realistic network. This vision led me to focus upon developing large-scale networked systems that avoid the common fallacies of distributed systems through a wide networking knowledge. In particular, I leveraged my network knowledge to not only offer basic support, such as reliable communication, but enhanced services such as a secured and private channel to distributed computing applications. Doing so I have developed several state-of-the-art tools and frameworks that are illustrated by the following three examples.
Distributing Network Functions (2014 – present)
In recent years, I have led the Network Group initiative in leveraging the paradigm shift that constitutes Software Defined Networking and Network Function Virtualisation. This group initiative has been accomplished on two separate directions. In the first one, I focused on enhancing application support through a more intelligent and agile network. In the second one, I have also applied theoretical results from distributed systems, such as the two-phase commit algorithm, and optimisation techniques to also enhance the management of Software Defined Network.
In the first direction, where we used SDN in order to improve application services, I have proposed numerous enhanced networking services, such as secu- rity and privacy preserving techniques, that could not be deployed in nowadays networks due to the legacy lack of network support. These services include:
In addition to the above, I have applied distributed computing concepts in order to enhance how Software Defined Networks are operated. In particular, I have proposed and published in the IEEE Communication Magazine (impact factor of 10.435) the first ever distributed algorithm able to garbage-collect forwarding rules in the presence of an asynchronous network, (e.g. when no delay upper bound can be established).
Rethinking Network-bound Applications (2016 – present)
Networking applications currently use the Berkeley socket model to interface with a networking stack that resides in the operating system kernel. This model requires costly context switching between applications and the kernel, as well as memory copies on both the sending and receiving path. This model imposes a limitation on performance which becomes even more apparent with the doubling of bandwidth of network bandwidth every 17-18 months, compared with CPU and DRAM performance doubling only every 26-27 months. For example, The Memcached application spends over 80% of CPU time in the kernel networking stack, using less than 5% of the available networking bandwidth.
This research aims to evaluate the impact of using kernel bypass technologies, to accelerate network bound applications. Some research questions include:
Cyber-Fraud Detection (2014 – present)
Facebook and other Online Social Networks have become one of primary outlet for advertisements. How can we detect and prevent frauds on these systems and ensure that businesses are getting bang for their buck? In this project we aim to provide new forms of cyber-defence. This will lead us to offer: