Dependable Computing over the Internet

It is hard to exaggerate the importance of the Internet in our daily lives today. Internet has influenced every aspect of our lives including business, education, shopping, and entertainment. Increasingly, people are using Internet for accessing different types of information and for performing different types of computation using this information. It is expected that the use of Internet will continue to grow exponentially over the next decade, particularly as a vehicle for computing.

The main goal of our research is to investigate design and implementation techniques at the system level to promote dependable computing over the Internet. Despite the ever-increasing popularity of the Internet, the task of performing any reasonable computation over the Internet is still very complicated for most users. There are four important reasons for this complexity:

Internet is unreliable: transient and permanent communication failures, communication partitions, and node failures are common in the Internet.
Internet is insecure: information transmitted over the Internet is vulnerable to sniffing and modification over the communication links, and at the intermediate switches and routers.
Internet is heterogeneous: computing components in the Internet differ from one another in their hardware and software specifications.
Internet is typically accessed via low-end computing devices that are limited in their processing capability, have low-bandwidth connections, and are mobile in nature.

At present, we have three research projects, each addressing one or more of these reasons. The details of these projects can be found the following links: