My Summer Schedule- 2006

I will be working as a summer Intern in Pervasive Platforms and Architectures Lab at Motorola's Research division.

Privacy Issues with Web2.0

The concept of 'Software as a Service' is the driving force behind the so-called Web2.0. This emerging paradigm will lead to lots of user data being gathered at web servers of the service provider. Google's upcoming GDrive is one such service that is in the offing. Imagine users storing their personal data at Google servers with Google's assurance that their privacy will be preserved. But what happens if the Governement decides that it needs to have access to that data and force Google to make it accessible ? This would obvioulsy mean a serious blow to the user trust in a specific technology. So what about storing the data in encrypted format? Problem is that the encryption/decryption of the data will still be done at the server side making it a potential weak link. What if the employed cryptosystem allows users to do the encryption and decryption in their local resources and then store the data in encrypted form at the server. In a naive implementation, this would lead to serious performance issues. But our work on Homomorphic cryptosystems, suggests that certain operations can be performed on the ciphertext such that corresponding functions will be automatically applied to the underlying plaintext. This has useful implications in allowing users to efficiently manage encrypted data on the remote untrusted servers.

Warner Bros. realise the potential of P2P Infrastructure

WSJ reports that Warner Bros. will be using BitTorrent Inc.'s peer-to-peer technology to distribute its movies to legal subscribers. P2P technology has been wrongly percieved as an enabler of illegal distribution of copyrighted data. I have been strongly advocating the use of P2P Infrastructure for scalable, secure and authenticated dissemination of multimedia content. Our research work in context of the project TRIUMF has focused on delivering efficient and practical solutions in this regard. SAMcast protocol is an ongoing research effort at Multimedia Research Lab at UIC. This protocol has following features

• It employs highly scalable P2P communication primitives.
• It uses an efficient content aware assymetric cryptographic protocol for secure dissemination of multimedia data.
• Novel finger printing techniques coupled with efficient water marking procedures provide collusion resistent authenticated content delievery.
  
• Highly efficient and high performance incremental hashing schemes are used for authentication of the multimedia content.
• An incentive based approach encourages users to allow their network resources to be used for P2P commnication based content delivery.

We have also developed a novel solution to the secure aggregation problem in large scale overlay networks. P2P networks have tremendous potential in creating vibrant social communities which in turn provide exciting business opportunities. It is all upto us to create technologies which keep users' trust in the network and in the ever-booming online communities.