Dr. Hitesh Tewari

The global cloud productivity market exceeds $50 billion annually, yet dominant platforms such as Google Workspace and Microsoft 365 retain provider access to document content, enabling data mining, AI training on user intellectual property, and forced disclosure under the U.S. CLOUD Act. InvizCrypt eliminates this risk entirely, client-side encryption ensures no server ever sees document content, making it structurally impossible for any provider to read, mine, or surrender user data. Starting with collaborative research writing and LaTeX, InvizCrypt is building a full encrypted productivity suite for individuals, enterprises, public institutions, and governments that demand genuine digital sovereignty. InvizCrypt (www.invizcrypt.com) emerged from a recognition that the research community and the wider population of individuals and organisations handling sensitive documents have no credible encrypted collaborative options. Every major cloud platform handling research writing or sensitive documents in general e.g., Microsoft 365, Overleaf, Google Workspace etc., store document content in plaintext on provider-controlled servers. This is not a gap that policy commitments or contractual assurances can close; it requires a fundamentally different architecture. Building InvizCrypt has meant solving the hard problem of delivering real-time multi-user collaboration while maintaining a server-blind design, where the platform operator genuinely cannot access document content, and not merely claims not to.

Social media networks have become a part of our everyday lives where we share ideas, thoughts, photos etc. with others in our network. However we are also sharing all of the above with the organizations that control the networks that we are using, and giving them carte blanche permission to use our personal data as they fit! Increasingly users are realizing the downside of such "free" services and looking to other networks where their right to anonymity and privacy is put before company profit. In this project we aim to develop a novel architecture for a blockchain based social media network. Specifically we wish to explore the use of "community shards" which hold pointers to the user data stored securely on third party servers or clouds such as IPFS. We envisage that our blockchain will consist of one main (vertical) chain, off which there will be many community (horizontal) shards. This will enable us to limit the length of the main chain and only users that are interested in a particular community will need to download the corresponding horizontal chain. The community will also have the ability to up or down vote posts thereby allowing the most popular posts to be widely circulated and for posts with "fake or false" news to be flagged and removed from the platform. There could even be an incentive mechanism to pay content creators for popular posts. Each community will be administered by a set of nodes and we will make use of smart contracts and proof-of-stake (PoS) protocols to elect their members. These nodes will be responsible for the policies of that particular community and will decide on who can or cannot join the community. In the event of a user violating the community policy the administrator nodes will have the ability to remove a user from their community. Such a decentralized network would give power back to the users that are the core contributors of the system, and eliminate large centralized organizations that control their data. We will build a prototype system modelled on a well-known social media network, and measure parameters such as how quickly the blockchain grows with an increasing number of communities and users, the latency on the network in fetching data from external servers etc. The learnings from this project will hopefully inform the development of the next generation of decentralized social media networks.

Today when a wind turbine produces 1MW of electricity, it can be bought by a large energy users such as data centres using an instrument called a "Guarantee of Origin Certificate", so that companies can achieve their net zero targets. The problem with the current system is that these certificates are usually generated up to a week after the electricity has been produced, and it can be up to a month before the certificates are bought by a company. So in practice there is no correlation between when the electricity was generated and when it was actually consumed. In fact, the company has no idea if the energy they consumed was produced from a renewable source or not. In this project our aim is to create a "real-time" system in which the certificates are consumed within minutes of the electricity being generated, such that there is a real correlation between the electricity that is produced for example by a wind turbine and subsequently consumed by a data centre. We do this by making use of a Distributed Ledger (commonly referred to as a Blockchain) that allows all participants in the system to quickly and securely share information, thereby allowing companies to have confidence that they are consuming electricity from renewable sources.

This project looks at designing a 5G trusted ecosystem architecture and simulation model which enables 5G network slicing support for diverse services by deploying a distributed General Slicing Provision (GSP) solution. Each operator and network provider will be in charge of managing their own GSP instance. An innovative trusted scheme will enable the overall slice allocation and management for diverse service types. Although the proposed solution is general, the project will focus on assessing its benefit in a concrete use-case: vehicle platooning. The project aims to achieve the following outcomes: 1. Proposing the overall architecture 2. Designing a distributed solution incorporating aspects from the blockchain concept 3. Testing the proposed solution using a simulation environment in the context of a concrete use-case (i.e. platooning).