Pantos Project Update: The way towards instant cross-blockchain communication

Pantos
Pantos
Dec 16, 2020 · 14 min read

Read this article in German.

While the general objective of the Pantos project hasn’t changed since its inception over 2.5 years ago, the industry as a whole definitely has. In late 2017 and early 2018, the goal was to simply connect fragmented blockchains. Now, we see a lot of those projects coming to maturity. We also see the traditional financial industry embracing the possibilities of blockchain technologies. In this blog post, we would like to give an overview of what has happened so far and where we are heading further on. We are also addressing ways to eventually reward PAN token holders.

Over the past 2.5 years, we already achieved a lot. But guess what… the future will be even more exciting. That’s why we’ll keep this part short and give you a concise overview, including the most important milestone we hit so far with the Pantos project:

Participants in the Pantos Technology ICO contributed 616 BTC in total during the 4 week long ICO phase.

We published the first Technical White Paper on Token Atomic Swap Technology (TAST), which was followed by nine more Technical White Papers between May 2018 and October 2020, providing the scientific groundwork necessary to go from a purely research-based phase to a product focussed phase.

The release of the fourth TAST Technical White Paper marked the end of the initial research phase. With this update, we also released the first prototype for Pantos’ technology based on our scientific findings.

Having completed the basic research on the requirements for cross-blockchain token transfers and the publication of the DeXTT prototype, the first TAST research prototype, we continued to improve the prototype. With it, we were able to present a fully-fledged solution on how cross-blockchain token transfers could look in practice.

After 2.5 years of research and development, we reached a major milestone by entering into a partnership with Raiffeisen Bank International (RBI). In the scope of this partnership, Pantos technology will be made use of for RBI Coin, an e-money solution developed by RBI. This milestone also comes with structural updates, a more distinct separation between research and development. We will not publish any further White Papers for the time being and will instead make more use of the blog post format.

A long time in the making and finally opened at the end of November 2020, we announced a new Christian Doppler Laboratory for Blockchain Technologies for the Internet of Things (CDL-BOT) led by Prof. Stefan Schulte. Together with our new partner IOTA, and with the support of the Austrian Federal Ministry for Digital and Economic Affairs and the Christian Doppler Gesellschaft, the research focus of the new laboratory will be on Internet of Things (IoT) applications of Distributed Ledger Technologies (DLTs) and blockchain-interoperability.

Current Prototype and Outlook

For this outlook, we start from a prototype illustrating the feasibility of decentralised, near-instant cross-blockchain communication. We now consider our earlier approaches (DeXTT in particular) to be blueprints for cross-blockchain balance synchronisation. The Testimonium relay can be used to link different blockchains as long as both allow specific blockchain logic, commonly referred to as smart contracts. For our work on token transfers and cross-blockchain contract invocation, we use the abstract notion of transaction inclusion verification. Relay solutions are specific examples, allowing transaction inclusion verification. The most prominent other example is centralised authorisation as used, for instance, by Tether. This abstraction is to highlight that the protocols developed do not depend on the specific transaction inclusion verification, but also work with other cross-blockchain communication mechanisms yet to be specified. For the remainder of this text, we will focus on future work.

Potential rewards and incentives for PAN Holders?

The Pantos community keeps asking for a timeline, for plans to reward ICO-participants and, essentially, for a finished product. Let it be known: we hear you, we feel very similar needs. The Pantos project up until about now was mostly a scientific research project. One of the dogmas of scientific research is that it needs to be independent and hence, may not be driven mainly by economic interests. Scientific research will continue to be an important part of the Pantos project, but at this point in time, we consider the science part in our research to be less pure and practical aspects are becoming a similarly important factor.

We have looked at various possibilities to enable cross-blockchain communication, to enable the transfer of tokens from one blockchain to another. Usually, blockchain transactions are not executed for free and similarly, our approaches for such communication involve off-chain-clients that are paid by other clients for their cross-blockchain services. This could suggest the use of PAN as a payment currency and thus, an intrinsic use case for the PAN token.

There are two substantially different approaches to enable cross-blockchain communication, from a general point of view and a specialised point of view:

  1. General approaches are important to look at for starting points. They aim at the most encompassing and elegant solution, which as a side product, also solves special cases such as token transfer. Such general solutions are more likely to be independent and rely on as few parameters as possible. One problem with general solutions is that they tend to lack incentives for maintaining the infrastructure.

Possibility: Earn transaction fees for maintaining the network

In the past, we have mostly worked on general approaches. Cross-blockchain communication as a general question seems to have a tendency of preferring native blockchain currencies over custom tokens. As an alternative approach, the future could also bring special first designs. Some of the possible solutions we have in mind follow principles established by proof-of-stake consensus protocols, where holders of the ecosystem token participate in maintaining the network, all the while making an income from cross-blockchain transaction fees.

Disclaimer: Note that we are currently exploring the ideas represented in this article, but have not yet committed to any particular solution. Our goal is to reward PAN holders in some way, but we can’t promise a certain direction at this stage.

Pantos use cases (besides Pantos infrastructure necessities) are likely going to happen but are not the main focus of our current work. For instance, occasionally we dream about solutions allowing us to use PAN tokens (without holding any native blockchain currencies) to initiate any activity on any connected blockchain. For example, given a dedicated Pantos P2P network, it would then be possible to submit a complex transaction only there. The initiator of this transaction would pay only with PAN, the transaction could involve exchange of PAN from the Ethereum blockchain for RBI Euro on the Billon TLD.

The Future of Cross-Blockchain Interaction

Solutions for cross-blockchain interaction (and cross-blockchain token transfers) can be classified by three main parameters. Decentralisation factor, transaction speed and operational costs. As is natural to such classifications, we strive for optimal values for each of these parameters, but should expect that the prioritisation of one parameter means less optimal values for another parameter.

  1. The decentralisation factor is the intrinsic dogma of the blockchain world. Ideally, blockchain solutions do not depend on centralised entities, allow arbitrary actors to participate in their incentive structures, and make transactions verifiable publicly.

Our previous approach of installing blockchain relays to allow cross-blockchain communication can be considered as optimal in terms of decentralisation. It can be implemented as a blockchain-agnostic technology. From there, the resulting transaction speed is merely a question of security. As a countermeasure to double spending attacks, crypto exchanges usually require a fixed count of blocks to verify some transaction for incoming assets. We have spent a great deal of work on optimising our operational costs. By facilitating a dispute solution, we minimise the amount of information necessarily stored on-chain while securing the network with the security of the participating blockchains as long as at least one honest actor is on board. “Minimised operational cost” that is, if each and every block of the source blockchain is, by requirement, immediately mirrored on the destination blockchain.

With blockchain and cryptocurrencies still in their early stages, it is not clear yet which requirements blockchain applications will come with, and even less clear what kind of communication will be needed across different blockchains. From our point of view, the most pressing question is a quantitative one. How many cross-blockchain transactions can we expect? If almost every block of the source blockchain contains some information that needs to be verified by the destination blockchain, then arguably relay solutions are also optimal in terms of operational cost. If however, such cross-blockchain communication is expected to be far less frequent than the usual on-chain communication, relay solutions, that in some sense replicate entire blockchains onto other blockchains, come with a lot of unnecessary overhead. In other words, pure relay solutions do not have a dormant state.

Batching Solutions

Batching approaches are natural solutions for less dense cross-blockchain transaction intervals. By collecting a number of blocks for simultaneous submission later on, we might not meet the strict relay definition anymore, but prevent the cross-blockchain communication protocols from clogging their host blockchain networks. By using a batching approach however, it is evident that we lose the property of near-instant, cross-blockchain communication. Validation takes time. There is a difference between validating only one block at a time, or a collection of many blocks. The benefit of batching solutions is that they allow for a dormant state, where the cross-blockchain network can be activated only when needed.

For instance, zero knowledge proofs allow compressing the required on-chain verification to feasible amounts. Zero knowledge proofs are mathematical constructs that allow verification of some solution without knowledge of the details of its proof. For block header verification, this solution could, for instance, be proof that some block header is preceded by 100 other equally-valid block headers. The verification of this statement ideally does not require all 100 headers as input but only the first and the very last block header. Similar technologies are used for privacy blockchains, such as Monero and ZCash to allow zero knowledge transactions that can be shown to be valid without third parties gaining any insights on the amount or account details. The particular problem we face for our needs is that computing zero knowledge proofs seem to be computationally hard and thus expensive in terms of electricity cost.

Another batching approach would be to give more power to our dispute mechanism. So far, we have introduced single step disputes. Submitted block headers if falsifiable can be rejected in a single step. Multi-step dispute systems, as for instance discussed in work already completed on Dogethereum could allow such disputes to turn into a sort of discussion, where one party proposes some block header, another party doubts its validity, the first party provides further information necessary for verification and so on. Naively, this could mean that a dispute would result in tracking a block header back to the last generally accepted block header, one by one, with two separate transactions for each previous block header. This approach would require sophisticated collateral and incentive structures. It also requires a detailed game theoretic analysis of the security issues encountered, but would allow for, arguably, minimal transaction costs. For the case of token transfer, taken to the extreme the initial statement (to possibly start a dispute from) would then be the mere claim on the destination blockchain.

As a last contender, for the time being, we also want to highlight the concept of stateless SPV proofs, which is used to allow more efficient communication from the Bitcoin blockchain, to the Ethereum blockchain. Here, instead of continuously submitting block headers, each requested transaction simply comes with a sequence of block headers. The authenticity of such a transaction then relies on the difficulty of computing such sequences. It remains to be seen whether similar solutions are possible for other pairs of blockchains.

Pantos network on a carrier blockchain

In some sense, relay solutions focus less on operational cost optimisations. Similarly, batching solutions can be said to care less about speed requirements. As a means to reduce operational costs, we will now take a look into a delicate matter, a loosening of the decentralisation dogma.

As a starting observation, it should be pointed out that an ideal operational Pantos network comes with several characteristics. For the sake of simplicity, we consider the abstract notion of an “asset” to refer to the smallest unit of the PAN token.

  • At any given time, each asset is owned by one actor at most.

We observe that the Pantos network itself can be seen as an abstract notion of a ledger, a virtual blockchain in some sense. With this in mind, we turn our flashlight of interest to various levels of centralisation.

The most centralised solution is arguably to implement smart contracts on each connected blockchain that require signatures by authorities for any cross-blockchain interaction. In practice, this could be Bitpanda and/or other trusted sources. Such authorities (or oracles) are optimal in terms of speed and operational cost. By using authorities, it might even be possible to cut the native blockchain time limits imposed by the block times short. Authorities however, thwart the decentralisation dogma in a spectacular manner.

By using a collection of trusted authorities, this centralisation can be slightly defused. A collection of authorities or delegates can be voted upon. A delegation voting process can be achieved via blockchain technology. In other words, we could utilise an existing but alien yet suitable blockchain to serve as provider of authority permissions. So, part of our work for the coming months will be to investigate and evaluate various blockchains for their applicability as carrier blockchain for the Pantos network. As pointed out by Waterloo, bridging, for instance, Ethereum and EOS in their current states seems to be economically feasible. The delegated staking used by EOS allows for some efficiency in that regard. The danger of using an alien carrier blockchain is a danger of dependence, thus counterfeiting the idea of blockchain agnosticism.

PAN should play a vital role in transactions in the Pantos network. Our prior work seems to indicate that native cryptocurrencies might be favourable over foreign tokens for incentive structures. Otherwise, the need for conversion rates and additional cost for conversion come into play. This insight combined with the idea of utilising an alien blockchain for our purposes leads to the question of whether a dedicated Pantos blockchain would be a good idea. The benefits of such a blockchain include: being able to tailor the inbuilt logic to the cryptographic needs required to cooperate with several other blockchains, to create a natural incentive for paying fees with PAN or holding PAN for staking purposes and being able to optimise speed and validation mechanisms to suit our needs. A Pantos blockchain would further allow a simple mechanism to use PAN to pay for any action on any other connected blockchain.

As a closing remark, we want to highlight that the thought of alien carrier blockchains or a dedicated Pantos blockchain might turn into a solution, but should also be seen as some form of placeholder. Ideally, the Pantos network turns out to become a virtual blockchain, a ledger that makes use of several other, independent blockchains, and takes advantage of the benefits of each connected blockchain in a symbiotic way. This is where finally cross-blockchain balance synchronisation (as investigated in our first White Papers) might come back into play, as a means to allow more expensive but less frequent authority updates.

Project state and roadmap updates

We are currently in between testing, reevaluating and improving prototypes and further researching. From a wider perspective with this blogpost we enter a new phase where part of our work becomes actual development. At the same time the evaluation work of the coming months will allow us to make first steps regarding the actual implementation we choose to enable cross-blockchain communication. For the time being we do not have further Whitepapers planned, the scientific research will rather be published using more conventional scientific venues. We are currently actively discussing the future Pantos outreaching strategy. Community communication has always been an important part of this outreach work, among other formats we consider regular Twitter updates and in depth blog posts. High priority community questions could be a starting point for the former, topics like “multi-step dispute schemes” or “the virtual Pantos network” could be starting points for the latter.

Glossary

  • Batching: To increase efficiency it is a common practice to partition complex tasks into smaller tasks to be solved by independent entities. A batch then is a package of work forwarded from one entity to the next. For computational programs in particular often it is cheaper to transfer a batch of information for combined processing, instead of transferring and processing the smallest units.

Pantos

The first multi-blockchain token system.