Papyrus will cope with blockchain scalability issue
It goes without saying that the blockchain is a fashionable trend in 2017. There are a lot of projects in adtech industry that aim to disrupt the way we interact with advertising. In a nutshell most of the projects claim that they will change the whole user experience regarding all the Internet, though these claims might be overvalued.
One of the biggest problem for digital advertising is ad fraud. The price for that is quite clear — over $16.4 billion and it’s still growing. In digital advertising the transfer of value goes through the following axes: advertiser-publisher-user. Advertisers bid for ad placements, publishers accept the bid, the ad impression is delivered, the ad is filled and the user’s ID is matched with that ad impression ID.
As blockchain is essentially a decentralised storage for all the transactions, and it has proven efficiency in finance and multiple other industries, then it could be a great idea to use in digital advertising, so that the value can be saved and not wasted if we implement mathematical proof and decentralisation, right?
Actually, there’s some doubts that blockchain can change the situation. Recent article by Forbes presents some concerns regarding it. The argumentation is simple: up to this date blockchain technology is very slow to handle that much number of transactions. For instance, Bitcoin’s blockchain process 7 transactions per second, whereas Ethereum’s one — around 13 transactions per second. In case of advertising, a single transaction occurs every time as ad is viewed. With millions of as impressions being delivered every second, it’s just not possible for the current technology development to process that much information.
So the question is really important. Why the Papyrus digital advertising platform will be a game changer?
It’s true that large ad networks and publishers generate hundreds of billions of ad clicks and at least two times more ad impressions annually. Using current technology any attempt to execute smart contracts for digital advertising will generate excessive costs. Our solution utilises an extension of the technology called state channel that is capable of supporting these high loads.
The idea of a state channel is not new. It underlies the scale-up solutions for Bitcoin and Ethereum ( as well as Lightning Network and Raiden Network). Papyrus is one of the first to customize it for the specific needs of AdTech. The technology aggregates state logs from DSPs, SSPs, and auditors. So the realisation of payment is based on the constant log exchange among parties.
In the typical state channels system the transaction takes place when the critical mass of obligations occurs (e.g. Bob owes Alice an amount of $). In the case of adtech, such a system would be feasible because of the need for a revision of impressions/clicks by an external auditor that will come later on. In the Papyrus ecosystem the auditor takes a direct part in the formation of the payment. This system is more efficient for adtech as all parties receive payments after all verification procedures.
Here is an example:
- Participants make deposits in the locker (thereby fixing some of its state) and then directly exchange with each other payments in the form of appropriately drawn up and signed transactions of the relevant block. This changes the balance of deposits in favor of one or the other participant.
- Participants transfer the updated state of the blockchain (in the part related to their deposits) to each other and they can do this arbitrarily (of course, within the limits of deposits) without the costs of carrying out transactions in the detachment.
- The final state is recorded in the blockchain only once — when the channel is closed. Or if one of the participants finds the incorrect behavior of the other. Let’s say that the participants will send about 200 thousand micropayments (for a total of $ 200) through one channel. In our initial assumptions this gives several tens of millions of channel opening / closing transactions per month. This is not trillions of transactions per month, but still too much for immediate processing in a public blockchain.
Papyrus has another solution: we can develop the idea of direct transmission of updates to the state of the blockchain outside of it and apply this approach directly to the opening / closing logic of the channel (originally performed in the blockchain). Let’s look at the opening / closing transactions for the dedicated nodes that will exchange the calculated updates of the public lockdown state with each other only occasionally publishing them to the blockchain itself (to update the balance of payments between the participants). The work of such nodes can be organized as a highly specialized block based on the Proof-of-Stake consensus where the blocks consist exclusively of opening / closing transactions. Participants will make large PoS deposits in a public blockchain and will be penalized for improper handling of these transactions.
To ensure guarantees of invariability and verifiability a specialized blockchain can lean on a public block writing down hashes of its blocks of transactions. Suppose one block of a specialized blockchain contains about 1 thousand transactions of opening / closing channels — in our initial assumptions this gives several tens of thousands of transactions in a public detachment a month. It already looks realistic, for example in the Ethereum network the current daily volume is hundreds of thousands of transactions.
The use of state channel libraries in Papyrus will facilitate the opening and maintaining of a peer-to-peer state channels network. Leveraging the developments of Lightning Network and Raiden Network the Papyrus ecosystem will introduce state channels adopted for usage in decentralized RTB protocol. The overall idea under that is to enhance the scalability of the Papyrus ecosystem.
In the first half of 2018 we are planning to develop and test Papyrus Protostar Release with:
- Libraries for dRTB Event Log Storage
- Libraries for dRTB State Channels Network
- Libraries for Auditor integrations and more.
With 20 connected partners (at the moment) we will present a first approximation of our future digital advertising platform. This is about real work, our passionate and talented team, and the deep desire to make advertising better. Stay tuned!