Viktor Tron, Dr. Aron Fischer: Swarm Incentive System for Exchange, Guarantee and Fraud

Reporter: Cathy

A robust peer-to-peer content storage and retrieval system requires a careful balance of the needs of service providers and consumers. Based on the new opportunities provided by Ethereum, the Swarm network is uniquely capable of meeting these needs through micropayments and smart contract implementation incentives.

The talk introduced the Swarm incentive system, which is built around three pillars we call Swap, Swear, and Swindle.

The Swap system records bandwidth usage and provides compensation to nodes that serve content. The dynamic nature of the system means that popular content will automatically be more widely distributed and more quickly accessible. Micropayments are handled by custom checkbook smart contracts, and a full payment system is being actively integrated into the system.

The Swear contract handles long-term storage, allowing nodes to sell their storage space while allowing others to store their content reliably and long-term on the swarm.

Finally, the swindle contract consists of a litigation engine for resolving disputes. The nature of the swindle contract makes it a natural candidate for a state bypass adjudicator, and we will discuss topics such as off-chain state bypass decision contracts, circular payments, and litigation dispute resolution mediation. Therefore, the discussion of Swarm will touch on multiple topics in the current Ethereum ecosystem and the research happening around it.

Swarm nodes and data blocks have their own IDs. If you need to get your own information, you need an address. This is the way of point-to-point information transmission. The address is stored in the key space. There is always a node that is closest to the address. At this time, the request can be sent to the node. This is the process of how we transmit information.

About Swarm incentives: This is a point-to-point content transmission system, so in this case, it is possible to ensure the content given to this system, or how much content is obtained. We cannot pay for each database, which will take a long time, and several payments together will cause the blockchain to swell. So we introduced the checkbook smart contract, which can pass Swarm off-chain and then dock the final check. Comparing the checkbook and the payment channel, the payment channel is the Raiden Network, and the checkbook has relatively little obstacles, because in this checkbook, everyone can use this checkbook, and in the payment channel everyone must enter the network. However, the disadvantage of the checkbook is that there will be repetitions, so payment cannot be guaranteed. Combining Swarm and Swap can be programmed.

There is a certain incentive system now. Content can be delivered through nodes, and Swarm CDN will automatically scale up and down. Only data nodes with high browsing rate can be charged. Others that are not in demand will be deleted.

Each node will save the data with high popularity, but you need to pay for it. If it is not very popular, it will be deleted over time. Through this incentive mechanism: others give information, you pay.

Victom: About the storage of information.

The incentive for storage is completely different from the incentive for transfer. Storage is a promise, not an action that can be executed immediately. So for swap, popular content can be restored quickly, but no one can guarantee that unpopular content will exist forever, and someone will delete it. So we have to find a way to solve this problem.

First, to ensure that the content you want to store will not be deleted over time, you need to pay the node to save the content. For example, if you want someone to take care of your child, you have to pay, but you don’t know who to find, at least they have to be qualified and controllable.

The basic idea is that we need to pay for the storage of data. Payment can be made in installments, such as a babysitter who may have to take care of the child every day, but you can pay after a period of time. So we need a contract to ensure the security of content storage. At the same time, when restoring, we can also ensure that all data and content are obtained.

The storage fee can be paid in installments. If the payment is delayed, it will constitute a condition for a third party to keep it. That is, the custodian can send a valid custody certificate to obtain fees from the third party. Therefore, another concept is that the entire process can be operated off-chain and can be integrated into the payment channel.

So we need a payment channel to ensure that the custodian has the custody receipt. However, information loss is bound to cause losses, but it will be punished. We need other methods to punish, that is, swarm, this contract can be registered as a long-term storage node. First pay a deposit, which is what we call a swear contract. If you want to promise not to lose the content, otherwise you will lose the submitted deposit.

Regarding how to obtain data, we have discussed recovery in swarm. The recovery of this information is the same as our synchronization. This synchronization means that the data block must be kept in the node, and the address of the saved node is closest to the data block ID address. How to reach the nearest node. There will always be a node closest to the data block, so the information needs to be transmitted from one node to another node, and then reach the nearest node. The whole process is carried out through node transmission. If we insure the storage, how can we further protect the storage?

In every transaction, we will have an invoiced swap that is executed in the registered node. So every node can use this receipt, and if you have this receipt, you can appeal. For example, if your insured data is lost, the insured will lose the deposit, and the data owner can also sue such storage.

The prosecution is, for example, that the storer was supposed to store a data block, but failed to find the data block, and then he lost the deposit and was sued.

In this case, the storage can defend itself by either helping to find the data block, such as if the babysitter lost the child, then he had better find the child online.

In fact, the final prosecution requires a final result, and the system has an interesting property. As for the storage commitment contract, it has been signed successfully at the same time as uploading the data. At this time, we don’t need to know who the real storage is. For example, if you go to an agency to find a nanny, you sign a contract with the agency, not the nanny. At this time, you can define the responsibility for storage. So, this is swindle. There is no clear translation yet. Swindle is used to protect data blocks and prevent loss, and then it is guaranteed by deposits, prosecutions and third-party custody. If the storage cannot provide proof, the judge will determine a certain penalty. This is an Orange Book related series in the Ethereum field. If you have any comments, please tell us and let us continue our research.

Regarding the current status and use of Swarm: We are conducting key tests, including its scalability, etc. We use Microsoft's Azure cloud platform and hope to test it in three regions and one hundred nodes.