In the model physical infrastructure network)DimineThe application, a network of suppliers is assigned to provide a service – weather readings in specific sites, access to wide range, or energy, for example – for the customer. Since the interested suppliers in particular cannot be trusted in the self -report on their status, the question then becomes: How and when can the network provoke the resource condition honestly?
For example, suppose the customer depends on an accessible and reliable physical service in a specific area. The customer may want to verify the presence of the service on this site, and he is enthusiastic enough to provide compensation for providing the service, but he cannot accurately trust the provider of the self -report because it may be one of the cheapest provides to provide the service elsewhere and the site is tried only. This is a problem Check the siteAnd that we will use as an illustrative example all the time.
In new research, we formally distinguish from the two basic decades-manipulation and self-dealing-which stands in the way of checking the level of service provided by the supplier (in our example, its location). More importantly, knowing that this is it only Two obstacles also indicate approaches to how to solve them. Our business framework is the first to provide mainly the types of services that can be stimulated in DePin applications.
Our model
Our model contains three types of participants: clientWhoever asks for service; the sourceWhich claims to provide the service; and ObserversThose who can partially verify the source of the expected service. We focus on the source and observers.
The source of the information in the example of checking our site is the object whose location must be reported. Observers (For example, sensors) receive signals from the source that may be noisy and/or processed by the source. For example, every observer may communicate with the source and measure the time of the circular journey transmission; The source may deliberately delay before responding to contact requests and thus manipulating the calculated distance.
Arrange operations as follows:
- The source chooses a strategy. The person or entity gets the information of the information receives by observers. He chooses a “processing” method that the observers will get, using the permitted options only based on their own information (for example, its real location) and other restrictions (for example, it can use the messages delay to increase the perceived distances but it cannot reduce the perceived distances across the same way).
- Signs are created. Next, the signals are created (for example, the distances between the observers and the source) (perhaps probably) are created using the method that was chosen.
- Every observer only sees their own signal. Observers get only seeing their own signal – they do not know what others have seen.
- Each participant decides themselves their information. After seeing their references, the source and observers participate in the mechanism of deduction/verification, where the source and observers are what they like – including incorrect things.
The goal is to design a mechanism in which the source and observers are stimulated to report sincerely about their own information and signs, respectively.
Two basic challenges have been revealed
Our paper proves that there are two challenges rooted in accurately deduction of information in the signal networks. We explain that these challenges are essential to the problem, instead of just being an artifact for our currently emerging.
1. Tampering
The first challenge is related to the manipulation of the service measures provided, and the fact that any person or even all observers may not be able to discover this manipulation. That is, the source may lie to the observers – and it may be difficult to know when to do so. To stimulate accurate measurements, the benefit of the participants in the network (i.e. how they are rewarded) must be linked to visual behavior in the protocol mechanism (for example, whether the monitors reports are logically consistent with each other). For example, a participant in the network (source or observer) can be rewarded in the original symbol of the project to provide the information requested by the protocol that corresponds to the information provided by other participants.
Our first official theory proves that the problem of stimulating accurate information from the source and observers is not solved unless The preparation satisfies a condition to be “identifying the source.” This condition requires that the source information be completely and always (in the statistical sense) to look exactly as it is everyone Observers under different real source signals; We will say more about this below. We explain that when this mystery is present, there is nothing a protocol designer can do to ensure sincere benefit: the source can always take advantage of mystery to distort his illogical information. This impossible result keeps even when the protocol has local symbols or other incentives at its disposal.
2
In incentive design, sometimes therapy is worse than the disease. In the context of providing decentralized service, there is a special concern that the source may be denied as one or more observers (the so -called Sybils) Protecting or verifying service from themselves, or the team (i.e. collusion) with other participants in the network to do so. In the language of our model, the source may collude with some observers (or itself, under the identity of Sybil) with the intention of exploiting the mechanism of its reward.
Our second theory proves that when this collusion exists, honest deduction of the source information is essentially impossible (unless the source of the source identity is ignored, as the factors that can correspond to potential factors) are ignored.
Good news
Our results show that the lack of identity and self -dealing are essential obstacles to verifying DePin applications. The good news is that we show a sense of it only Two obstacles: We give an official mechanism, according to the assumptions of determining the source of the source and the lack of self -agreement, achieving the full revelation of information in the balance.
How can the Depin protocol designers go forward?
Our results refer to two main lessons of DePin designers. The first is to take steps to ensure that the source identity is possible. This is not as it seems as it seems – for example, in checking the site, the source identity is equivalent to the engineering situation in which the source site lies in the convex body of observers’ sites. So, if you want to make sure that you can honestly get the source site, make sure to “surround” its potential sites appropriately with the observers.
The second lesson is that the DePin protocol designers need to process self -fears outside the protocol. The methods of doing this include, for example, restrictions on entry without permission (and thus justifying stronger confidence assumptions) or random identification between many different sources that may perform a customer service (and thus set a non -zero cost to demonstrate to request service).
***
We hope our work in thinking clearly helps in verification problems and different options to address them. DePin protocols explore these relevant ideas and ideas, and we are excited to find out how to run it.
***
For deeper technical analysis and discuss our model of signal and results, please see our paper.Recovery compatible with incentives from treated signals, with applications on decentralized physical infrastructure“
***
Jason Millions He is a student in the Department of Computer Science at the University of Colombia, where Christos Babadimrio and Tim Rogarden advise him. It is widely interested in the game theory, especially in conjunction with machine learning and Defi. He was also trained at the A16z Crypto Research Lab Lab.
Jens Ernstberger It works at the intersection of encryption and computer security. Recently, he was trained in A16z Crypto research. Previously, a doctorate student was in Munich Technical UniversityUnder the supervision Professor Sebastian Steinhurst and Dr. Arthur Jerefis.
Joseph Bono He is a technical adviser to Crypto A16z and Associate Professor in the Department of Computer Science at the Kurant Institute, New York University.
Scott Duke Commons He is Professor of Business Administration at Saroufim Rock in Harvard Business CollegeAffiliated with faculty members The Ministry of Economy at Harvard UniversityAnd research partner in A16z Crypto. A number of companies also advise on the Web3 strategy, as well as the design of Marketplace and incentives; For more disclosures, see Its location on the web. He is also a co -author A symbol of everything: How will Nfts and Web3 transform the way we buy, sell and create.
Tim Khadrden Professor of Computer Science and a member of the Columbia University Data Science Institute, and the head of research A16z Crypto.
The views of the AH Capital Management, LLC (“A16Z”) are adapted and not A16Z views or its subsidiaries. Some information here was obtained from third -party sources, including from the conservative companies from the money managed by the A16Z. Although it is taken from sources believed to be reliable, the A16z has not independently verified this information and does not provide any representation about the current or permanent accuracy of the information or its suitability for a specific situation. In addition, this content may include third -party ads; A16Z did not review such ads and does not support any advertising content in it.
This content is provided for media purposes only, and it should not be relied upon as legal, commercial, investment or tax advice. You must consult your advisers on these matters. References to any securities or digital assets are for illustrative purposes only, and they do not constitute an investment recommendation or offer to provide advisory services for investment. Moreover, this content is not directed or intended for use by any potential investors or investors, and it may not be relied upon under any circumstances when making a decision to invest in any fund managed by A16Z. (An investment offer will not be made in the A16Z Fund except through the Special Employment Memorandum, the Subscription Agreement, and other relevant documents for any such fund and must be read entirely. There is a list of investments provided by the funds managed by Andressen Horowitz (except for investments that did not provide the source for A16Z to detect public and unannounced investments in the public assets circulating) in https://a16z.com/Investments/.
The content speaks only from the date indicated. Any expectations, estimates, expectations, goals, horizons, and/or opinions expressed in these materials are subject to change without notice and may differ or differ with the opinions that others express. Please see https://a16z.com/disclosures For additional important information.
