On November 21st, at the Baidu Smart Cloud Techday Blockchain Product Salon, Baidu Smart Cloud team joined hands with industry players such as Shanghai Pudong Development Bank, Intel, Yuxin Technology, and Firecoin. They shared their development trend on blockchain and industry. Thinking about products and technologies.
The blockchain layout of Dachang has always been one of the focuses of Babbitt. We have had detailed reports on Baidu Super Chain , and this is the first collective debut of Baidu Smart Cloud Blockchain team. According to the reporter of Babbitt, Baidu Super Chain and Baidu Intelligent Cloud Team are all committed to the research and development of blockchain technology, but the focus is different. The former focuses on the underlying public chain, namely “XuperChain”, while the latter focuses on BaaS. The service, the "sky chain", includes not only the self-developed "XuperChain", but also the community's most powerful public chain "Ethereum", the alliance chain "super account book", the enterprise Ethereum "Quorum", and even the first A Chinese enterprise BaaS platform that supports the cross-chain project Cosmos.
The following content comes from the salon site, Baidu intelligent cloud blockchain product leader Liu Wei and Baidu intelligent cloud blockchain architect Chen Haodong share the scene, they introduced the overall layout of Baidu blockchain, "sky chain" BaaS platform Technical architecture and trusted computing platform business logic. The text was compiled by Babbitt reporters based on live recordings and PPT.
Baidu blockchain overall layout
From the application practice, Baidu began to try in 2016. At that time, for many enterprises, it was very difficult to accept the blockchain technology. Our overall idea at the time was to try first in our own business. For example, Baidu Totem, Baijiahao, Baidu Encyclopedia, we do copyright protection on it, when you edit the entry, you can see the hash value; such as the Leitz dog, the universe and the oasis, we do through the blockchain Game points, as well as game motivation systems.
In terms of technical layout, we have also done a lot of work, including the open source underlying framework "superchain" of the benchmark shared chain, which has relatively strong performance when managing multiple nodes and thousands of participants; Including our Baidu Smart Cloud has done a complete set of "Tian Chain" BaaS system, it will be compatible with many technologies, such as Ethereum, Super Book, Cosmos, Quorum, XuperChain, can be used on our platform, we should also be China A BaaS platform that supports Cosmos.
In terms of ecological development, we feel that we can't go it alone, so we joined the LINUX Foundation's privacy computing project as a founding member. We are also a member of the Superbook Foundation, or a member of the Enterprise Ethereum (EEA). Also joined the ICT Trusted Blockchain program and launched its own Super Partner Program. We hope to explore with you the possibility of discovering the blockchain technology in the enterprise.
From the perspective of venue landing, we made a chain of supply chain financial asset securitization in 2017, and also released the "Baidu Blockchain White Paper", which has a lot of scenes to introduce, you can go and see.
Sky Chain BaaS Platform
Next, we will give you a simple disassembly. What kind of platform is “Tian Chain”? How should everyone use it?
Our bottom layer is a BaaS platform, which contains a lot of functional components, including the management of the alliance chain, the module of privacy calculation, the comprehensive monitoring of contracts and certificates, and the integration of the underlying cloud technology, including container orchestration and resource monitoring. Turn our chain into a deliverable, engineered software that can be deployed to a cloud data center or directly through a public cloud platform.
The upper layer is a business trusted platform. The blockchain is a distributed ledger. It is actually a machine for creating trust. We believe that its solution should have trusted attributes, so our trusted computing platform can The digital identity platform and the trusted digital rights platform have made a vertical solution with three dimensions.
The Trusted Computing Platform is a multi-party computing, secure and fair data computing platform; the trusted digital identity platform is a multi-party DID system that uses the mainstream standard of W3C. Distributed identity, different from centralized identity, such as Facebook. We believe that in the Web3.0 era, distributed identity systems will be very important; trusted digital rights platform, based on your identity, the data generated on the future value of the Internet is your copyright, we will help protect you rights and interests. This part will be described in detail below.
The next level is the scenario solution, including deposit certificate, data, and points. This is where we think there is a certain possibility in the scene and the industry attributes are not so strong.
Credible deposits, such as bill tracing, copyright protection, product traceability and IOT, the core is to put the data flowing in real life into the chain, through distributed management to ensure that it can not be falsified and transparent and credible; Data, such as joint credit, joint marketing, and genetic data analysis, build a multi-party, trustworthy data federated computing platform while protecting the power of data owners.
The two scenarios we are currently focusing on, one is smart government and the other is credible finance. These two are the most likely to land. We have done a lot of scenario-based “blockchain+industry” practices in Trusted Finance, such as supply chain finance, asset securitization, financial information sharing, and intelligent collection. Later, the government industry scene will be introduced, such as one-stop government affairs, organization catalog, digital identity, public interest traceability and so on.
Background and selection of trusted computing
Trusted computing is a type of technology derived from the related requirements of data collaboration in the context of the big data era. The development of information technology is mainly reflected in our ability to control data. If you look at the time dimension, you have experienced 3 to 4 periods.
The first period was digital. In the 1970s, most of the world's data began to be digitally converted by trading engines. The second period was the beginning of the big data era in 2005, mainly the Hadoop ecosystem and HDFS. Data storage services, MapReduce parallel data processing services, these industry standards began to build; the third period is 2014, we entered the big data sharing, the background is that the mobile ecology caused a large-scale outbreak of data, and giant companies have formed a monopoly on data. At this stage, the data business value is highlighted, the data circulation demand is prominent, the big data exchange is established; the fourth period is from 2018, and we have already transitioned to the stage of data collaboration. Data islanding problem, user data privacy problem, data ontology rights problem, convergence analysis under the condition that data ontology rights are unchanged, these issues are getting more and more attention.
From the perspective of data collaboration, the demand side and the data side are calculated. Their relationship is not a simple sale but a partnership. In the process of cooperation, we concluded that there are five characteristics that need to be focused on: First, data privacy protection, which means that only the owner of the data in the calculation process can see the original data, and the other parties cannot know the data itself. The second is the calculation verifiable, which is for the calculation of the demand side, it wants to achieve its purpose in the calculation process, is it really calculated its expected calculation function, it needs to be verifiable; The third is the transparency of computing requirements. This is for the data side. That is to say, before the data party provides the data, it can view what the demand side has done with my data, and view the whole process in plain text. Fourth, the calculation result is confidential. The calculation results are only open to the calculation of the demand side, and the other parties are not visible; the fifth is that the process can be traced, which is set for the needs of the regulatory body.
With the requirements, what technologies do we have now that can meet such standards? Below we list and compare some of the safety data collaboration technology solutions that can be researched by academics and engineering circles.
Directly speaking, for the technical solutions of universal data collaboration, the TEE of the trusted computing environment has a comparative advantage. The other three types of technology Baidu also have a dedicated team to track and research. At present, there is still a lot of work from the commercial landing. To do it, including how to improve its security, reduce its resource consumption, and enrich its ecology.
In a brief introduction, a secure area on the main processor of the TEE computing device is itself a black box, which is inherently capable of providing isolated execution and secure storage. At the same time, TEE also agreed on a set of interaction rules, that is, I have a way to dynamically verify that TEE is legal outside, including 2 elements, one is TEE hardware is not genuine, the other is whether the running program and the agreement are consistent We call this process remote authentication. Combined with the three functions of isolation execution, secure storage and remote authentication, it is possible to do black box operations when the process stack and data are completely confidential.
This is the design architecture of our product. It is divided into three layers. The bottom layer is based on TEE technology. Combined with SGX security hardware and MesaTEE security software, it realizes universal security computing capabilities. The middle layer is a general-purpose collaboration system, which is based on the blockchain. Various collaborative relationships are defined on the blockchain, including trusted environment registration authentication, algorithm chain consensus, calculation process monitoring, etc., which can be done in weak trust or no. A multi-party collaboration model is established in the trust environment; the top layer is an open platform for algorithmic capabilities, and the Baidu intelligent cloud big data platform and the financial joint modeling platform are docked to complete the application of specific data analysis models and data negative models.