Lambda is a Disruptor in Blockchain-Based Storage Solution, bringing to the market, a fast, safe, and scalable blockchain infrastructure, which provides decentralized applications (DAPPs) data storage capabilities with unlimited scalability. The CEO HE Xiao Yang will be discussing the Lambda project with us in this interview.
Please tell us your name and about your background?
My name is HE Xiao Yang and I am the founder of Lambda. I have a deep passion for technology innovation and software development. I am well known in China for my extensive experience in infrastructure software and active involvement in the open source community.
Previously, I was the co-founder of OneAPM which is a fast growing infrastructure software focusing on ITOM (IT operation management) in China. Some refer to OneAPM as the “New Relic or AppDynamics of China” and the company has received strong VC backing from Matrix Partners, Chengwei Capital, and Qiming Venture. Prior to my entrepreneur experiences, I worked at BEA as a R&D software engineer. I am also a blogger and some of my articles have been published by major media such as Forbes China, 36Kr, Sina, etc.
Lambda is a fast, safe, and scalable blockchain infrastructure project, which provides decentralized applications (DAPPs) data storage capabilities with unlimited scalability.
What are the major services and features provided by Lambda?
Lambda provides a new solution to separately save different types of data in distinct chains and blocks. It is devoted to providing an infrastructure for DAPPs, on which it intends to offer a series of infrastructure capability services including storage with unlimited scalability, high throughput computing, and fast network transmission, making it easy for DAPPs to fulfill data generation, transmission, storage, retrieving, and computing.
Lambda services include: multi-chain data co-storage, cross-chain data management, data privacy protection, Proof of Recoverability of data (POR), Provable Data Possession (PDP), and distributed intelligent computing through logic decoupling and independent implementation of Lambda Chain and Lambda DB.
Additionally, developers are allowed to freely use Lambda project codes and modify them under Apache License while users are permitted to download binary files of Lambda and deploy them in the private environment without connecting to the extranet.
What is the technology behind Lambda? Can you give us a walk-through of the architecture design?
We use Lambda Chain’s homogeneous multi-chain structure design to randomly select a set of verifier nodes to check the integrity of the data when we read it. Through separate validation of multiple verification nodes and Byzantine Fault Intolerance (BFT) consensus, we can complete Proof of Recoverability of data (POR) verification and Provable of Data Possession (PDP), which is our advantage over IPFS. IPFS currently has no mature chain design and complete data storage proof mechanism.
In addition, Lambda ensures the following:
- Our innovative design includes the separation of blockchain and storage chain
- We have addressed the challenge of Provable Data Possession (PDP), with a perfect integration with blockchain.
- Our solution can support high throughput data storage and retrieval by leveraging Sharding technology, which provides possibilities for future DAPP application scenarios.
- By leveraging the subchain technology and cooperating with open source community, we will be able to offer an ecosystem platform with data collection, storage, computing, exchange capabilities in addition to the core block file storage capability.
2017/2018 introduced many blockchain-based data projects. What makes the Lambda trusted and secure data storage unique?
We believe that all storage projects follow the core requirement of Provable Data Integrity. Also, it is crucial for data projects to store data in an infeasible storage node and make them trusted and secure, in the academic world, there are mainly two approaches which are Provable Data Possession (PDP) and Proofs of Retrievability (POR) depending on the usage scenarios. This technology is quite mature in cloud storage and Lambda has applied these two approaches in blockchain through academic research and engineering verification, which has eliminated the core obstacles of decentralized storage.
How does the Lambda public data access and transaction work?
Lambda’s public data is primarily the main chain and sub-chain ledgers, as well as the TAG information that must be stored for data verification. We provide a special data structure to store data verification TAG.
For transactions, Lambda is divided into two parts: on chain trading and off chain trading. Due to current technical constraints, we support the requirements of trading between demand and supply sides of storage service as well as the subsequent data storage, by leveraging the similar approach utilized in the decentralized exchanges.
What specific types of data protection and personal privacy will Lambda focus on?
Lambda believes that the stored data is owned by the users themselves, and no third party has the right to access the data unless authorized by the user.
Currently, data on the blockchains are public and accessible, which limits the application scenarios to a large extent. Lambda has adopted an Online/Offline Multi-Authority Attribute Based Encryption (OO-MA-ABE) plan, which aims to transfer the online computing cost of the client to the off-line stage or cloud servers.
The primary advantages are listed as follows:
- With the use of online-offline and outsourced deciphering technology, Lambda formulates an efficient mobile cloud storage data access plan. In the encryption stage, it pre-processes a large number of matching operations in advance; in the deciphering stage, it outsources the matching operations to the cloud storage server. Consequently, the client can fulfill encryption and deciphering with simple calculation operations only, significantly decreasing the online computing expenses of the client.
- Lambda promotes a method to verify the correctness of outsourced deciphering. In the encryption stage, the user generates Hash values by encrypting private keys and clear texts as the verification token of data; in the deciphering stage, the user adopts the token to verify the correctness of the deciphering result, and therefore check whether the deciphering of cloud storage server is correct. Meanwhile, the plan specified in the paper can resist from the identity information acquisition by single institutions, protecting the privacy of users.
- The Lambda team performs the security analysis and simulation experiments for the plan, the results of which indicate its security and efficiency.
How do the Lambda token incentives work with the IoT data storage services?
As a decentralized storage system, Lambda is ideal for storing time-based valuable data. Users can spend tokens to store such data. They can also authorize certain organizations or other users to use their own data to earn tokens. The demand side and the supplier of data can reach an agreement through the data trading platform.
Please tell us more about the Lambda data trading platform and how you utilize artificial intelligence data.
The first layer of the Lambda trading platform / exchange is for storage service transaction, which covers the order matching and payment between the storage service demand side and the provider side.
On the second level, we will encourage developers in the ecosystem to implement computational sub-chains in the data computing domain and access them to the Lambda system, as well as order matching, payment, and result delivery for data calculations in data exchanges.
Within the Lambda system, trading is a very important part of us, because the Lambda team constructs only code and rules that are publicly placed on Github and other open source platforms. However the actual operation, the runtime environment on each host, the corresponding blockchain ledger, the actual storage, etc. are all provided by contributors within the ecosystem, or miners.
In the Lambda system, there are several different exchanges built in. Currently, we are still adjusting the rules of these exchanges, but the functions are certain.
- The Mining Exchange – The storage miner and the validator pledge the Lambda tokens, gaining the rights to become a miner. Unlike previous systems, the pledge price of this exchange follows a certain curve. The result of this curve can be guaranteed in such way, a) so that the number of storage nodes and verification nodes can always follow the optimal ratio; b) in a specific time period, the storage space provided will be close to a certain threshold, neither too much nor too little.
- The Storage Space Exchange – In this exchange, tenants can initiate bid-type transactions requesting storage space. Storage miners can rent out storage space here. Because storage space is a must for data transactions, for our system, the storage space exchange is similar to the EOS RAM exchange, we will also adopt a trading system similar to Bancor, but we will not set CW parameters that may cause the price too steep.
- The Data Exchange – For tenants, after obtaining storage space, users can buy and sell data on the data exchange. Our data exchange adopts a mechanism similar to secure multi-party computing. After the transaction is over, the owner of the data does not know what data the buyer specifically purchased, and the seller of the data does not know what data the buyer has.
What is next on the Lambda roadmap? Do you have any current opportunities for investors?
Q3 2018 will be a very important milestone for Lambda, in which we need to implement Lambda Chain and Lambda FS and Lambda FS-based object storage. We will also conduct public offerings, so stay tuned for our official news.
Could you tell us about your team and advisors?
Our core team is a group of developers who have extensive experience in infrastructure software R&D and open source community. We contribute codes to various open source projects under the Apahce Foundation, such as Camel, Akka, Drill, etc. Our team members are deeply involved in all of the projects. We are also actively promoting the development of open source and communities in China and have established the China Users Groups of Druid and Clickhouse.
In addition to myself, our core team includes: Lucy Wang CMO at Lambda; Alex Lototskiy, senior principal security engineer; He Bingqing, OneAPM’s co-founder & CTO; Jiang Ning, Apache Camel Mentor; Oleg Lyamtsev, engineer; Li Monan, China’s top programmer and founder of Coreseek and Log Insight; Gao Haiqiang, computer science expert; Zhang Xiaoguang, Apache Akka Commiter; Guo Hongqiang, data scientist.
We also have an experienced team of advisors: Tian Jia, Wang Tao, Chen Yuan, Liu Chunhua.
How secure is Lambda? Would you like to talk about your legal and security measures?
Lambda uses an online-offline and outsourced decryption technology to construct an efficient mobile cloud storage data access control scheme. In the encryption phase, a large number of pairing operations are pre-processed in advance; in the decryption phase, the pairing operation is outsourced to the cloud storage server; the client only needs to run a simple operation to complete the encryption and decryption, thereby greatly reducing the online computing overhead of the client.
Lambda proposes a method to verify the correctness of outsourced decryption. In the encryption phase, the user uses the encryption key and the plaintext to generate a hash value as the data verification token; in the decryption phase, the user verifies the decryption result with the verification token and then verify whether the cloud storage server decryption is correct.
Additionally, Lambda utilizes PDP and POR technology to detect whether the data is stored. At the same time, the data adopts multiple backups. According to different storage levels, for example, 1/3 of the copy loss can be detected in time and the data is restored and transferred by POR technology and Erasure Coding technology.
Moreover, the decentralized cloud storage and cloud database, compared to the centralized cloud, is the biggest difference is the Permissionless storage node. Also,the cloud user is not a verification initiator, nor a fully trusted TPA node. In Lambda’s design, we are using two upgraded versions of the PDP method that support public validation, BLS-PDP and MF-PDP, and we will work together on a trusted TPA through the consensus of multiple verifier nodes to complete the data holding and integrity verification, and write the verification results to the chain for FisherMan inspection.
Where should we direct our readers for more information?
Official website: http://www.lambda.im
Telegram Community: https://t.me/HelloLambda
White Paper: http://www.lambda.im/doc/Lambda-WhitePaper-en.pdf