Search

The existing blockchain architecture have structural redundancy and performance weakness when facing vertical structural lightweight applications. Therefore, this paper proposed a lightweight blockchain architecture. In the data layer, it designed a data storage mechanism by row to improve data read and write efficiency, and presented a storage methods based on MongoDB to support offline isolated storage of massive data. In the consensus layer, a hierarchical consensus mechanism divided nodes into decision-making and execution layers according to transaction granularity, reducing the number of nodes participating in the consensus and improving the efficiency of consensus. In the network layer, it introduced a distributed structured network model to improve the efficiency of data broadcasting. In the application layer, it proposed the user-defined transaction interface to support the function customization of blockchain system in different scenarios, and used the user-defined finite state machine replace the Turing machine to improve the controllability and complexity of smart contracts. The simulation experiments show that the LBA architecture can effectively improve the data storage efficiency, network communication and consensus efficiency in the blockchain system. LBA can be applied to the lightweight small and medium-sized application scenarios of the blockchain. [ABSTRACT FROM AUTHOR]

DNA computing is a novel parallel computation paradigm. Existing models were mostly based on molecular biological technology and different from the traditional computers, they were less universal than traditional ones, so the algorithm or a DNA computer just for one kind of problem could not be transplanted to solve other problems without any modification. This paper put forward a new generalized molecular computational model (designated by GMCM) based on Turing machine, which did not rely on biotechnology and included an ordinary single tape Turing machine, a finite write-only tape, and a special working tape. Through the topological mapping between the write-only tape and the working tape, it could realize read or write in parallel. Validation test proves that GMCM can solve the satisfiability problem in a polynomial time and has obvious advantages compared with existing DNA computer on computation accuracy and universal property. [ABSTRACT FROM AUTHOR]