Reducing Redundancy and Enhancing Security in Blockchain through Adaptive Reward and Weight Mechanisms

This paper presents reforms to improve the Conflux blockchain protocol, including a continuous reward mechanism and redundancy reduction, enhancing performance while preserving security and decentralization. Our proposed scheme introduces an adaptive exponential weight system to address liveness and double-spending attacks more effectively. The protocol adapts to potential stagnation by exponentially scaling block weights, ensuring honest miners dominate the chain and achieve liveness. Additionally, the adaptive transaction broadcasting mechanism reduces network overhead while maintaining or improving security. Comparative analysis demonstrates that our approach not only achieves better resilience against liveness attacks and double-spending but also improves overall system security when contrasted with Conflux. The proposed modifications were implemented and tested on a light network deployed on two systems: one using an AMD Ryzen 5 processor with 16 GB of RAM and the other using an Intel Xeon processor with 6 GB of RAM. Our approach reduces duplication by a factor of around 50% while reducing memory consumption by a factor of 17% when compared with Conflux.