Author: Neiman 2018-01-29 13:34:20
Published on: 2018-01-29T13:34:20+00:00
The author is conducting a research project on blockchain timestamping and suggests that all such projects save data in a block and rely on the block timestamp as proof of the data's existence at a specific time. However, they question the accuracy of Bitcoin timestamps and found no discussion or research on this topic. They provide a simple analysis of timestamp accuracy and suggest improvements. The author observes that the block timestamp may not be the time that the block was created but rather when the miner started mining it. Timestamps are not necessary to avoid double-spending, but they are included in Bitcoin's protocol only to adjust the difficulty of PoW. The protocol controls timestamp accuracy based on the network time concept, which defines the median time of other clients, with new blocks being accepted by peers if their timestamps are less than the network time plus two hours and greater than the median timestamp of the previous 11 blocks. While the second rule doesn't give a tight lower bound, the author suggests that the PoW adjusting algorithm yields a second mechanism of accuracy control if miners have no motive to increase difficulty artificially. If we want to be pedantic, the best lower bound for a block timestamp is the timestamp of the block that closes the adjustment interval in which it resides. The author suggests exchanging the average with standard deviation in the difficulty adjustment formula to better mirror changes in hash power along the interval and disable the option to manipulate timestamps without affecting the difficulty, but acknowledges that this change requires a hardfork and won't happen anytime soon.
Updated on: 2023-06-13T00:21:09.402309+00:00