Bitcoin requires decentralization of miners (or mining pools) and full nodes to achieve what some consider its core property: censorship resistance. As such, the block-size dispute represents a trade-off. Bigger blocks allow for more transactions on the Bitcoin network, but take more time to propagate, favoring larger miners and pools, while the increased data transmission disincentivizes users to run full nodes. Fortunately, there are proposals to increase Bitcoin’s efficiency that reduce the risk of bigger blocks. One of the most promising innovations in this regard, are Invertible Bloom Lookup Table , or IBLTs. First introduced by Bitcoin XT and Bitcoin Core developer Gavin Andresen, this idea was picked up and is currently worked on as a side-project by Linux veteran and Blockstream’s Lightning Network developer Paul “Rusty” Russell. “If we can make this work, it means less bandwidth requirements and less block data, which should be good for network health all over,” says Russell. Redundancy So what problem do IBLTs solve? Typically, all Bitcoin transactions are transmitted from node to node over the peer-to-peer network, to be stored by the mempools (the record of unconfirmed transactions) of individual nodes. When a miner finds a block, it includes (some of) these transactions in that block, and subsequently transmits this block over the same peer-to-peer network. Of course, this means that all transactions in the block are effectively sent over the network twice: once as a transaction, and once as part of a block. Speaking to Bitcoin Magazine , Russell explained: “We’ve got redundancy in blocks. Most nodes already know some of the content in that block; they have already seen it. If we can optimize that, we can speed up block transmission. That decreases centralization pressure because miners can get their blocks out faster, while the network works better … […]
