We currently aggregate the proposer signature with the rest of the proofs, which is unnecessary: the proposer signature is ~2KB and takes >1ms to verify, so having it travel alongside the block proof adds almost no overhead.
What's more, avoiding the aggregation of it reduces aggregation time by about 50% in the common case of 1 attestation blocks (since we aggregate 1 proof instead of 2, so half the work). It would also make empty blocks trivial to produce and propagate over the network, since they require no aggregation and weigh only as much as the signature plus the block (we assume the block proof is empty when there are no attestations).
We currently aggregate the proposer signature with the rest of the proofs, which is unnecessary: the proposer signature is ~2KB and takes >1ms to verify, so having it travel alongside the block proof adds almost no overhead.
What's more, avoiding the aggregation of it reduces aggregation time by about 50% in the common case of 1 attestation blocks (since we aggregate 1 proof instead of 2, so half the work). It would also make empty blocks trivial to produce and propagate over the network, since they require no aggregation and weigh only as much as the signature plus the block (we assume the block proof is empty when there are no attestations).