Eqs. (17)–(20), p. 11 — if then for every
ProvedLNChannels.RateBounds.eq_17_20lightning-networkp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1payment-channels
Let , , and
If
then for every real .
With , is the cost of a unidirectional channel of size minus the on-chain cost , so under this condition on-chain settlement is strictly cheaper than a unidirectional channel of any size. This is the sufficient condition (17)–(20) from which Theorem 2(i) follows.
Formalization Note The paper concludes " by (17)", but (17) is strict and the strict conclusion is what Theorem 2(i)'s strict inequality needs; the strict form is stated, for all rather than only at the minimizer . The paper's argument chooses , which needs ; no sign is imposed on here, since for every term of is nonnegative and .
Preamble
import Mathlib import Definitions.Def_LNChannels_RateBounds_ClosedForms
Formal statement
namespace LNChannels.RateBounds
/-- (17)–(20), p. 11: if `κ ≥ (e^C − 1)/B` then `F(m) = C/(1 − e^κ) + B/(e^{κm} − 1) + m` is
strictly positive for every `m > 0` (on-chain settlement is strictly cheaper than a unidirectional
channel of any size). -/
theorem eq_17_20 (B κ C : ℝ) (hB : 0 < B) (hκ : 0 < κ) (hκC : (Real.exp C - 1) / B ≤ κ) :
∀ m : ℝ, 0 < m → 0 < C / (1 - Real.exp κ) + B / (Real.exp (κ * m) - 1) + m := by sorry
end LNChannels.RateBounds
Source
Guasoni, Huberman & Shikhelman, Lightning Network Economics: Channels, Management Science (Articles in Advance, Nov. 6, 2023), DOI 10.1287/mnsc.2022.01664, p. 11 (PDF p. 12), proof of Theorem 2(i), eqs. (17)–(20)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.