Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 3.20 — existence and uniqueness of the Erlang fixed point

Proved
KellyStochasticNetworks.erlang_fixed_point_unique

by naimengye · Sep 18, 2026 · Mathlib 0df444a (Lean v4.33.1)

convex-optimizationloss-networksstochastic-networksteletraffic

Consider a loss network with JJJ links and RRR routes: link jjj carries Cj≥1C_j \ge 1Cj​≥1 circuits, a call on route rrr requires AjrA_{jr}Ajr​ circuits from link jjj, and calls on route rrr arrive at rate νr>0\nu_r > 0νr​>0. The Erlang fixed point equations are

Ej=E ⁣((1−Ej)−1∑rAjr νr∏i(1−Ei)Air,  Cj),j=1,…,J,(3.7)E_j = E\!\left((1 - E_j)^{-1}\sum_r A_{jr}\,\nu_r\prod_i (1 - E_i)^{A_{ir}},\; C_j\right), \qquad j = 1,\dots,J, \tag{3.7}Ej​=E((1−Ej​)−1r∑​Ajr​νr​i∏​(1−Ei​)Air​,Cj​),j=1,…,J,(3.7)

where E(ν,C)E(\nu,C)E(ν,C) is Erlang's formula. Then there is exactly one vector (E1,…,EJ)∈[0,1]J(E_1,\dots,E_J) \in [0,1]^J(E1​,…,EJ​)∈[0,1]J satisfying them.

The equations are the reduced-load approximation: pretend the links block independently, so that the traffic offered to link jjj is the traffic on the routes through jjj, thinned by the blocking probability of every other link each route uses, and then apply Erlang's formula link by link. The factor (1−Ej)−1(1 - E_j)^{-1}(1−Ej​)−1 removes link jjj's own thinning from the product. This approximation is what practitioners compute, and the theorem is what makes "the Erlang fixed point" a well-formed phrase: repeated substitution, or a damped iteration, converges to a vector that does not depend on where it started.

Existence is a short argument: the right-hand side defines a continuous map of the compact convex cube [0,1]J[0,1]^J[0,1]J into itself, so Brouwer's theorem applies. Uniqueness is the substance, and the book obtains it by rewriting (3.7) as the stationary conditions of a strictly convex minimization, the revised dual problem (3.8).

Formalization Note The incidence matrix has non-negative integer entries, which is the general case of section 3.3 rather than only the 000–111 case. The hypothesis Cj≥1C_j \ge 1Cj​≥1 is needed for the statement to mean anything: a link with no circuits has E(ν,0)=1E(\nu,0) = 1E(ν,0)=1 identically, and the factor (1−Ej)−1(1 - E_j)^{-1}(1−Ej​)−1 would then be a division by zero. Under Cj≥1C_j \ge 1Cj​≥1 no solution has Ej=1E_j = 1Ej​=1, so the quotient is taken at a non-zero denominator throughout.

Preamble
import Mathlib
import Definitions.Def_KellyStochasticNetworks_Balance
import Definitions.Def_KellyStochasticNetworks_Erlang
import Definitions.Def_KellyStochasticNetworks_Migration
import Definitions.Def_KellyStochasticNetworks_LossNetwork
Formal statement
namespace KellyStochasticNetworks

theorem erlang_fixed_point_unique {J R : ℕ} (A : Fin J → Fin R → ℕ) (ν : Fin R → ℝ)
    (C : Fin J → ℕ) (hν : ∀ r, 0 < ν r) (hC : ∀ j, 1 ≤ C j) :
    ∃! E : Fin J → ℝ, (∀ j, E j ∈ Set.Icc (0 : ℝ) 1) ∧ ErlangFixedPoint A ν C E := by sorry

end KellyStochasticNetworks
Source
Kelly & Yudovina, Stochastic Networks, CUP 2014, p. 67 (PDF p. 75), equation (3.7) and Theorem 3.20: 'Consider the equations E_j = E((1 - E_j)^{-1} sum_r A_{jr} nu_r prod_i (1 - E_i)^{A_{ir}}, C_j), j = 1, ..., J, (3.7) the generalization of the Erlang fixed point equations to matrices A that may not be 0-1. ... Theorem 3.20 There exists a unique solution (E_1, ..., E_J) in [0,1]^J satisfying the Erlang fixed point equations (3.7).' sha256 ec271d555059aee58613e5e9a98b8346214b16185c527314f5d94c1ac8b17b6a
Human review
  • Endorsed by Shuze Chen · Sep 25, 2026

    Confirmed by the moderator at approval.

  • Endorsed by naimengye · Sep 25, 2026

    Confirmed by the mission captain (proposal self-audit).

  • Endorsed by Community (Bot) · Oct 22, 2026

    Confirmed by the moderator at approval.

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me