Preliminaries on the counting path. By clauses 1 and 2 of Counting Path and Its Jump Times, p is nondecreasing with p(0)=0 and takes values in {0}∪N, so 0≤p(x)≤p(R) for x∈[0,R] and every difference ∣p(u′)−p(u)−(u′−u)∣ with 0≤u≤u′≤R is at most p(R)+R; hence Discw(p) is a well-defined real number (the set is nonempty, containing the value for u=u′=0). For a natural number k≥1 let τk=τk(p)∈[0,∞] be the k-th jump time of that definition, the greatest lower bound of {t≥0:p(t)≥k}. We claim that for x≥0, p(x)≥k if and only if x≥τk. If p(x)≥k then x belongs to the set whose greatest lower bound is τk, so x≥τk. Conversely, let x≥τk (so τk<∞). For every t>τk there is, by the definition of the greatest lower bound, some t′∈[τk,t) with p(t′)≥k, hence p(t)≥p(t′)≥k by monotonicity; by right-continuity (clause 3) p(τk) is the greatest lower bound of {p(t):t>τk}, so p(τk)≥k, and p(x)≥p(τk)≥k.
Claim 1. Let C′ be measurable. For k≥1 the set {u∈[0,s]:p(Cu′)≥k} equals {u:Cu′≥τk} by the preliminary claim; if τk=+∞ this set is empty, and otherwise it is the preimage under the measurable map C′ of the closed set [τk,∞); in both cases it is measurable; so each indicator u↦1{p(Cu′)≥k} is measurable. Since p(Cu′)≤p(R), one has p(Cu′)=∑k=1p(R)1{p(Cu′)≥k} (an integer n with 0≤n≤p(R) equals the number of k∈{1,…,p(R)} with k≤n), a finite sum of bounded measurable indicator functions, hence bounded measurable by claims 1 and 2 of Arithmetic, Absolute Values, and Pointwise Limits of Measurable Real-Valued Functions. Then u↦Mˉ(Cu′)=p(Cu′)−Cu′ is bounded measurable (claim 2 of that lemma), and so is each component of u↦HuMˉ(Cu′), a product of bounded measurable functions (claim 3 of the same lemma). Likewise {u:Cˉu≥bj}=Cˉ−1([bj,∞)) is measurable, so u↦1{bj≤Cˉu}Hu is bounded measurable.
For the decomposition fix x∈[0,R] and let j− be the largest index j∈{0,1,…,J} with bj≤x (it exists since b0=0≤x). Then 1{bj≤x}=1 exactly for j≤j−, and, telescoping with p(b0)=p(0)=0 and b0=0,
j=1∑J(Kj−μj)1{bj≤x}=j=1∑j−(p(bj)−p(bj−1)−(bj−bj−1))=p(bj−)−bj−.
Hence Mˉ(x)−∑j(Kj−μj)1{bj≤x}=p(x)−p(bj−)−(x−bj−). If j−=J then x=bJ=R and this vanishes. Otherwise bj−≤x<bj−+1, so 0≤x−bj−<μj−+1≤μmax, and the window [bj−,x] has length at most μmax; by the definition of the window discrepancy the absolute value is at most Discμmax(p).
Claim 2. Fix u∈[0,s] and let x≤x′ be the two numbers Cu, Cˉu in increasing order; then x′−x≤w1 and Mˉ(x′)−Mˉ(x)=p(x′)−p(x)−(x′−x), whose absolute value is at most Discw1(p) by definition. Consequently
Λ(C)−Λ(Cˉ)=v(Mˉ(Cs)−Mˉ(Cˉs))+∫[0,s]Hu(Mˉ(Cu)−Mˉ(Cˉu))du
by linearity of the integral, and by the triangle inequality, Norm Bound for a Vector-Valued Lebesgue Integral over a Compact Interval and monotonicity, its norm is at most ∣v∣Discw1(p)+∫[0,s]∣Hu∣Discw1(p)du=(∣v∣+∥H∥1)Discw1(p).
Claim 3. For x∈[0,R] write r(x)=Mˉ(x)−∑j(Kj−μj)1{bj≤x}, so that ∣r(x)∣≤Discμmax(p) by claim 1. The map u↦r(Cˉu) is bounded measurable, being Mˉ(Cˉu) minus a finite linear combination of the measurable indicators of claim 1. Substituting Mˉ(Cˉu)=∑j(Kj−μj)1{bj≤Cˉu}+r(Cˉu) into Λ(Cˉ) and using linearity of the integral for the finite sum,
Λ(Cˉ)=j=1∑J(Kj−μj)(v1{bj≤Cˉs}+∫[0,s]1{bj≤Cˉu}Hudu)+vr(Cˉs)+∫[0,s]Hur(Cˉu)du,
and the first sum is ∑j(Kj−μj)αj. The remaining two terms have norm at most ∣v∣Discμmax(p)+∥H∥1Discμmax(p) by the same argument as in claim 2.
Claim 4. Combine claims 2 and 3 with the triangle inequality.