YES(?,O(n^2)) Problem: f(f(X)) -> f(a(b(f(X)))) f(a(g(X))) -> b(X) b(X) -> a(X) Proof: Complexity Transformation Processor: strict: f(f(X)) -> f(a(b(f(X)))) f(a(g(X))) -> b(X) b(X) -> a(X) weak: Matrix Interpretation Processor: dimension: 1 max_matrix: 1 interpretation: [g](x0) = x0, [a](x0) = x0, [b](x0) = x0 + 1, [f](x0) = x0 orientation: f(f(X)) = X >= X + 1 = f(a(b(f(X)))) f(a(g(X))) = X >= X + 1 = b(X) b(X) = X + 1 >= X = a(X) problem: strict: f(f(X)) -> f(a(b(f(X)))) f(a(g(X))) -> b(X) weak: b(X) -> a(X) Matrix Interpretation Processor: dimension: 1 max_matrix: 1 interpretation: [g](x0) = x0 + 1, [a](x0) = x0 + 1, [b](x0) = x0 + 1, [f](x0) = x0 orientation: f(f(X)) = X >= X + 2 = f(a(b(f(X)))) f(a(g(X))) = X + 2 >= X + 1 = b(X) b(X) = X + 1 >= X + 1 = a(X) problem: strict: f(f(X)) -> f(a(b(f(X)))) weak: f(a(g(X))) -> b(X) b(X) -> a(X) Matrix Interpretation Processor: dimension: 2 max_matrix: [1 1] [0 1] interpretation: [1 0] [g](x0) = [0 0]x0, [1 0] [a](x0) = [0 0]x0, [1 0] [b](x0) = [0 0]x0, [1 1] [0] [f](x0) = [0 1]x0 + [1] orientation: [1 2] [1] [1 1] [0] f(f(X)) = [0 1]X + [2] >= [0 0]X + [1] = f(a(b(f(X)))) [1 0] [0] [1 0] f(a(g(X))) = [0 0]X + [1] >= [0 0]X = b(X) [1 0] [1 0] b(X) = [0 0]X >= [0 0]X = a(X) problem: strict: weak: f(f(X)) -> f(a(b(f(X)))) f(a(g(X))) -> b(X) b(X) -> a(X) Qed