Problem 5.77

Problems and Theorems in Analysis · 6

Show
LaTeX
Problem 5.77

Let a1,a2,…,an,λ1,λ2,…,λna_1, a_2, \ldots , a_n, \lambda_1, \lambda_2, \ldots , \lambda_n be real constants, λ1<λ2<λ3<⋯<λn\lambda_1 < \lambda_2 < \lambda_3 < \cdots < \lambda_n. Denote by ZZ the number of real zeros of the entire function

F(x)=a1eλ1x+a2eλ2x+⋯+aneλnx F(x) = a_1 e^{\lambda _1 x} + a_2 e^{\lambda _2 x} + \cdots + a_n e^{\lambda _n x}

and by CC the number of changes of sign in the sequence of numbers a1,a2,…,ana_1, a_2, \ldots , a_n. Then C−ZC-Z is a non-negative even integer. (A generalization of 36, 37 different from 41: Replace exe^x by xx.)

Proof. We may assume without loss of generality that a1,a2,…ana_1, a_2, \ldots a_n are all different from 00. That C−ZC-Z is even is evident from the fact that as x→−∞x \to -\infty the term a1eλ1xa_1 e^{\lambda_1 x} is the dominant term, and as x→+∞x \to +\infty the term aneλnxa_n e^{\lambda_n x} is the dominant term. [8, 9, 37.] That C−Z≥0C-Z \geq 0 may be proved by mathematical induction; we deduce the relation for CC from that for C−1C-1, with the aid of Rolle's theorem [solution 75]. Indeed, if there are no changes of sign, evidently ZZ is also =0=0 and the statement holds. Assume that it holds in the case that there are C−1C-1 changes of sign. F(x)F(x) has CC changes of sign, C≥1C \geq 1; e.g. let α+1\alpha +1 be an index of change, 1≤α<n1 \leq \alpha < n, aαaα+1<0a_{\alpha }a_{\alpha +1} < 0. Choose a number λ\lambda in the interval λα<λ<λα+1\lambda_{\alpha } < \lambda < \lambda_{\alpha +1} and consider the function

F∗(x)=eλxd[e−λxF(x)]dx=a1(λ1−λ)eλ1x+⋯+aα(λα−λ)eλαx+aα+1(λα+1−λ)eλα+1x+⋯+an(λn−λ)eλnx. \begin{aligned} F^*(x) & = e^{\lambda x} \frac{d[e^{-\lambda x}F(x)]}{dx} \\ & = a_1(\lambda _1-\lambda )e^{\lambda _1 x}+\cdots +a_{\alpha }(\lambda _{\alpha }-\lambda )e^{\lambda _{\alpha }x} \\ & \quad + a_{\alpha +1}(\lambda _{\alpha +1}-\lambda )e^{\lambda _{\alpha +1}x}+\cdots +a_n(\lambda _n-\lambda )e^{\lambda _n x}. \end{aligned}

Let the number of zeros of F∗(x)F^*(x) be Z∗Z^*. We have [6, 12]

Z∗≥Z−1. Z^* \geq Z-1.

Let the number of changes of sign of the sequence of coefficients

−a1(λ−λ1),−a2(λ−λ2),…,−aα(λ−λα),aα+1(λα+1−λ),…,an(λn−λ) \begin{gathered} -a_1(\lambda -\lambda _1), \quad -a_2(\lambda -\lambda _2), \quad \ldots , \quad -a_{\alpha }(\lambda -\lambda _{\alpha }), \\ a_{\alpha +1}(\lambda _{\alpha +1}-\lambda ), \quad \ldots , \quad a_n(\lambda _n-\lambda ) \end{gathered}

be C∗C^*. Clearly

C∗=C−1. C^* = C-1.

By the assumption of the mathematical induction argument we have

C∗≥Z∗. C^* \geq Z^*.

From the three relations above it follows that

C≥Z, C \geq Z,

q.e.d.---Was it necessary to assume from the outset that a1,a2,…,ana_1, a_2, \ldots , a_n are all ≠0\neq 0? Could we not also have chosen λ=λα\lambda =\lambda_{\alpha } or λ=λα+1\lambda =\lambda_{\alpha +1}?

?
Problem 5.77.1

The entire function

F(x)=∑k=1n(−1)n−k(nk)kx F(x) = \sum _{k=1}^{n} (-1)^{n-k} \binom {n}{k} k^x

has simple zeros at the points

x=1,2,3,…,n−1 x = 1, 2, 3, \ldots , n-1

and no other real zeros. (Interpret kxk^x as exlog⁡ke^{x \log k}, with the principal value of log⁡k\log k.)

?