• Non ci sono risultati.

Let {an}n≥1 be a sequence of real numbers

N/A
N/A
Protected

Academic year: 2021

Condividi "Let {an}n≥1 be a sequence of real numbers"

Copied!
2
0
0

Testo completo

(1)

Problem 11809

(American Mathematical Monthly, Vol.121, December 2014) Proposed by O. Kouba (Syria).

Let {an}n≥1 be a sequence of real numbers.

(a) Suppose {an}n≥1 consists of nonnegative numbers and is nonincreasing, and P

n=1an/√

n converges. Prove thatP

n=1(−1)nan converges.

(b) Find a nonincreasing sequence {an}n≥1 of positive numbers such that limn→∞√nan= 0 andP

n=1(−1)nan diverges.

Solution proposed by Moubinool Omarjee, Lyc´ee Henri IV, Paris, France, and Roberto Tauraso, Dipartimento di Matematica, Universit`a di Roma “Tor Vergata”, Italy.

(a) We first note that limn→∞

nan= 0. Otherwise there exist ε > 0 and n0≥ 1, such that for all n ≥ n0,√nan> ε and

n

X

k=n0

ak

√k > ε

n

X

k=n0

1 k, which contradicts the fact thatP

n=1an/√

n converges.

Moreover, we have that

Sn:=

n

X

k=1

(−1)k=

a−1

X

k=1

(−1)k((k + 1)2− k2) + (−1)a(n + 1 − a2) = −(−1)aa − 1 + (−1)a(n + 1 − a2)

where a = ⌊√

n⌋. Therefore |Sn| ≤ |n − a2− a| + 2 ≤ a + 2 ≤√ n + 2.

Now, for 1 ≤ n ≤ m,

m

X

k=n

(−1)kak =

m

X

k=n

(Sk− Sk−1)ak=

m−1

X

k=n

Sk(ak− ak+1) − Sn−1an+ Smam.

Since {an}n≥1 is a nonincreasing sequence, we have

m−1

X

k=n

|Sk(ak− ak+1)| ≤

m−1

X

k=n

(√

k + 2)(ak− ak+1)

=

m

X

k=n

(√ k −√

k − 1)ak+ (√

n − 1 + 2)an− (√

m + 2)am

m

X

k=n

ak

√k+ (√

n − 1 + 2)an− (√

m + 2)am.

Hence, it follows

m

X

k=n

(−1)kak

m−1

X

k=n

|Sk(ak− ak+1)| + |Sn−1|an+ |Sm|am

m

X

k=n

ak

√k + 2(√

n − 1 + 2)an

which implies that {Pn

k=1(−1)kak}n≥1 is a Cauchy sequence because each term on the right hand side goes to zero as n and m go to infinity.

(2)

(b) For n ≥ 1, let

an = 1

xnln(1 + xn) with xn= ⌊(√

n + 1)/2⌋.

Then {an}n≥1 is a nonincreasing sequence of positive numbers such that for n > 1,

0 <√

nan< 2√ n (√

n − 1) ln(√

n + 1)/2) because x − 1 < ⌊x⌋. Hence limn→∞

nan= 0. Moreover

(−1)n= −1 and xn= m for (2m − 1)2≤ n < (2m)2, (−1)n= +1 and xn= m for (2m)2≤ n < (2m + 1)2, which implies that for N < (2m + 1)2,

N

X

n=1

(−1)nan

m−1

X

n=1

2

n ln(1 + n)− (4m − 1) m ln(1 + m). Therefore the series diverges

X

n=1

(−1)nan

X

n=1

2

n ln(1 + n) − limm

→∞

(4m − 1)

m ln(1 + m) = +∞.



Riferimenti

Documenti correlati

[r]

3) Algebra: polynomials and their roots, quadratic polynomials, polynomial division, finding roots, binomial coefficients.. 4) Equations: types of

Then the following assertions are equivalent... the hermitian part of A

Solution proposed by Roberto Tauraso, Dipartimento di Matematica, Universit`a di Roma “Tor Vergata”, via della Ricerca Scientifica, 00133 Roma,

[r]

[r]

Fon-Der-Flaass (Russia)

in such a way that U 1 receives at least one ball, and while any balls remain, each successive urn receives at least as many balls as in all the previous urns combined.. Let b(n)