Maths Olympiad Prep

Library / /296 of 383

, 2021

Number theory Difficulty 8.9 Shortlist Prove it IMO

Let a1,a2,a3,a_{1}, a_{2}, a_{3}, \ldots be an infinite sequence of positive integers such that an+2ma_{n+2m} divides an+an+ma_{n} + a_{n+m} for all positive integers nn and mm. Prove that this sequence is eventually periodic, i.e. there exist positive integers NN and dd such that an=an+da_{n} = a_{n+d} for all n>Nn > N.

Solution

We will make repeated use of the following simple observation:

Lemma 1. If a positive integer dd divides ana_{n} and anma_{n-m} for some mm and n>2mn > 2m, it also divides an2ma_{n-2m}. If dd divides ana_{n} and an2ma_{n-2m}, it also divides anma_{n-m}.

Proof. Both parts are obvious since ana_{n} divides an2m+anma_{n-2m} + a_{n-m}.

Claim. The sequence (an)(a_{n}) is bounded.

Proof. Suppose the contrary. Then there exist infinitely many indices nn such that ana_{n} is greater than each of the previous terms a1,a2,,an1a_{1}, a_{2}, \ldots, a_{n-1}. Let an=ka_{n} = k be such a term, n>10n > 10. For each s<n2s < \frac{n}{2} the number an=ka_{n} = k divides ans+an2s<2ka_{n-s} + a_{n-2s} < 2k, therefore
ans+an2s=k a_{n-s} + a_{n-2s} = k
In particular,
an=an1+an2=an2+an4=an4+an8 a_{n} = a_{n-1} + a_{n-2} = a_{n-2} + a_{n-4} = a_{n-4} + a_{n-8}
that is, an1=an4a_{n-1} = a_{n-4} and an2=an8a_{n-2} = a_{n-8}. It follows from Lemma 1 that an1a_{n-1} divides an13sa_{n-1-3s} for 3s<n13s < n-1 and an2a_{n-2} divides an26sa_{n-2-6s} for 6s<n26s < n-2. Since at least one of the numbers an1a_{n-1} and an2a_{n-2} is at least an/2a_{n}/2, so is some aia_{i} with i6i \leq 6. However, ana_{n} can be arbitrarily large, a contradiction.

Since (an)(a_{n}) is bounded, there exist only finitely many ii for which aia_{i} appears in the sequence finitely many times. In other words, there exists NN such that if ai=ta_{i} = t and i>Ni > N, then aj=ta_{j} = t for infinitely many jj.

Clearly the sequence (an+N)n>0(a_{n+N})_{n > 0} satisfies the divisibility condition, and it is enough to prove that this sequence is eventually periodic. Thus truncating the sequence if necessary, we can assume that each number appears infinitely many times in the sequence. Let kk be the maximum number appearing in the sequence.

Lemma 2. If a positive integer dd divides ana_{n} for some nn, then the numbers ii such that dd divides aia_{i} form an arithmetical progression with an odd difference.

Proof. Let i1<i2<i3<i_{1} < i_{2} < i_{3} < \ldots be all the indices ii such that dd divides aia_{i}. If is+is+1i_{s} + i_{s+1} is even, it follows from Lemma 1 that dd also divides ais+is+12a_{\frac{i_{s} + i_{s+1}}{2}}, impossible since is<is+is+12<is+1i_{s} < \frac{i_{s} + i_{s+1}}{2} < i_{s+1}. Thus isi_{s} and is+1i_{s+1} are always of different parity, and therefore is+is+2i_{s} + i_{s+2} is even. Applying Lemma 1 again, we see that dd divides ais+is+22a_{\frac{i_{s} + i_{s+2}}{2}}, hence is+is+22=is+1\frac{i_{s} + i_{s+2}}{2} = i_{s+1}.

We are ready now to solve the problem.

The number of positive divisors of all terms of the progression is finite. Let dsd_{s} be the difference of the progression corresponding to ss, that is, ss divides ana_{n} if and only if it divides an+tdsa_{n + t d_{s}} for any positive integer tt. Let DD be the product of all dsd_{s}. Then each ss dividing a term of the progression divides ana_{n} if and only if it divides an+Da_{n+D}. This means that the sets of divisors of ana_{n} and an+Da_{n+D} coincide, and an+D=ana_{n+D} = a_{n}. Thus DD is a period of the sequence.

Want a route through all this instead of an archive? The track puts 2,000 problems in a working order, from AMC 10 level to the IMO shortlist.

Source: MathNet, licensed CC-BY-4.0. Statement reproduced verbatim; metadata (topic, difficulty) added by this project.