Let n be a natural number. We define sequences ai and bi as follows. We let a0=1 and b0=n. For i>0, we let (ai,bi)=⎩⎨⎧(2ai−1+1,bi−1−ai−1−1)(ai−1−bi−1−1,2bi−1+1)(ai−1,bi−1)if ai−1<bi−1,if ai−1>bi−1,if ai−1=bi−1. Given that ak=bk for some natural number k, prove that n+3 is a power of 2.
Solution
We first note that ai+bi=n+1 for all natural numbers i. We define three sequences {ri}, {si}, {ti} as follows. Let r0=1, s0=t0=0. For i>0, we let (ri,si,ti)=⎩⎨⎧(2ri−1,2si−1,2ti−1+1)(2ri−1−1,2si−1−1,2ti−1−1)(ri−1,si−1,ti−1)if ai−1<bi−1,if ai−1>bi−1,if ai−1=bi−1. It is then easy to verify that ai=ria0+sib0+ti and bi=(1−ri)a0+(1−si)b0−ti for all i≥0.
Now let k be the smallest non-negative integer such that ak=bk. For any integer l with 0≤l≤k we claim that rl−sl=tl+1−2sl=2l. We prove this claim by induction on l. The claim is straightforward to check for l=0. Now, suppose that the claim is true for some non-negative integer l<k. Since l+1≤k it follows that (rl+1,sl+1,tl+1) equals either (2rl,2sl,2tl+1) or (2rl−1,2sl−1,2tl−1). In either case, rl+1−sl+1=2(rl−sl) and tl+1+1−2sl+1=2(tl+1−2sl). Hence the claim.
Now if ak=bk then rka0+skb0+tk=(1−rk)a0+(1−sk)b0−tk. This implies (2rk−1)+2tk=(1−2sk)n. Therefore (n+3)(1−2sk)=2(rk−sk)+2tk−4sk+2=2k+2. Hence n+3 is a power of 2. In fact, since 1−2sk is also a power of 2, it follows that sk=0 and hence n+3=2k+2.
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.