Saturday, 19 August 2017

Proving an inequality for a sequence by induction



I'm having some trouble with the following problem:



Let an be a sequence defined iteratively for n0 as follows:



an=am+1+2am+anm1+2 where m is defined as n21, where . denotes the standard floor function.




The inequality I'm trying to show is:



ann22+n



I've shown the result for the the cases n=0,1 (a0=0 and a1=1 by definition, by the way) so I just need to prove the inductive step for n2.



Edit - I should add that all the terms of the sequence are non-negative, so in particular if k1k2 then we have ak1ak2



Here's roughly what I've tried so far:




Unless I've made a mistake, I can assume that n is even (since if n is odd just replace n by n+1 which is n).



So n2=n2 and hence an=an2+2an21+ann2+2=2(an2+an21+1)



which is 2((n2)22+(n21)22+n2+n21+1)=2((n2)22+(n21)22+n) by the inductive hypothesis.



However, nothing I try from this point seems to make much progress.



I'm not exactly sure what the best way to manipulate the two floor functions above is, to get the required result.




Any help or hints would be greatly appreciated.



Thanks in advance.


Answer



an=am+1+2am+anm1+2=an21+1+2an21+ann2+11+2
=an2+2an21+ann2+2 Let's try by induction: suppose ann22+n for all kn (inductive hypothesis), let's prove it's valid also for n+1.




  • If n is even (n=2k):
    an=a2k=2(ak+ak1+1)

    and
    an+1=a2k+1=2(ak+ak1+1)=ann22+n(n+1)22+n+1

  • If n is odd (n=2k+1):
    an=a2k+1=2(ak+ak1+1)(2k+1)22+2k+1=4k2+4k+12+2k+1=2k2+2k+2k+1=2k2+4k+1
    and an+1=a2k+2=a2(k+1)=2(ak+1+ak+1)
    but for the inductive hypothesis, ak+1(k+1)22+k+1 and ak(k)22+k which means that an+1=2(ak+1+ak+1)2((k+1)22+k+1(k)22+k+1)
    Now,

  • if k=2x an+1=a2k+2=a4x+22((k+1)22+k+1(k)22+k+1)=8x2+12x+4(4x+2)22+4x+2=8x2+8x+2+4x+2=8x2+12x+4

  • instead if k=2x+1 an+1=a2k+2=a4x+42((2x+2)22+2x+2+(2x+1)22+2x+2)=8x2+20x+12(4x+4)22+4x+4 so the proof is finally concluded, because it's been proved that if ann22+n, so does an+1, and since a00 it follows that ann22+n for every nN




Q.E.D


No comments:

Post a Comment

real analysis - How to find limhrightarrow0fracsin(ha)h

How to find lim without lhopital rule? I know when I use lhopital I easy get $$ \lim_{h\rightarrow 0}...