Without loss of generality, suppose s1≥r1+1. Let d=gcd(a,k), a=da1 and k=dk1. By using this equation and rewriting the problem's equality it is obtained that
i=1∏n(dria1ri+dk1)⟹i=1∏n(dri−1a1ri+k1)=i=1∏n(dsia1si+dk1)=i=1∏n(dsi−1a1si+k1)
Now consider the latest two equations modulo dr1a1r1+1. Since the sequences are strongly ascending, it is deduced that
⟹dr1a1r1+1(moddr1−1a1r1k1n−1)da1(modk1n−1)(dr1−1a1r1+k1)k1n−1≡k1n(moddr1a1r1+1)
So a1∣k1n−1, but gcd(a1,k1)=1. Therefore a1=1 that results in a∣k and it's in contradiction with a>k. Therefore s1=r1. After removing s1, r1 from the assumption and continuing the same thing for r2, s2 we take r2=s2. So for each i there is ri=si. ■