# Prove that there are infinite integer solutions to an equation (Use Pell?)

I neeed to prove that this equation: $$k(2a+(1+k))^2-(k+1)(2b+k)^2=k(k+1)$$ has infinite solution for $a,b \in \mathbb{N}^+$ and $k$ constant (positive integer too).

but I have a sort of hint (doesn't really help me):

If $k$ is odd I can substitute $kc$ to $b$, obtaining $$(2a+k+1)^2-(k(k+1))(2c+1)^2=(k+1)$$ And apparently I can do some trick like this if $k+1$ is odd, subtituting $a$ with something. But I can't reach anything like Pell's.

Then I can prove with Pell's that $$m^2-(k(k+1))n^2=k+1$$ has infinite positive integer solutions for $m$ and $n$, and the hint now says that it is done. But I wouldn't say so, because if in all the solutions $n$ turns out to be even $2c+1 \neq n$ and therefore there are no solution fo the original equation?

Thank you!

First, note that the original equation has the solution $a=0,b=0$.
Now multiply through by $k$, and let $u=k(2a+k+1),v=2b+k$ to get $$u^2-k(k+1)v^2=k^2(k+1)\tag1$$ Again, this has the solution $u=k^2+k,v=k$. Now consider the Pellian $$u^2-k(k+1)v^2=1\tag2$$ Since $k(k+1)$ is not a square, this Pellian has infinitely many solutions. Do you know how to combine a solution of (2) with the solution we have for (1) to get another solution for (1)? If so, you now have infinitely many solutions for (1). Then see if you can work your way back from $u,v$ to $a,b$.
• Ok, but how can I show that all the solutions will have: $$u \equiv k+1 \pmod 2$$ and $$v \equiv k \pmod 2$$? – Problemsolved Jun 22 '12 at 7:26
• Why would you want to do that? From $u=k(2a+k+1)$ it's clear that $u$ is even, and from $v=2b+k$ it's clear that $v\equiv k\pmod2$. – Gerry Myerson Jun 22 '12 at 7:37
• Sorry, I wasn't clear enough. How can I work my way back from $u, v$ to $a, b$? – Problemsolved Jun 22 '12 at 7:58
• Can you not solve $u=k(2a+k+1)$ for $a$, and $v=2b+k$ for $b$? – Gerry Myerson Jun 22 '12 at 9:12