For $j = 1,\ldots,n$, let $X_j = 0$ if the $j$-th coin flip is tails and $X_j = 1$ if the $j$-th coin flip is heads. The number of subsequences of $k$ consecutive heads is $$N = \sum_{j = 0}^{n-k}X_{j+1}X_{j+2}\cdots X_{j+k}.$$ Do you see why this is true? Now, using linearity of expectation followed by the fact that $X_1,\ldots,X_n$ are independent, we have
\begin{align*}
E[N] &= E\left[\sum_{j = 0}^{n-k}X_{j+1}X_{j+2}\cdots X_{j+k}\right]
\\
&= \sum_{j = 0}^{n-k}E[X_{j+1}X_{j+2}\cdots X_{j+k}]
\\
&= \sum_{j = 0}^{n-k}E[X_{j+1}]E[X_{j+2}] \cdots E[X_{j+k}].
\end{align*}
Can you finish the problem from here?