It's a somewhat different kind of ring buffer, because there's just one index, but I used it in my signal processing class for a finite-impulse-response filter.
Choose N to be a power of two >= the length of your filter.
Increment index i mod N, write the sample at buffer position x[i], output sum of x[i-k mod N] * a[k] where a[k] are your filter coefficients, repeat with next sample at next time step.
Comments
It's a somewhat different kind of ring buffer, because there's just one index, but I used it in my signal processing class for a finite-impulse-response filter.
Choose N to be a power of two >= the length of your filter.
Increment index i mod N, write the sample at buffer position x[i], output sum of x[i-k mod N] * a[k] where a[k] are your filter coefficients, repeat with next sample at next time step.