The circuit can be optimized by avoiding the storage and evaluation of full-degree N-1 polynomials when many coefficients are zero.
Currently, message polynomials are encoded with all N coefficients, even though in practice their degree is often much lower (the minimum degree for the plaintext is $2^{28}$ now).
The circuit can be optimized by avoiding the storage and evaluation of full-degree N-1 polynomials when many coefficients are zero.
Currently, message polynomials are encoded with all N coefficients, even though in practice their degree is often much lower (the minimum degree for the plaintext is$2^{28}$ now).