https://arxiv.org/abs/2102.12951 (or view on SciRate)

Our paper provides a method to compile any quantum walk (i.e. any banded unitary matrix, possibly infinite, and not translation invariant) to a finite sequence of shift and coin operations on a cell structure which can be chosen at will (or according to experimental limitations).

Important here is to note that the notions of “shift” and “coin” operations make sense only after a cell structure is fixed.

Of course, if you define a quantum walk (say a Grover walk) from the outset as such a protocol the question is trivial *on that cell structure*, but it is non-trivial to find such a protocol on some (any) other cell structure.

Moreover, our algorithm does not require that a walk is given in the beginning by shifts and coins - any banded unitary is fine.

tags:

quantum-walks