e1fa8ef99e7ead98.tex
1: \begin{abstract}
2: Quantum walks, both discrete (coined) and continuous time,
3: form the basis of several recent quantum algorithms.
4: Here we use numerical simulations to study the properties of discrete,
5: coined quantum walks.
6: We investigate the variation in the entanglement between the coin and
7: the position of the particle by calculating the entropy of
8: the reduced density matrix of the coin.
9: We consider both dynamical evolution and asymptotic limits for coins of
10: dimensions from two to eight on regular graphs.
11: For low coin dimensions, quantum walks which spread faster (as measured by
12: the mean square deviation of their distribution from uniform)
13: also exhibit faster
14: convergence towards the asymptotic value of the entanglement between
15: the coin and particle's position.  For high
16: dimensional coins, the DFT coin operator is more efficient at spreading
17: than the Grover coin.
18: We study the entanglement of the coin on regular finite graphs such as
19: cycles, and also show that on complete bipartite graphs, a quantum walk with a
20: Grover coin is always periodic with period four.
21: We generalise the ``glued trees'' graph used by Childs et al.~[STOC, 59, (2003)]
22: to higher branching rate (fan out) and verify that the scaling with 
23: branching rate and with tree depth is polynomial.
24: \end{abstract}
25: