A problem of the ,,monotone'' CFTP algorithm discussed in
Sections 3.5.3 and 3.5.4 is
the necessity to save all innovations
where denotes the coupling
time defined in (95), i.e.
Therefore, in the year 2000, David Wilson suggested the following
modifications of the CFTP algorithm aiming at a reduction of the
necessary memory allocation.
The main idea of the modification is to realize coupling to the
past (see Sections 3.5.2 - 3.5.4)
based on a sequence of independent and identically distributed
blocks of ,,forward simulation'', where
the (potential) ,,initial times''
of the
Markov chain
can be picked at random.
The innovation sequences
are
chosen identical with probability ,
i.e., we merely consider a single sequence
of independent
and uniformly distributed random variables and define
.
Furthermore, we assume that the Markov chains
and
defined by (87) and (90) have finite
forward and backward coupling times
bzw.
with probability .
Now we consider blocks of forward simulation of (at first
deterministic) length for some .
For arbitrary and
, let
and
Furthermore, for each we consider the event
where the length of the blocks is chosen such that
(105)
Starting at the read-once modification of the CFTP
algorithm is given as follows.
Simulate
via
and
for
.
Set and . If the event has occurred
proceed with step 3, otherwise return to step 1.
Repeat steps 1 and 2 until the event
occurs for some
and return the value of
for an arbitrary
as a realization of
.
Example
For states we consider the irreducible and aperiodic
transition matrix
For block length and the -uniformly distributes
pseudo-random numbers
As the simulation blocks and hence the events
are independent and as
,
the first blocks of forward simulation of the algorithm
described above in particular yield the coupling from the past
discussed in Section 3.5.2 if they are considered in
reversed order.
Therefore,
for all
.
The last, i.e. the
st block of forward simulation
serves only to define a stopping rule.
The read-once modification of the CFTP algorithm terminates with
probability
having the same distribution as the forward coupling time and
which is independent of the innovation sequence
,
then by the following elementary but useful properties of the
coupling times and we obtain
.
Theorem 3.25 The random variables
and have the same distribution, i.e.
. Moreover, if the coupling times and
are independent and almost surely finite, then
(106)
Proof
By the homogeneity of the Markov chains
and
, for any natural number
we have
Let now the coupling times and be independent and
finite with probability .
This implies
where the last equality follows from
which has
been shown in the first part of the proof.