is multiobjective xcorr possible?

multiobjective xcorr?

I'm hoping to automate the synchronisation of signals from two very similar sensors, s_1 and s_2,which were used to record the same experimental stimulus, but with the following complicating factors:

1) While s_1 recorded data continuously, s_2 only recorded short segments of data.

2) The stimulus was approximately periodic and so comparing a single segment of data from s_2 with the signal from s_1 using xcorr gives periodic similarity scores.

However, the time-ordering of the segments recorded by s_2 is known, as well as the fact that the segments from s_2 are non-overlapping.

Can I use use xcorr to define a multiobjective optimisation problem where these time-ordering+non-overlapping conditions are imposed to recover the correct result?

Below is an attempt to show the problem visually (text formatting allowing?):

s_1 = [1 2 3 4 3 2 1 2 3 4 3 2 1 2 3 4 3 2 1];

s_2 ={[1 2 3 4], [2 1 2], ___[2 1 2 3],__[2 1]};

[Most simplifying assumptions are acceptable, e.g., the largest gap between s_2 segments is small versus the period of the stimulus, the shortest length of s_2 segments is large versus the same.]