I have the following problem:
There are N items A(n) and associated weights W(n) for n=0...N-1.
Generate a sequence S(t) (where t= 0,1,2,3,...,T-1) of items chosen from A(n) such that:
Adjacent symbols in the sequence are not equal:
S(t) != S(t-1)
The W(n) describe the frequency of occurrence of A(n) in S over duration T:
W(n) / sum (k = 0...N-1) W(k) = sum (t=0...T-1) indicator ( S(t) == A(n) ) / T
where indicator(x) is 0 if x is false and 1 if x is true.
Does this sort of problem have a name other than weighted scheduling?
This may be related to the term "cyclic executive". Take a look and let me know if this is heading in the right direction.