Representations using negatively subscripted Fibonacci and Tribonacci numbers with applications (Q2883379)

The Fibonacci shuffle mapping \(s_n : [0, F_n - 1] \rightarrow [0, F_n - 1]\) is defined by NEWLINE\[NEWLINEs_n(k) \equiv \pm F_{n-1} k \pmod {F_n}.NEWLINE\]NEWLINE \((b_mb_{m-1} \dots b_2b_1)\) is called a \({\mathcal F}\)-representation of the positive integer \(N\) if NEWLINE\[NEWLINEN = \sum_{k=1}^m b_k F_k,NEWLINE\]NEWLINE where \(b_k \in \{0, 1\}, b_1 = 1, b_l+b_{k-1} \geq 1\) for \(2 \leq k \leq m\), that is no gaps of length more than one. New proofs are given of the assertions that (a) every integer can be expressed as a sum of distinct Fibonacci numbers from the set \(\{ F_n : n < 0 \}\) and (b) every integer can be expressed as a sum of distinct non-adjacent numbers in the set \(\{ F_n : n < 0 \}\). Higher-dimensional analogues of the Fibonacci shuffle that use Tribonacci numbers and vectors and G. Rauzy's fractals are discussed.