1: \begin{abstract}

2: In this paper, we introduce a new class of nonnegative tensors --- strictly nonnegative tensors.

3: A weakly irreducible

4: nonnegative tensor is a strictly nonnegative tensor but not vice

5: versa.   We show that the spectral radius of a strictly nonnegative

6: tensor is always positive. We give some sufficient and necessary

7: conditions for the six well-conditional classes of nonnegative

8: tensors, introduced in the literature, and a full relationship

9: picture about strictly nonnegative tensors with these six classes of

10: nonnegative tensors. We then establish global R-linear convergence

11: of a power method for finding the spectral radius of a nonnegative

12: tensor under the condition of weak irreducibility. We show that for

13: a nonnegative tensor $T$, there always exists a partition of the

14: index set such that every tensor induced by the partition is weakly

15: irreducible; and the spectral radius of $T$ can be obtained from

16: those spectral radii of the induced tensors. In this way, we develop

17: a convergent algorithm for finding the spectral radius of {\it a

18: general nonnegative tensor} without any additional assumption. The

19: preliminary numerical results demonstrate the feasibility and

20: effectiveness of the proposed algorithm. \vspace{3mm}

21: 

22: \noindent {\bf Key words:}\hspace{2mm} Nonnegative tensor, spectral

23: radius, strict nonnegativity, weak irreducibility, algorithm

24: \vspace{3mm}

25: 

26: \noindent {\bf AMS subjection classifications (2010):}\hspace{2mm}

27: 15-02; 15A18; 15A69; 65F15

28: \end{abstract}