# Linear independence in Finite Fields

How can we define linear independence in vectors over $\mathbb{F_{2^m}}$ ?

Let vectors $v_1,v_2,v_3$ $\in$ $\mathbb{F_{2^m}}$,

If $v_1,v_2,v_3$ are linearly independent,then $\alpha_1v_1+\alpha_2v_2+\alpha_3v_3$=0 if and only if $\alpha_1=\alpha_2=\alpha_3=0$ and $\alpha_1,\alpha_2,\alpha_3 \in \mathbb{F_2}$ ? or $\mathbb{F_{2^m}}$ ?

-
One comment: when making definitions, you shouldn't use the biconditional "iff". The first "if" there is also superfluous. – rschwieb Jan 24 '13 at 18:33
The edit seems to have made the formulation of the definition worse. In particular, an "iff" became "if" where what's really needed is "only if". – Andreas Blass Jan 24 '13 at 18:40
When you talk about linear independence, you need to mention what is your vector space, is it $F_{2^m}$ over $Z_2$? – mez May 10 '13 at 23:31

Linear independence is defined the same way in every vector space:

$\{v_i\mid i\in I\}$ is a linearly independent subset of $V$ if $\sum_{i=1}^n \lambda_i v_i=0$ implies all the $\lambda_i=0$ for all $i$, where the $\lambda_i$ are in the field.

In short, you definitely would not take the $\lambda_i$ from $F^m$. You are probably thinking of multiplying coordinate-wise. The definition of a linear combination, though, takes coefficents from the field (and $F^m$ is not a field).

Linear independence depends on the field (no pun intended.) If you want them to be linearly independent over $F$, then $\lambda_i$ can only come from $F$. If you want it to be linearly independent over $F_{2^k}$, then the $\lambda_i$ are all from $F_{2^k}$.
For a simple example, look at $F_2$ and $F_8$. If $x\in F_8\setminus F_2$, then $\{1,x\}$ is linearly independent over $F_2$, but it is linearly dependent over $F_8$.
Then the answer is $\alpha_1,\alpha_2,\alpha_3 \in$\mathbb{F_{2^m}}$? Sorry I could not understand – elif Jan 24 '13 at 18:37 Thank you for your answer. But if I have$k \times k$matrix, named$A$, such that each element is selected from$\mathbb{F_{2^m}}$. In which case$A$is invertible? if$A$'s rows are linearly independent over$\mathbb{F_{2^m}}$or over$\mathbb{F_2}$– elif Jan 24 '13 at 19:19 @elif In that case, if they are independent over$F_{2^m}\$. There is another way to see if it's invertible: check to see if its determinant is nonzero. – rschwieb Jan 24 '13 at 19:23