See, for example, the Java Primitive Variables article.
In short (sic), it is similar to int, but... just shorter!

just multiply your numbers (-1 to 1) by 2^15 - 1 (to give you a range of -(2^15 - 1) to (2^15 - 1)) and store the result, without any fractional part, in a short array of the same size as the float array and pass that to the method.

 

cool thanks!