How do I interpret haplotype network from software Network 5.0?

should be "median vector", as I guess

Simply read user guide: 

"A median vector is a hypothesised (often ancestral) sequence which is

required to connect existing sequences within the network with maximum parsimony. Without the median vector, there would be no shortest connection between the data set's sequences."

 Hey thanks Maria. But my point is does it take ancestral sequence (median vector) from the sequences which I have submitted .I am sure it does then  how do I know which one it has taken as ancestral sequence.

I am not an expert in the Network, but as far as I know there is no strict hypothesis about ancestral sequence, unlike you designate it yourself. The program just connects your haplotypes by the shortest way, and you have to decide where is the ancestor.  it's not obligatory  that the ancestor is the median vector. It could be one of your real haplotypes - most numerous, located deep in the tree, the closest to other species, given rise to many other haplotypes, etc.  

Well, I did not work with this software for a long time. Maybe it has gained now the feature of finding the ancestor... 

Median vectors (mv) , representing missing or not sampled haplotypes