Unitary groups are very important objects in the setting of Langland's Conjecture because of the existence of Shimura Variety ( which I don't know) and also because people know how to attach a galois representation to a automorphic representation of unitary groups in almost all cases (By the work of Taylor, Harris and many others)(Which also I don't know).

I was trying to learn about unitary groups attached to $(k,E,D,*)$ where say $k$ is a totally real field,$E$ CM field of deg 2 over $k$, $D$ is a central simple algebra of rank $n^2$ over E, and $\*$ $k$ algebra anti-involution of 2nd kind on $D$ (i.e. coinciding with the action of non-trivial element of $Gal(E/k)$ on $E$).

But I could not really find a reference for this. Essentially some authors define it a inner form of a particular quasi-split unitary group, and some authors define it as functor of points. Also it is commented that there is some sort of Global-Local patching going on. Can any one give me a reference where unitary groups is covered in some what details rather than a overview in 2 pages?