Yes, the result is very nice. Indeed, after deriving the first order expansion on the marginal distribution of each Wishart eigenvalue,I find that the exponent of $x$ remains the same even in the presence of spatial correlation and line-of-sight effect. In other words, the diversity order of each MIMO sub-channel is independent of the two factors above. This is very interesting and somewhat amazing, at least to me.
PS. Thank you, Rahul Vaze, for the nice paper. Also, WindoWSIL is a nice place. ^_^

]]>