We should not, therefore, be asking subjects how, why, and by what right they can agree to being subjugated, but showing how actual relations of subjugation manufacture subjects. 
So his analysis of sovereignty is going to have little to do with things like compacts, charters, constitutions, etc. -- or at least as we consciously perceive them -- and more to do with the role of power relations in creating subjects. To that extent, there's a remarkable similarity to Althusser's theory of Ideological State Apparatuses, though Foucault is quick to criticize the totalizing aspect of Althusser's theory ("I think we can analyze them effectively only if we do not see them as an overall unity, only if we do not try to derive them from something like the Statist unity of sovereignty" . -- a pretty clear critique of the Althusserian ISA).
Working backward, we can also attribute Foucault's methodology to Marx's famous dictum from A Contribution to the Critique of Political Economy (1859): "It is not the consciousness of men that determines their existence, but their social existence that determines their consciousness." However, Foucault would still see the prevailing Marxist interpretations of culture as too totalizing, as he preferred the play of systems against one another and the power relations that came out of competing systems, or as by-products of those systems -- or I suppose more to the point, the interplay of forces that throught their interplay create systems. In the end, though, I don't see much difference, because his mapping out of the regimes of surveillance, from the panopticon to the clinic, could all contribute to the logic of capitalism -- as opposed to the logic of the ruling class.
Anyway, before I get too far afield, Foucault takes "sovereignty" as his ostensible topic, but the exploration of sovereignty, while interesting in itself, also moves him toward a more general goal of establishing historical discourse as a battlefield: historical interpretation, historical representation, inclusion or exclusion as "world historical" are the ground where an eternal war is fought and never completely won or lost.
Perhaps that sounds simple, and maybe it is. Nietzsche more or less argues the same thing in the 19th century, but Foucault is making a much more sustained analysis than Nietzsche ever did, and in the end his point isn't simply that historical discourse is the location of interpretive battles, but also that it can only be so. There is no fixed finality. The battle is never ever over.