My impression is that it's not at all that consistent. The total spending is constant at a few billion per year, but what that gets spent on depends entirely on the taste of the program managers, which in turn depends on the darpa director. The program managers only serve for a few years at a time, and the director for a bit longer, so it can change drastically. I think the trends are typically last a decade: in the 1980s after the success with the F-117, they were funding all kinds of airplane projects. Also in the 1980s they funded a lot of AI research, until they suddently didn't (the "AI winter"). In the 1990s, they were into networked combat simulators and command & control systems.
Starting around 2010, DARPA suddenly began to fund a lot of computer science research on formal verification. I was a phd student at the time, and it was pretty striking, programming language theory used to be a really fringe subject, and now suddenly it was awash in cash and all the professors could hire huge herds of postdocs. But some of those professors also told me that I should not expect it to last: sooner or later there will be new people in DARPA and all the money will instantly stop.
Certainly it's up to the vagaries of fate and such, but when I was in school I worked in a lab that for several years got fairly reliable DARPA funding for NLP work, and while they never built the department around it or anything it was nice to have that consistent year over year assumption that one or two projects would be interesting to the DARPA folks.
For them (us?) it was more a matter of "well this doesn't necessarily merit or reach the requirements for quote-unquote 'real' research, but it's really interesting and you know some investigation couldn't hurt". Those got pitched to DARPA, they asked "can you do it with Arabic and Chinese too?" and we said "no" and they funded us anyway a few times out of ten. Not bad, really.
Starting around 2010, DARPA suddenly began to fund a lot of computer science research on formal verification. I was a phd student at the time, and it was pretty striking, programming language theory used to be a really fringe subject, and now suddenly it was awash in cash and all the professors could hire huge herds of postdocs. But some of those professors also told me that I should not expect it to last: sooner or later there will be new people in DARPA and all the money will instantly stop.