Hey @jeno, glad to hear about your interest in the subject... To answer your question, well... That's the ultimate question. I've seen a few people get great results from using a simple lawn-light solar panel as a power source going directly into the soil. These typically output something between 1.5 and 3V, for the small panels.
DIY Experimenting
To start experimenting, all that's needed is a power source like that, or perhaps an old cell-phone charger power supply or even a D-cell battery (the big one is best so it has enough charge to last for awhile). For electrodes, all you really need are some iron nails... Stainless is fine, but there may be toxicity issues depending upon how they react in electrolysis reactions with the soil-water.
Regarding the link, sorry that it's a for-fee one... You may be able to find some snippets of interest through google books, or find a way of checking it our for free at a university library. I have no affiliation with them, just figured its a good resource to share (but I don't have the right to share the file underneath).
Optimizing Output
In terms of a circuit to try out with regards to optimization, I see 3 approaches:
1) use a cell-phone charger power charger on an hourly interval timer (this will probably be sub optimal, but may do well (giving the plant an hourly break from stimulation stress)
2) use a microcontroller circuit like an arduino
3) use a simple 555 timer circuit that you can find online to create a series of pulses.
