We present a new observational campaign, DWARF, aimed at detection of circumbinary extrasolar planets using the timing of the minima of low-mass eclipsing binaries. The observations will be performed within an extensive network of relatively small to
medium-size telescopes with apertures of ~20-200 cm. The starting sample of the objects to be monitored contains (i) low-mass eclipsing binaries with M and K components, (ii) short-period binaries with sdB or sdO component, and (iii) post-common-envelope systems containing a WD, which enable to determine minima with high precision. Since the amplitude of the timing signal increases with the orbital period of an invisible third component, the timescale of project is long, at least 5-10 years. The paper gives simple formulas to estimate suitability of individual eclipsing binaries for the circumbinary planet detection. Intrinsic variability of the binaries (photospheric spots, flares, pulsation etc.) limiting the accuracy of the minima timing is also discussed. The manuscript also describes the best observing strategy and methods to detect cyclic timing variability in the minima times indicating presence of circumbinary planets. First test observation of the selected targets are presented.
