The author in this paper discusses the possibility of detecting artificial non-spherical objects with Kepler and COROT, including single objects and multiple objects. The three single object cases discussed in this paper are triangle, two-screen and louver-like six screen shaped objects. The transit depths generated by their simulations are on the order of 100ppm which is detectable with Kepler. Then the author moves on to discuss detecting multiple-object transit signals, specifically, multiple transits, grouped by prime numbers.
Further, the author discusses the efficiency of using the above mentioned artificial transits as a communication tool. The author compares this way of communication with laser beacons and finds similar communicating efficiencies. The author additionally shows that with our current technology, to communicate with each target star using laser pulses, the time required is on the order of days. The communication efficiency will be much improved now since the launch of GAIA satellite.
Finally, the author argues that transit signals will be used for attention-getting and laser pulses will be used for data transfer since it is more directional.
There are several limitations to this paper, first, in the communication efficiency part, the author does not take into account the factor of distance. Second, in the communication efficiency part, the author assumes ETI has knowledge of proper motions and distances to the target stars which is not always the case.
The main concern I have with this paper is that the author does not at all discuss what could mimic the artificial signals during the limb-darkening period for natural sources such as stellar activities.