[hvalentim]

It also seems antenna gain has a relation with directivity. The more "directive" the design the higher the gain, the drawback being you need to more accurately point the antenna to the source of the signal.

Performance wise, it seems then a Yagi should be the best option for signal reception.

A Yagi can actually be made to look "cool"/decorative:

source : Antenna for Superheterodyne pin???

The issue being Yagis tend to be quit large/occupy a lot of space. Since the signal will come from above/the roof I guess it would be placed vertically (???) which would help.

According to this calculator it is perhaps possible to have a 6dB gain with a design "as small" as 24 x 17 cm (results below).


**** Logarithmic Periodic Dipole Antenna ****
http://www.changpuak.ch/electronics/lpda.php
JavaScript Version 12.01.2014, based on ARRL ANTENNA BOOK, 17th ED.
------------------------------------------------------------------------------
Min. Frequency : 433 MHz
Max. Frequency : 434 MHz
Boomlength : 237 mm
Gain : 6.70 dB
Elements : 4
tau : 0.792
sigma : 0.142
------------------------------------------------------------------------------
Bandwidth (active region) : 2.01
Bandwidth (structural) : 2.01
Bandwidth (effective) : 1.00
------------------------------------------------------------------------------
Element 1, Rod : 173 mm, Position : 0 mm, Diameter : 3 mm
Element 2, Rod : 137 mm, Position : 98 mm, Diameter : 3 mm
Element 3, Rod : 109 mm, Position : 176 mm, Diameter : 2 mm
Element 4, Rod : 86 mm, Position : 237 mm, Diameter : 2 mm
------------------------------------------------------------------------------


More interesting stuff, often referenced:

---

Another "trick" on the transmitter part seems it could be to increase voltage to 12V.

Judging by the look of the RX board (below), seems this is already the case?

   

"433MHz Superheterodyne 3400 RF Transmitter and Receiver
Transmitter Specs:
Working voltage: 3V~12V"