This third randomly shaped loop design by VE9SRB is the second attempt at achieving better performance than the MI2 Fractal Loop and the first in achieving better performance than the Random2 design in the same physical aperture size.  

As with the Random2 design, the issue in improving performance is increasing the resonant resistance.  This can be accomplished by moving as much of the wire length to the outside of the loop as possible.  At the same time, the total length of wire will have to be adjusted for resonant frequency.

The total wire length in this design is about 41.35 m, which is considerably more than the MI2 or the Random1 design.  The physical aperture size remains unchanged at 2.8 m (X) by 2.66 m (Y).

The first resonance occurs between 14.96 and 14.97 MHz.  The resonant resistance is 41.4 ohms and the antenna gain is 1.91dBi.  The 2:1 SWR bandwidth wrt 50 ohms is about 2.14%.

Again, the tradeoff for improved SWR, gain and bandwidth is additional wire length.

The real question at this point is the following:  With any fixed aperture size and fixed total wire length, does the Fractal geometry provide optimum performance as compared to an arbitrary geometry?  Clearly, based on the fact that one single arbitrary geometry provides performance that is virtually identical to that of the MI2, it is possible that other geometries may provide identical or better performance that the MI2.  The only way to adequately address this issue is to compare several geometries to the MI2.  The work presented by N1IR more than adequately describes the performance of the MI2 and other Fractal geometries.  However, these works do not provide a comparison of Fractal and non-Fractal geometries.  

>From the results obtained with these four EZNEC files, it is clearly obvious that the performance of a small loop antenna is a function of numerous design variables, which include the feedpoint location, the total wire length, the physical aperture and the wire geometry.  The performance of any small loop cannot be solely attributed to the geometry alone.  In any fixed aperture, numerous geometries having the same total wire length will provide similar performance.  In any one instance, a random geometry may or may not provide better performance than a Fractal design.