HF Surface Wave Radar (HFSWR) seems to be the most relevant and lowest cost solution for the surveillance of the Exclusive Economical Zone [1]. The HFSWR is a land based system that can watch over large maritime areas and detect low-altitude aircrafts and surface vessels beyond the electromagnetic horizon [2]. However, one of its major problems lies in the emitting antennas: since HF wavelength extends between 10m and 100m (most often in the order of 30 m), antenna dimensions can become a restraint for radar deployment. For this reason, equivalent quarter-dipole antennas are usually employed. This simple solution lacks in directivity and therefore causes energy dispersion towards the sky. As a consequence, ionospheric clutter corrupts radar data. We aim to solve those problem (i.e., energy wasting and clutter corruption) right from the conception. We are of the opinion that the design of a radiator dedicated to surface wave excitation cannot be achieved if the electromagnetic problem is ill-posed. In a recent paper [3] we have introduced the modal decomposition method [4] to study the in°uence of the Zenneck wave on the ¯eld excited by any vertical source placed at the sea surface. It has been found that the Zenneck wave contribution is masked by an in¯nite spectrum of bulk waves that are not con¯ned at the sea surface. The excitation of the bulk waves does not permit to concentrate the radiated power at the interface between sea and air and therefore it lowers HFSWR performances. An interesting solution could come from the domain of negative parameters materials: an interface between air and a material having a negative permittivity can support a complex mode that may be not masked by the bulk waves. In this paper, with a similar approach of [5], we propose two di®erent structures capable of sustaining a guided complex wave and of concentrating the radiated power at lower elevation angles.