Terahertz Spectroscopy, Imaging, and Wireless Communications Lab

The increase in demand for high speed wireless communication systems, and the subsequent push to the Terahertz (THz) frequency carrier range has driven component development including sources, modulators, steerable mirrors, and detectors. Since diffractive effects in the far-field are much less prevalent in the THz frequency range compared to the few gigahertz range of current wireless systems (See our work on Terahertz Power Beaming), THz communications will be inherently line-of-sight (LoS) based. For certain scenarios, such as outdoor wireless links between buildings, this restriction is virtually no issue. However, LoS blockages may prove to be problematic. The essential freedom of maneuver provided by mobile receivers means that obstacles to the signal are inevitable. Mobile users can temporarily lose the signal if blocked by static or moving obstacles. For indoor situations, furniture, people or other objects can block LoS transmission.

Non-line-of-sight (NLoS) communication will require new technologies that can maintain the link between transmitters and receivers. When LoS links are blocked, one solution is to use a NLoS path by reflecting THz beams from walls or ceilings. We have adapted a new approach to THz mirror design for wireless communicatisis investigated: Van Atta array reflectors [6]. As an initial step in the development of THz Van Atta arrays for indoor THz wireless links, the first step is to design, fabricate, and characterize a passive planar retroreflector for the THz frequencies. Introduced in 1959, Van Atta arrays consist of antennas connected in pairs that retroreflect incident radiation. For proper operation, the antennas within each pair must be diametrically opposed about the center and connected by transmission lines of equal electrical length (integer multiples of the wavelength). Using this design structure, the emitted radiation, even from different antenna pairs, align in phase. The arrival of the wavefront at an angle θ causes a phase delay in the wavefront incident on the antennas. Since the pairs connected are opposite one another about the center, the phase delay in any receiving antenna is offset by the corresponding phase delay in its paired transmitting antenna and vice versa. There is potential to include electric switching and amplifying capabilities within the Van Atta structures, shown in the figure below, suggesting a promising approach to developing Van Atta structures as “smart” steerable mirrors.

The figure belwo is an illustration of a simple 2x2 Van Atta array for 200 GHz. The overall size of the structure is roughly 22mm square.

By repeating the 2x2 pattern into an array of van Atta reflectors, one can increase the refectivity of the structure. The graph below is the normalized RCS value at 200 GHz for a linear array of 2x2 Van Atta unit cells as a function of polar angle. The solid black, solid gray, dashed black, and dashed gray lines correspond to 1, 4, 6, and 8 unit cells, respectively. Note the increase in the RCS value with the number of unit cells or equivalently the area of the linear array. For comparison, the RCS value for a 22mm square flat metal plate is shown (dotted line with open circles).

Designing a 4x4 or larger Van Atta cell structure that can be printed onto a small space introduces technical limitations: not only do the wirings that connect pairs of Van Atta antennas become increasingly complicated, but the proximity of two stripline wave guides leads to performance degradation: specifically, the presence of one stripline structure in a printable structure changes the impedance of nearby stripline structures. To circumvent these technical difficulties, a variation in the Van Atta design, as shown in figure below was fabricated. The antennas are interconnected in one direction serially. Symmetric columns of antennas are connected with the typical Van Atta arrangement in the orthogonal direction. This Van Atta inspired structure is scalable to larger sizes since it minimizes the complexity of interconnects between antennas.

In the figures below is shown the nScrypt printer fabricating the van Atta inspired reflector. A comparison bewteen design and actual dimensions is also shown.

In the figure below are shown measured averaged values for resonant (black dashed) and non-resonant bands (gray dashed) for the parallel polarized Van Atta retroreflector. The black dashed line follows the trend of the Van Atta simulation (black dashed), averaging above -10dB out to a considerable angle. This a clear indication of Van Atta operation, since the non-resonant bands behave as flat plates (gray curve).