That’s an interesting question. Actually, there are two main reason for replacing patch antenna with DRAs at mm-wave frequency spectrum. The first reason as you said, is that the antenna efficiency is better in DRAs compared to the patch antenna; however, it is very difficult if you want to prove the mathematical relation between the metallic losses and frequency with exact closed formula. Anyway you can simulate both antenna in a full wave simulator like HFSS to see the difference. At low frequencies (microwave frequencies), antenna efficiency is not that much critical; however, at mm-wave frequencies it is vital due to the fact that the propagation losses are huge, and as a result we need to high performance transmitters and receivers.

The second reason that the DRAs are more interesting in mm-wave applications is the fractional bandwidth. For example, mm-wave 60GHz frequency band system works from 57GH to 64 GHz, which is 7GHz bandwidth. A single conventional patch antenna feed by a quarter-wave transformer or with other types of feeding can give you about 2GHz bandwidth at 60GHz; however a single conventional DRA can cover the entire frequency band from 57GH to 64GHz.

That’s why we usually use DRAs at mm-wave frequencies.

This is a very relevant question, mainly because of the current push from the industry towards mm-wave 5G technology.

In addition to the excellent points mentioned by Farahani, I would like to add that microstrip technology in general, is inherently low-pass due to basic transmission line mechanism. The equivalent circuit of the transmission line unit cell makes it pretty obvious.

This is one of the reasons why planar waveguide technology is becoming more popular for high frequency mm-wave applications, as my colleagues working on SIW-fed slot-antennas or DRAs suggest. Since standard microstrip patches are probe-fed by co-ax or inset-fed by microstrip lines, their operation is limited to the fundamental cut-off of these transmission lines.

@Mohammadmahdi, I think it's not too difficult to calculate losses of metals at optical frequency. You just have to understand the physics behind the losses in metals. The loss components in metals come from imaginary part of epsilon. When you see the epsilon of metals at optical frequency, the imaginary part become very large. For an instance, gold epsilon at 100 THz is (-300 + J*60), so this large imaginary part is responsible for losses. The losses or simply called absorption in metals can be calculated from the following formula as: L = 0.5*wmega*E^2*(imag(epsilon)). In general, you can compare the losses of metals and dielectric at same frequency. Remember, dielectric also have losses because of same reason. We are going to published something on this topic soon, so stay tune :) .

I agree with Vishal Vashistha. You can try to characterize the metal with the Drude model. You are going to find the mathematical explanation for the metal loss, which becomes higher at mm waves.

Losses are due to Skin Effect. The penetration depth for a conductor decreases with the square root of frequency :

d=sqrt(rho/pi.freq.mu)

Dear Djamel Sayad ,

How penetration depth effects the losses in conductor? Can you explain it in detail?

The electric current flows mainly between the outer surface of the conductor and a level called the skin depth. Skin depth inversely proportional to square root of frequency. So, at mm-W frequencies, as skin depth decreases, current flows only the upper surface of the material. Hence, it increases effective resistance of the conductor. Consequently, it reduces effective cross-section of the conductor, which in turn increases losses.

I agree with Debdeep Sarkar. SIW fed dielectric resonator antennas are now promising approach for mm-W 5G technology as it minimizes parasitic radiations.

Hello,

In general Dielectric Resonator antennas having more attractive features compared to general microstrip patch antennas. The list of key advantages of DRAs over Microstrip patch antennas:- Much wider Bandwidth, More over operating Bandwidth of a DRA can be varied by the permittivity of the resonator material and its dimensions, Because DRA radiates through the whole antenna surface but in case of microstrip patch antenna radiates only through patch, Avoidance of surface wave and metal losses, DRA’s have high dielectric strength.

Because of that DRA is preferred over MSA, Hence DRA’s having higher power handling capacity and DRA’s can operate in a wide temperature range.

The main attractive feature in Dielectric Resonator antennas is Bandwidth Enhancement. By choosing proper structure for DRAs, we can easily increase the bandwidth.

Thanks,

Most of the papers and books (K W Leung and Aldo Petosa's books) reveal the lots of advantages of DRA over Microstrip Antenna (MA) which includes

1. high gain,

2. high radiation efficiency

3. low conductor loss

4. negligible amount of surface wave loss etc etc which are already discussed here.

with my limited knowledge, I would like to add another two important points which can be used to describe the use of DRA at high frequency over MA.

1. Resonant Frequency: for simplicity, consider one rectangular MA (RMA) and Rectangular DRA (RDRA) having same dimensions i.e. 30mm X 20mm X 0.762mm. RMA shows fundamental TMz_10 mode at at 2.8GHz (aaprox) for er = 2.32 whereas RDRA shows fundamental TEy_111 mode at 5.5GHz. (approx.) for er = 100. Please find the attachment ( MA_DRA.jpg ).Therefore, if we use er = 10, resonant frequency will shift from 5.5GHz to 55GHz. Therefore, using low-profile DRA, achievement of high frequency is very simple for DRA.

2. Size: Now, if we want generate 55GHz, the size of the RMA must be reduced by 10 times. Such small size of RMA may not be excited efficiency using conventional excitation system.

Excitation of higher order modes may solve this small dimensional problem for RMA but we can also excite higher order mode of the DRA.

Further, due to finite variation along the height of the DRA, DRA shows more no of modes compared to RMA for same frequency band.

For Skikn effect see the attached file

The calculation of different losses (conductor, dielectric and radiation loss) are explained in detail in the book: Handbook of Microstrip Antennas_James & Hall_1989 in Chapter 3. If you still need extra for conductor loss, please consult the book: Advanced Engineering Electromagnetics_C A Balanis_1989 (rectangular waveguide chapter)

Interesting question and nicely answered by all the commentators.

Despite all the great potentials of dielectric antennas, a major reason why DRAs are replacing MSA at high frequencies is reduction in relative size. At low frequencies, MSA has been a preferred technology as the 3D structure of a DRA is too significant to be used in practical applications. However at higher frequencies, the DRAs are becoming structurally feasible to be used in devices and systems. So in addition to their performance benefits, the size compatibility is a strong factor enabling wider implementation of DRA based systems.

This is the summary of all the answers given by the members:

1) DRA are more efficient with better fractional bandwidth at higher frequencies.

2) the relation between Conductor losses and frequency can be understood by either drude model or simple skin effect relation.

3) At mm wave frequencies the relative size of DRA is lesser than metallic antenna.

Dear Shailendra,

I would like to thank you first for adding extra points here. But, the 3rd point may not be true (3. At mm wave frequencies the relative size of DRA is lesser than metallic antenna. ). DRA has non-zero values of kx, ky and kz whereas patch has kz = 0 and kx and ky are non-zero.

Therefore, for a given resonant frequency, size of DRA > size of patch.

Sudipta Maity , I agree with that fact that K vector is non-zero in all three direction for DRA but I suppose, Shailendra Singh is referering to relative size of antenna. Rememebre the size of DRA is scaled by Lambda/ (2*n), where n is refrective index of DRA. Ofcourse the antenna will be three dimensions, since the Kz cannot be zero for DRA. So we can scale the diemnsinons of antenna by a factor of n for getting same K-vector. Also, there are further certain techniques for size reduction in DRA, which may not be possible in case of patch antenna. I recomends the following articles for further reading.

Correct me, if I misundesrstood something.

https://www.hindawi.com/journals/ijap/2016/6075680/

Article Size-Reduction Method for Dielectric-Resonator Antennas

Dear Vishal,

Don't misunderstood the point no 3. Applying symmetry is different. These two are different cases. Consider RMA and RDRA having dimensions a = 30mm,b = 20mm, d=0.762mm and er. The RMA shows fundamental mode at approx. 3GHz for er = 2.32 whereas RDRA shows fundamental mode at approx. 5GHz for for er = 100. If we take lower er (equal to 2.32), RDRA will show more than 30GHz.

Suppose, take this sample for this case. If we want to get 30GHz using that patch (er = 2.32), we have to reduce the size.

The value of er = 2.32 is just taken for example. DRA is normally made with er >= 10

Hi Sreenath,

Both the antennas are very close in some aspects. They both have similar feed mechanism and radiation patterns. But, at higher frequencies DRA is preferred over MPA because of higher gain and wider bandwidth when compared to MPA. The following paper provides useful info regarding DRA and MPAs

Microstrip Patch versus Dielectric Resonator Antenna Bearing All Commonly Used Feeds: An experimental study to choose the right element.

Gain flatness is also a matter to differentiate DRA with MSA at mm wave frequency which will be better in DRA. polarization flexibility will available in DRA by verity of current and field controlling mechanism without affecting other parameters.

One major drawback with metallic-based printed MIMO antennas is their relatively low efficiency (not exceeding 65%) .

The most suitable antenna to receive the microwave signals.

DRA has high gain, high rad. efficiency etc etc which are well known. new thing is that:

DRA supports higher order mode in the form of TE_mnp or TM_mnp or HEM_mnp whereas PAtch supports only TM_mn0 mode (p = 0) due to low thickness of substrate. Hence, DRA gives more flexibility to use higher order mode using large dimension of the antenna. Besides, good isolation is found between co-pol and cross-pol at higher order mode also.

In case of Microstrip Antenna, we try to use TM_10 mode i.e. fundamental mode, due to that reason, size of MA becomes too small which creates problem in practical realization.

Also, a metamaterials'[1] ("photonic"[2] like) antenna behaves, as if, it were, effectively with, a much larger (than its, geometrically, appearing) size, because its novel structure is able to store[2] the (high freq. EM-signal) energy in a special mode. This mode works not as a dummy quasi-FREE single (drude[3,4]) charge found in conventional antennas, but as a (damped, multi-charges) oscillator system.

So, a very low(er) power EM-signal might[5] be able to give (or lose[6]) more energy[7] stored[8] in a time period of some (multi) n-periods, n*To, where To=1/Fo, (Fo in your **GHz, range), e.g. a metamaterial antenna oscillator, effectively, functions better having a higher energy (due to a driven damped oscillator, modeling) gain[9].

1. Metamaterial antenna https://en.wikipedia.org/wiki/Metamaterial_antenna https://www.wisegeek.com/what-is-a-metamaterial-antenna.htm

2. https://www.sciencedirect.com/book/9781782424642/fundamentals-and-applications-of-nanophotonics

3. Energy of driven damped oscillator https://www.physicsforums.com/threads/energy-of-driven-damped-oscillator.292063/

4. Drude_model https://en.wikipedia.org/wiki/Drude_model

5. Lorentz-Drude Model https://optiwave.com/optifdtd-manuals/fdtd-lorentz-drude-model/

6. Physics of Resonance: Tacoma Narrows Bridge Collapse https://study.com/academy/lesson/physics-of-resonance-tacoma-narrows-bridge-collapse.html

7. THE DRIVEN, DAMPED HARMONIC OSCILLATOR https://physics.oregonstate.edu/~tatej/COURSES/ph421/lectures/L9.pdf

8. Energy in a Damped Harmonic Oscillator http://www.entropy.energy/scholar/node/damped-harmonic-oscillator-energy

9. METAMATERIAL ANTENNA DESIGNING AND ITS APPLICATION http://aguijet.com/images/short_pdf/1478594980_Ruchi_Thakur_3.pdf