 So good morning to everyone, my name is Tomasz Wolenski and I'm a product manager and a marketing manager at RF Elements. And today we'll continue with our series of webinars on antennas 101, today about the patch race. So in the previous webinar, we looked in detail at the horn sectors and today we'll speak about the patch race sectors. So where does the name come from? What they're built from, how do they function? And we'll look at their advantages and disadvantages in terms of application in unlicensed five gigahertz with networks. To begin with, take the word sector. So why is an antenna called sector? Because it's beam width in the azimuth plane it's not 360 degrees, meaning that it's not an omni. But only a part of the whole circle is covered. And this part is called sector, that's it. So sector is not an inherent quality of patch array antennas. It has to do with the beam width in the azimuth plane. Just a small recap from the previous part about horns. A great sector antenna has plenty of parameters that WISPs would ideally observe in order to choose the best possible antenna for a particular application. But of course we understand compromises are inevitable as you move in your daily life. But if you happen to have the time to be thorough definitely look through these parameters and give it a minute to see whether besides gain, beam width and broadband width, any other parameters might make sense to consider. And especially the amount of side loops when if the noise is what you're battling with in your networks. So patch arrays and horns are very different types of antennas which results in fundamental differences in their performance and properties of each of these types of antennas. So no longer only the price is the major decision factor when choosing between these two. Because of the interference conditions and the flood of all kinds of antennas to the market one should be really careful about what antenna technology to go with. In comparing horns and patch arrays their differences isn't in many if not most of the parameters you're looking at. So if you haven't seen the first part of this webinar series make sure you check it out. The part one about horns so that you have the complete picture of both technologies and you can find a recording of the first part about horns on our YouTube channel already. So this was a short intro and now we can move on to the details of the patch array antennas. Why are these antennas called patch arrays? So the core of this antenna is the printed circuit board with a number of patches stacked in the vertical direction and connected with the feeding lines which all originate at the coaxial connectors to which you connect the radio. So while the building unit is a single rectangular patch etched on the surface on the printed circuit board as soon as you stack two or more patches above each other and feed them with the same signal they form an antenna array. And this is valid regardless of the shape or size of the patches. Two or more patches makes the patch array or an antenna array. A single patch etched on the surface of a PCB laminate is a resonant type of an antenna. So the resonance is a phenomenon that occurs when the length of the patch is equal to half of the wavelength of the signal that is fed to it where by wavelength we mean the distance over which the feeding signal repeats its shape. So if this condition is fulfilled the antenna radiates the signal into free space. How does the resonant frequency change with the size of the patch? So as many things in the world of RF engineering the resonant frequency is inversely proportional to the size of the patch. As we can, as we increase the size of the patch the resonant frequency is decreasing and vice versa. When we decrease the size of the patch the resonant frequency is growing. So the graph is indicating where the resonant frequency FR is located. And the VSWR or voltage standing wave ratio is an antenna parameter that tells us how much of the input signal is reflected from the antenna connectors. So the smaller the VSWR is the less signal is reflected, which is what we want. So in other words, it is radiated by the antenna. So eventually you recognize the resonant frequency as a dip and the VSWR graph as you can see from this animation. It's not only the size of the patch that influences the properties of patch array antenna but also the substrate on which it is etched. So it has three main parameters that the patch antenna designers work with. So first is the substrate height or H. And this one influences the resonant frequency and the amount of unwanted radiation. Second is their permittivity epsilon R which is the property of the material of the substrate and tells us how strongly the substrate material influences the electric fields. And the third one is the loss tangent of the substrate. This one tells us how much power is dissipated in the material of the substrate which influences the resulting gain of an antenna. So naturally the higher the loss tangent the more signal is lost dissipated in the heat inside the substrate. This is how typical substrates look like. It's a thin sheet of a semi-rigid material with copper metallization on one or maybe even both surfaces. So there are many types of substrates with varying quality and properties with prices spanning quite a wide range. So naturally the very low loss and precise the electric constant substrates tend to be on the more expensive side. Patch antennas come in many different shapes. These different shapes with various cutouts as you can see from these images are another tool in the hands of the antenna designer. They all influence the bandwidth gain, manufacturer ability or price of the final product as well as other parameters like for example the stability of the radiation pattern and so on. So the intricacies of the patcher antenna design reside in knowing the effect of these modifications and be aware of the trade-offs that are always present. When we change one thing in a desired way well something else might go a little bit worse but it depends on the application what trade-offs are acceptable. The radiation pattern tells us how well an antenna radiates in any direction and single patcher antenna has wide radiation pattern in all directions as you can see. So sometimes you can also see the 2D version of the radiation pattern. It is important to understand that the 2D radiation pattern gives rather limited information about the antenna behavior because it shows only a single slice of the whole 3D image. Nevertheless it still depends of course on the application whether you should be interested in the full 3D radiation pattern or the 2D version gives enough information. And for wisps the 3D radiation pattern definitely makes more sense because the side lobes of the antennas are the root of the interference issues in unlicensed with networks. So the 3D radiation pattern will show you the whole information with all the side lobes you know so you have the full image and know if you want to avoid or use the antenna you're looking at. The reason why single patch antenna is not good for sector coverage is that its radiation pattern is wide and fixed and eventually its gain is also rather low. So there are the parameters that are fixed which is why the single patch is used in applications where a low gain and wide radiation angle are desirable which is definitely not the wisp networks. Another problem with the single patch antennas is their narrow band of operation which is connected to the narrow band nature of the resonance we mentioned earlier. So typically resulting to around half a gigahertz bandwidth in which would not be enough for unlicensed with networks operating in the five gigahertz frequency range but again in other types of applications it might be fine. To improve the shortcomings of a single patch antenna we can form an antenna array of two or more patches and this is the general principle you can apply with any radiating element stack two or more of them and you get an antenna array. Here we show a one dimensional patch array antenna and what happens with the radiation pattern as we keep adding more and more patches. So clearly the more patches in the array the higher the gain of the antenna will be and at the same time the radiation pattern is changing as well. You can see that the beam width is shrinking in the elevation plane. So when the patches are stacked in the vertical direction the beam width is shrinking in the elevation plane while in the azimuth the pattern keeps the shape of a single patch antenna. And also the side lows start to appear as we stack more and more patches which is the result of the physics of antenna arrays regardless what is the single or the building block of the radiating element. The array principles work in a similar way for two dimensional arrays. So adding more patches the main beam width is shrinking but now in both azimuth and elevation planes at the same time because we're adding more patches in horizontal and vertical directions. Same is valid for the side lobes. These images show results of simulation in an EM simulation software. So it's nicely visible how the side lobes progress as we increase the size of the array. And by now I'm sure you're getting the intuitive understanding of how the antenna arrays work. So if we for example stacked patches only along the horizontal direction the radiation pattern would be getting narrow in the azimuth plane and elevation would vice versa again remain the same. To align the line of thought with the first part of this webinar about horns you might have watched before we will start looking at the patch arrays in terms of their side lobes performance. There are two major causes of the side lobes patch arrays suffer from. So one of them as we mentioned earlier is the physics of the antenna arrays. To be a little bit more precise it's the interference of the waves that are radiated from each and every single patch or interference or in this case addition you can also call it an addition of the waves. So here it is nicely visible how the wave from two and more sources add up. In the areas where they add up constructively there's a maximum so you get a side lobe where they add up destructively there's a minimum and there's a dip in that radiation pattern. So in the symmetry axis of the array is located the main beam which is the strongest that's what we're interested in. Any other lobes outside that main beam are side lobes which in the RF engineering lingo they're called grating lobes if you will. And these are the side lobes a typical patch array has in the elevation plane. So when looking at the antenna from the side second major cause of the side lobes of patch arrays is parasitic radiation. And here we look at two components. First is the radiation of the feeding lines. So to bring the feeding signal to each patch in the array a network of feeding lines is used which is basically symmetrical division of the line until we get as many outputs as possible I mean or as needed as is the number of patches. In this case in this example you can see two. So at the top there is a vertical line and that's branching to two so we can feed our two patches. So the feeding line is composed of metal strips of varying width and shapes. And these have their own resonances that help on one hand extend their bandwidth of the whole antenna which is a good thing but also makes them radiate part of the energy before it reaches the patches which is a negative feature for a WISP sector antenna. Looking at the patch from the side so a lot more is happening there than we would actually like to. And main portion of the wave is radiated from the patch itself which is of course what we want. What portion of the energy is traveling inside the substrate as a so-called surface wave causing lateral radiation and diffraction at the edges of the substrate. And these parasitic sources of radiation cause additional side lobes in the azimuth plane of a typical patch array sector. In fact we can actually simulate how the feeding lines themselves radiate in the absence of the patches on the substrate. So here we can see the result of such simulation while the feeding lines are inevitable when designing a patch array. They introduce additional side lobes in the resulting antenna radiation pattern which is a very common thing with vast majority of patch arrays on the WISP market. One of the ways manufacturers try to deal with the azimuth side lobes are various shields and shrouds and that are attached to the back of the original antenna structure. Be it the aftermarket kits or the shielding kits provided by the manufacturer. So what is the effect of these shields on the radiation pattern of patch arrays? While the shield might help improve the front of the ratio a little bit the azimuth side lobes they aim to suppress are actually still strongly present as we can see from these near filled plots. On the left side is the generic patch array antenna when we're looking at it from the top. And a lot of radiation finds its way in the backward direction causing the side lobes that are harmful to WISP networks. And on the right side is the same antenna but with the shield which you can see as the two slanted lines at the edge of the antenna body. So the difference in the radiation is noticeable but altogether the desired effect of suppressing the side lobes is practically non-existent. The side lobes are merely rearranged slightly but are equally strong whether a shield is used or not. This slide shows the same comparison but the far-filled radiation patterns. So on the left is the patch array radiation pattern without the shield. And you can see the strong side lobes pointing backwards. On the right is the patch array radiation pattern with the shield. The change is noticeable. I won't be denying that but as you can see the strength of the side lobes didn't really change and some could actually even worse. So the conclusion for the shields is that they will really not help you mitigate the side lobes as intuitive as it sounds. One would think if I put that shield around it well surely it must be doing something but in this case this intuitive understanding does not apply. Like some things in the art of engineering or in physics in general are just weird. So some are intuitive, some are not and this one is definitely not but the conclusion here is that the shield unfortunately do not help to mitigate the side lobes. Here is another viewpoint on how the patch arrays perform with the shields. So the colorful images show the coverage pattern. In this case the MCS zones. So the dark blue for example says you can have the MCS rates of one or less or for example the green area says that the links can perform with the MCS rate from four to seven and so on. So instead of the typically expected oval shape of the coverage area you can see on the left one for the antenna without the shield you get this wild flower like looking area which is way different from what you expect when you use the shield. And this adds the other set of issues especially to the customers close to the edges of the sectors as you can imagine some of them will find themselves outside the coverage zone never before you install the shield and that will make their internet slow down. So is there another way how can we deal with this parasitic radiation and those side lobes? And the answer is yes. Here you see an example of a bit more complicated patch array structure that achieves the side lobe suppression. So this patch is circular and it's fed by a coaxial cable through the substrate as you can see on the side view. And there is an air core underneath the patch actually. So this air core is manufactured by milling of the substrate and has a cylindrical shape as is obvious from these images. And the presence of the air core effectively mitigates the surface wave we mentioned earlier and which would otherwise result into those azimuthal side lobes. On top of the air core to suppress the surface wave there are metal fences between the patches which furthest scatter the waves which would otherwise result into patch coupling between I mean the coupling between the patches. So on the right side you can see the resulting radiation pattern of this antenna. And the azimuthal and elevation cuts of the radiation pattern of a 16 element array clearly indicate the side lobe level below minus 20 dB which is a really good result for a patch array antenna. This amazing side lobe performance comes at a price though. So on top of the more complex structure to manufacture the non-resonant nature of the feeding structure and the presence of the air core make this antenna rather narrow band which is not a problem of course depending on the application. But again, for Wisp networks half a gigahertz bandwidth is simply not enough. Another factor preventing widespread adoption of solutions such as this one in the Wisp industry is actually the price, the end price. So this is not the simplest structure to manufacture which melts down to increasing the cost to the level not many wisps would be willing to pay. At RF elements, we came with our own solution to the parasitic radiation in the azimuth plane. We called it the back shield and it is a metal profile on which the antenna is attached. So the shape of the back shield with all the little notches and additional protrusions turns it into so-called frequency selective surface which effectively deals the parasitic radiation with the parasitic radiation in the azimuth plane. Comparing the far field radiation patterns with on the right side and without the back shield on the left, you can see a nice suppression of the lateral radiation of the side lobes which results in back radiation of rounder shape of the radiation pattern altogether forcing the field pointing forward rather than backward. Looking at the same radiation pattern from the side though we can see not much has changed really. Indeed the direction and the size of the side lobes has changed but that's it really. So unfortunately not even the back shield can deal with the elevation side lobes which are again an inherent property of any antenna array. The grading lobes we talked about a few slides back. Beam efficiency quantifies side lobes. So instead of just saying an antenna has or doesn't have side lobes, beam efficiency provides a numerical measure, a figure of side lobes that makes it super easy to compare antennas in terms of side lobe performance. Beam efficiency is the ratio of the energy contained in the main lobe to the total energy an antenna radiates. It tells us what part of the radiated energy is going into the main lobe. So the higher the beam efficiency is the more energy is in the main lobe or in other words where we want it to be unless everywhere else or in side lobes. So beam efficiency can be 100% at most and that's the maximum value the beam efficiency can have. And this means that all the energy the antenna radiates is in the main lobe meaning that it literally has zero side lobes but the smaller the beam efficiency is the more side lobes an antenna has. So for example, if beam efficiency is 35%, the remaining 65% of the energy the antenna radiates is in the side lobes which makes up a very poor antenna in terms of side lobe performance and noise suppression. So to give you a particular example, this is the radiation pattern of a generic battery antenna. So if it's beam efficiency is 58%. The 58% of the energy antenna radiates is in the main lobe where we want it. And the remaining 42% therefore must be in the side lobes which is completely undesired. And note that all the side lobes are highlighted. So beam efficiency includes all the side lobes of an antenna not just one or a slice of the radiation pattern but the whole package, the full 3D data unlike the other noise suppression measures you might know such as front to back ratio side lobe level or Etsy masks. The vast majority of antennas used for sectorial coverage in WIST networks are either patch arrays or horns. So reflecting the issues the patch arrays have for the side lobes meaning their size and the frequency dependence. The beam efficiency of patch arrays is around 60%. Depending on the manufacturing and design quality. So now you can directly compare how much side lobes patch arrays and horns have. With patch arrays around 40% of the energy they radiate is in the side lobes. On the contrary to horns, horn sectors only have around 5% of the energy going into side lobes. See, as easy as that. You can actually also see other horns in this graph as well. And this is to highlight that it takes a considerable effort to design a horn antenna so that its beam efficiency is high. So the stable and zero side lobe performance is not a given. It's not a default when you have a horn. And if you wanna know more about beam efficiency we have a whole webinar about this topic which we did some time ago. And you can find the recording of that webinar on our YouTube channel. The balance between the horizontal and vertical chains is important for link reliability. So if both chains perform identically, switching between them makes no difference whatsoever to the performance you see. The gain of the patch arrays is stable in a very narrow band of frequencies because of the very narrow band nature of the resonances of the patches and the feeding lines as we explained earlier. So we see that even between the five and six gigahertz the gain is changing substantially which eventually reflects into instability of the coverage provided by this antenna. On top of that, there is also a mismatch between the horizontal and vertical gains. You see that there, which means that the curves now on the left graph are not overlapping which when looking at the radiation pattern reflects as a different coverage pattern for both polarizations. So in the end, your customers will be the ones to perceive these discrepancies as unstable connection speed. And there is more issues with the patch arrays we could talk about but I'm only going to talk a little bit about the frequency dependence. So we already know the patch arrays have many frequency dependent side loops that collect and transmit unnecessary noise increasing the noise floor your radio is working with and so on and so on. But not only the side loops but also the main beam and the main beam width changes with frequency which is yet another addition to the instability of the coverage provided by patch array sectors. So see on the animation how the side loops are pulsating and regrouping and so on. This is exactly this feature that makes the patch arrays so unreliable and random switching between the channels often changes the link performance quite significantly. So patch arrays are unstable in the whole unlicensed spectrum wisps use. Here we show the frequency instability of patch array how the instability of the patch arrays reflects on the actual coverage area. So seeing this animation you might not be surprised because you probably experienced it in practice where by switching the channels hoping to use a cleaner bit off the spectrum the results seem even worse than before leaving you kind of scratching your head. And this is exactly what happens when you are changing the channels. The radiation pattern is changing a lot with frequency which is causing the fluctuations and ultimately unreliability of your network and your life as a wisp, you know a person who's running a business really constantly busy servicing the links that change whenever you switch channels. And mind you, these are not just some artistic impressions we made up in Photoshop. These are, you know, this animation is a result of real physics based simulation when we imported the 3D radiation pattern off this antenna into our program and run the simulation to see how the coverage changes with frequency. So these are as close to reality as it gets. And here are a short note on the mounting of the patch erase. They're usually composed, the parts are composed of many, many, many smaller parts. And, you know, you need to screw 12 screws to finish the installation. So this makes especially the collocation of these antennas on a small space rather challenging logistical task. So going to the strengths of the patch erase. So the number one strength of patch erase is their gain. It is super easy to keep increasing the gain with the number of patches in the array. An example of variation pattern of a typical patch array has a main beam which is wide in the azimuth plane and, you know, which is good for the angular coverage. Despite that you can theoretically grow the gain of patch erase indefinitely. The real life limitation is given by the beam width. So the elevation beam width is increasingly narrower with growing gain. So in order to preserve decent sectorial coverage you cannot increase the gain of patch erase indefinitely. So as the number of patches in the array increases the gain grows, which is desirable but so does the elevation beam width decrease. So the narrow beam width in the elevation causes increase of the null zone. And this is covered by the, increasingly covered by the side lobes you can see in this animation. So as illustrated here, this is how the coverage looks like in the case of high gain patch array antenna. The area near the tower is covered by the side lobes if at all and these side lobes change with frequency meaning you can't really provide a stable coverage to the customers near the tower. Now this zone of uncertainty coverage increases with the gain of the patch erase. So as we keep increasing the gain of these antennas these issues get worse and worse. Another strength of the patch erase is their cost. Simply said they're cheaper to manufacture. The development of the PCB technology started at the beginning of the 20th century by a German inventor, Albert Hansen. And actually the first experiments were done by Thomas Edison. Since then it has gone quite a long way and today it is one of the cheapest technologies for circuit antenna manufacturing at low frequencies partially also thanks to the fact that virtually all other electronics are based on this technology as well which has put a tremendous pressure and effort into inventing cheap and really sustainable ways of manufacturing these PCBs. So thanks to the well-developed PCB manufacturing technology scaling of the patch array antennas for higher gain is not very difficult nor expensive. So the surface area will increase but the thickness of the 30 millimeters in this case in this example will remain the same. So the added expenses will correspond to the cost of the PCB manufacturing which is one of the really the biggest advantages of patch array antennas. For example, to scale the patch array from 18 to 24 DBI gain all you need to do is to increase the area of the antenna five times naturally including the enclosure of course which will not add much cost in the end. So while patch arrays undeniably have the advantage of easily growing the gain and simple manufacturing resulting into attractive pricing from the point of view of what is needed in WISP networks there are many cons. So the present state of the industry is that for majority of WISPs claiming their biggest issue is the noise or interference. And this shows the side lobes and other cons of the patch arrays are the issue that should be mitigating. Also the thinking that higher gain equals better performance kind of falls apart in the face of those high noise levels. One should really use an adequate gain antenna for the given job otherwise you're hurting yourself and actually also others at the same time. So for those of you who participated in the previous webinar about horns we prepared a short summary table comparing both types of antennas in the light of the most important antenna parameters in WISP networks. Here you can see the final comparison table. So in terms of gain the patch arrays definitely have upper hand compared to horns because the higher gain comes easily but horns are definitely not far behind. In terms of the gain stability, horns are impeccable. The patch arrays unfortunately because of all the physics we just talked about suffer from the instability of their gain. In terms of side lobes, same thing. The horns are definitely winning this one. But as we've shown the patch arrays can be also optimized or can the side lobes of the patch arrays at least to a degree can be dealt with but the price is simply way too high to be implemented in the WISP industry. In terms of bandwidth, horns are really wide band covering pretty much the whole spectrum unlicensed spectrum the WISPs work with. And unfortunately the patch arrays again due to the resonant nature of their functioning is the bandwidth is not so wide. In terms of the pattern stability, horns again are winning this point simply because yeah, again, I'll be repeating myself. The physics just is unrelenting. In terms of the balance of the performance of the horizontal vertical chains patch arrays are doing, let's say, okay but horns are definitely way better. Switching the channels, there's nothing to the performance when using horn antennas. And in terms of manufacturing patch arrays are really cheap which makes them very affordable. And it's of course understandable for the WISPs that are just starting out and cannot afford to go with the better technologies. But if you have at least a little space to move away from them, it's a good idea simply because I mean, despite that horns cost a little bit more in the long run this upgrade or change of technology is definitely worth it. So we at our developments, we like to pride ourselves in setting new industry standards. So we're helping the WISPs to deal with the most pressing problem of the interference, self-interference and noise by through our toolset of the horn antennas that are very efficient and effective at rejecting the noise. So our choice port environment enables super easy connection of the radios you might be using while actually also delivering almost 100% of the power from the radio to the antenna which cannot be said about the coaxial cables. And all together, we like to call it a toolset of horns provides you the options to scale your networks indefinitely so you can, there is really no limit to how many horns you can put up on a single side except obviously the physical limit of the space I mean, adding the horns will not degrade the performance of the sectors you already have in place. So that's pretty much the gist of the massive scalability which allows your business as a WISP to grow. I would also like to invite you to check our YouTube channel and check the WISP traveler playlist where WISPs like yourselves talk about their own experience with our horns and how they helped them to deal with the problems they were facing. We also have another playlist called Inside Wireless on our YouTube channel and this playlist contains quite a few short around three minute long videos about all kinds of things from the world of RF engineering. So whether you still have some questions from the RF engineering world or just want to refresh what you already know I encourage you to check this channel out. It is definitely helpful. And we also have a online discussion forum where you can search through the questions our customers asked about our products or register and ask your own questions regarding whatever interests you about our products. So thank you for your attention and I wish you have a nice rest of the day and the week. Bye bye.