I was giving an update in our weekly staff meeting about our upcoming cnPilot E500 WiFi Access Point (AP) release, and somebody asked me why we are performing better than most products in the market at the given price point. As I answered, it occurred to me that in today’s world of marketing, when looking for reasons why a product is successful, most of us look for a buzzword, a tag line, or an obscure technology.
Take the IPhone®—there was no tagline, no new technology—the success lay in using the existing technologies in the right way. Great products are rarely an outcome of one magic bullet; rather they are an outcome of using the existing technologies in the right way and improving the user experience at the right price.
There are a number of good outdoor WiFi access points in the market that provide acceptable range and throughput. But they use high-gain external antennas that are large in size and high in cost. There is also a set of WiFi APs that have a smaller form factor in the $300–400 price range, but this group lacks the range, capacity, and ruggedness needed for an outdoor deployment. Service providers need a product that can be reached by cell phones from 200 meters away, can serve video to hundreds of end devices, is small and unobtrusive, and is ruggedized, at a sub-$400 price point.
When designing an AP, we chose an acceptable compromise between a few key factors: transmit (TX) power, antenna, the number of input/output chains, the radio’s sensitivity, and the packet processing in software. Most of the good quality Systems on a Chip (SoC) in the market today have very similar sensitivity and improving on them would add significant cost. The software has to be lean and mean with no packet copy. With those two points said, let’s focus on TX power, antenna, and the number of chains.
As the TX power is raised, the range increases. But that increase is only in the downstream direction. It has no impact on the upstream range. The upstream range is decided by the TX power and antenna gain of wireless client devices, which are much lower than that of an AP. So, even a very high TX power AP will not be accessible by a client beyond a certain distance. Antenna design comes into play here. An antenna focuses the energy in a direction or in a plane. There is an advantage of antenna gain over TX power—the gain of antenna is symmetric. So it enhances the potential to receive the signals from a distance along with being able to send a signal to a farther distance. Typically, in outdoor deployments, the antenna needs to be Omni-directional to transmit and receive in all directions. The general goal is to try to focus the energy more in the horizontal plane—in something like a donut shape. We also try to reduce the energy going above the horizon. But the more powerful the antenna is, the larger in size, and the more expensive it becomes. So this requires good engineering judgement on antenna gain and TX power. Also, there is typically more downstream traffic than upstream. So the data rates need to be higher in downstream than upstream, and the total gain in downstream direction should be more than the upstream. This is another factor that needs to be taken into account while balancing TX power and antenna gain.
Having a larger number of input/output chains gives higher performance. Most of the wireless devices that are installed in public hotspots (cell phones, tablets, ultra-portables) have only 1×1 or 2×2 chains. So the APs with more than 2×2 chains have only marginal advantage (mostly coming from the diversity) and at significant addition of cost. Also, if the attached antenna to the chains are not physically separated, this advantage can be limited. Keeping aesthetics in mind, it is impossible to keep a large number of antennas separated. So a 2×2 chain is a logical configuration. The orientation of mobile devices is never stationary, so keeping one chain vertically polarized and another horizontally polarized maximizes the chances of good performance irrespective of the orientation of the wireless device. This is exactly what we did in cnPilot E500.
Beamforming and smart antennas also play a significant role. Beamforming directs the antenna pattern electronically. Some of this is done in digital signal processing (DSP), and most WiFi SoCs in the market are capable of performing this function. It has limited gain in smaller number of chains; the real advantage appears in APs with more than 4×4 chains, which are typically much higher in cost and with an antenna array significantly larger in size. Some of the WiFi APs have used a large number of array elements, and by altering the delays and power through each element, one is able to steer the beam very effectively. But there is a catch. One needs significant space to fit the large number of antenna elements required. If the antenna elements are arranged too close together, though the beam steering will yield a few dB gain, the total gain of the antenna will be smaller. Beamforming with smart antenna only makes sense for tower-mounted, high-altitude APs, which can be large in size. Thus, even though we have used the beamforming extensively in our tower mountable PMP APs, we did not use it in cnPilot E500.
For reliability in an outdoor environment, a ruggedized IP67 grade housing is vitally important. The largest part of the cost in this ruggedization is associated with the glands for each of the antenna connectors and enclosure. BecausecnPilot E500 is equipped with an internal antenna, the AP has a more regular shape instead of the protruding antennas that are typically seen.
The high-performance of the cnPilot E500 is in the innovative application of existing technologies: the antenna, TX power, and number of chains. Like a great work of art, the innovation is not in creating a new color, rather it lies in carefully, imaginatively combining the colors to create a masterpiece.