New technology for building cellular networks

As you know, from time to time we do translations for you of interesting publications on new products and / or various discussion issues in the telecom industry. This time we bring to your attention a translation of an article about a promising technology that can significantly increase the speed of wireless Internet in existing cellular networks.

In mid-February of this year, American startup Artemis Networks announced plans to commercialize its own pCell technology. It is based on a new wireless communication scheme, which should save us from congestion in cellular networks and provide faster and more reliable data transmission. And what’s most interesting, supposedly you don’t even have to change our smartphones for this, pCell works on existing 4G devices.

If the scalability of this technology is proven, pCell can radically change the operation of wireless networks, replacing existing cellular networks with new ones built on a completely different principle. They combine the signal from many distributed antennas, which allows you to create a unique cocoon of stable communication around each subscriber device. In such a cocoon, the entire spectrum width of this network will be available.

PCell technology was first introduced in 2011 under the name DIDO(distributed input, distributed output). At that time, the idea looked like pure fantasy. And there is no doubt that Artemis will have to reasonably answer to numerous criticisms and doubts before major operators decide to pay attention to them. But there are several reasons why their technology can really be implemented.

Firstly, this is an elegant solution to the general problem of annual traffic doubling , which operators solve as they can . According to Steve Perlman, CEO of Artemis, the lack of spectrum width limits the ability of operators to innovate. The reason is not a lack of ideas, in recent years a variety of solutions to the problem of traffic growth have been proposed. For example, small cells , interference coordination(interference coordination), as well as the use of the millimeter range, which we already wrote about ( part 1 , part 2 ). But Perlman believes that all of these measures are crutches for a morally obsolete cellular communications scheme. He argues that the main problem lies in the very principle of dividing the network into cells.

The main problem of cells is signal interference. Base stations and subscriber units must precisely coordinate power and transmission spectrum so as not to interfere with each other's signals. This division of the available bandwidth of the spectrum between users was incorporated into the technology of cellular communications in the 1980s. And as a result, the throughput of each session decreases very much if many subscribers are connected to the base station.

Artemis decided to abandon this dynamic division of the range; instead, each subscriber unit can use the entire available network band, regardless of the number of other users in the district. As if your smartphone is constantly connected to your personal base station. Hence the name pCell - personal cell, personal cell.

How exactly does this technology work? To deploy such a network, the operator needs to create a cloud data center that will take on all the main load of heavy computing for the system. Then it is necessary to install antennas in places where subscribers are crowded: in residential and office buildings, shopping centers, on the streets. Although the antenna modules (the size of a hat box) look like small honeycombs, they are not. These are not base stations, but rather simple devices that serve to redirect and decrypt signals. Each module connected to the data center can be installed anywhere.



Now suppose your smartphone is trying to connect to the pCell network. To do this, a traditional request is sent, which is received by all antennas nearby (say, 10), which report this to the data center. Let's say you watch a YouTube video on your smartphone. The data center requests it from Google servers, and then the video is transmitted to your device through 10 antennas that have accepted your connection request. Moreover, not one of them will broadcast the entire stream or even part of it. Instead, the data center uses the location of the antennas relative to the subscriber unit and channel parameters, such as multipath and attenuation, to generate 10 unique wave signals emitted by these antennas. Being meaningless on their own, together these 10 waves form the desired signal.

While you are moving in space, while other devices are disconnected from the network, the data center constantly recalculates the wave parameters for each antenna so that all subscriber devices receive the correct resulting signals. No competition for the resources of the base station, you can literally cover the entire city in the entire available spectrum band.

On hand pCell plays compatibility with existing 4G-devices. This is done using software emulation of LTE stations, and the data center can use these virtual base stations to establish a connection with unsuspecting smartphones and tablets. In addition, each gadget will consider that at the moment it is the only one connected to the BS. The same can be done on 3G and Wi-Fi networks.

Will pCell operators adopt? It is unlikely that any of them will soon want to abandon the created LTE networks, even if the technology from Artemis really turns out to be so miraculous. But compatibility with LTE is a trump card pCell. For example, operators can deploy antennas in places where base stations are most often congested: at airports, stadiums, and city centers. That is, where funds have already been invested in infrastructure. And users can seamlessly switch between networks without having to buy new smartphones.

Artemis wants to license its technology for use by mobile operators and Internet providers. Large-scale field trials are also beginning in San Francisco, with technology likely to reach commercial availability by the end of this year.



Of course, this technology cannot be called a contender for the title of 5G. We are talking about another architecture - a self-optimizing cellular network. We are traditionally interested in the opinion of experts from the telecommunications industry: what do you think are the prospects for integrating the described technology into existing LTE networks without serious additional investments? How do you assess the feasibility of such a solution?