Private data: future networks today

CSIRO technology promises higher-performance radio networks without extra spectrum.

Public wireless networks built by the major telcos enable us to use our mobile phones and tablets when we’re out and about. These networks cost-effectively provide the entire population with basic access to voice and data, and increasingly, video-based services. Coverage, cost and consistency of performance drivers are on a ‘best endeavours’ basis, but are not mandates.

Private networks, by contrast, have a different set of drivers and performance objectives - rapid deployment, security, ability to deal with severe environmental conditions and life-threatening situations, demanding remote and regional performance, and 99.99% availability and reliability … especially in crisis situations that can occur literally anywhere. Deployments are typically for emergency services, and availability of radio spectrum is often limited and/or fragmented.

So, public networks are deployed by private companies and private networks are deployed by the public service.

The underlying wireless technologies, both public and private, have now evolved to be all internet protocol (IP) based. This has fostered a view held by some that convergence around IP will see the scale of public networks dominate to defeat the private networks. So why isn’t it happening?

Private networks have evolved from delivering voice to emergency services, to adding data and (more recently) data-intense applications such as video delivery. From its voice-only origins, spectrum allocation has been based on either 25 or 12.5 kHz bands typically in the sub 1 GHz bands, ie, VHF, UHF and 700/800 MHz.

The public networks have also evolved from mass-market voice to now being dominated by data and video, occupying from 5 to 40 MHz of contiguous spectrum in bands from about 850 MHz up to about 2.6 GHz.

Private vs public

The spectrum difference between public and private is what begins to answer the question, “why is public not replacing private?” But there are many other factors too, and they can be conceptualised by realising that the two network environments are evolving with very similar underlying radio technologies, but the needs, demands and expectations are different enough to drive different solutions into different markets.

Public networks are becoming all IP and are optimised for very high-density traffic with smaller and smaller cell sizes to accommodate the very large number of devices, connections and bandwidth required to serve the highly connected consumers in a huge IP pool. Furthermore, the revenue generated can support the funding of forever-reducing cell sizes that enable this traffic and device growth.

Private networks are evolving to solve a different problem set: security, spectral efficiency, inconsistent spectrum availability, coverage during intense random events, service availability during a crisis (exactly when the public networks congest) and rapid deployment, often using multiple technologies, while ensuring IP direct connection is the most efficient it can be.

Terminal devices are another point of necessary difference. While public networks support the wonderful range of smartphones and tablets, private networks must support vehicle-installed devices that are designed for the very specific needs of police, ambulance, fire and other public safety situations. These purpose-built devices can leverage components developed for the consumer products, but they will remain the domain of specialist solutions far removed from the fashion pressures of the consumer market.

Spectral efficiency

RF Technology has partnered with the CSIRO to deliver a unique solution to these issues. CSIRO’s wireless researchers have been working for the past few years to develop new software and hardware that can help improve the efficiency, data rate and distance of wireless technology. The result is Ngara, a state-of-the-art technology offering the opportunity to deliver scalable solutions from a single narrowband channel to a hundred channels.

One of Ngara’s unique features is its ‘beam forming’ approach, concentrating signals through the air between antennas rather than spreading the signal over a large area. This means, for example, that the signal requires much less spectrum to send the same amount of data.

“This radio technology leads the way with a significant improvement in the state of the art in spectral efficiency and a software-defined radio architecture providing a flexible and scalable solution,” said Dr Mark Hedley, CSIRO’s research director for the Wireless and Networks Program. “Solutions can now target the performance and features needed for the emergency services market much more effectively than the more generic scalable and coverage-driven LTE approach.”

How Ngara works. Courtesy CSIRO.

RF Technology and its US-based subsidiary IP Mobilenet are developing a new generation of emergency services radio products leveraging Ngara. While public carrier infrastructure is meeting the demands of very high-density, large user numbers and extended coverage with large numbers of smaller cells, the private networks enabled by RF Technology and IP Mobilenet will offer a more scalable and flexible solution set with the following features:

• 30 to 900 MHz bands with a single radio
• high spectral efficiency
• software channel selection and allocation
• software-defined beam forming
• ability to leverage non-contiguous spectrum
• dynamic performance of narrowband and broadband capacity
• spectrum on demand
• all IP-based communications
• simulcast
• numerous simultaneous users without degradation of bandwidth
• the ability to work in narrow band

With all these new areas of flexibility it is increasingly possible to meet the ever-more-demanding requirements of the emergency communications sector.

One example of this new flexibility can result in ‘adaptive’ video delivery to/from police vehicles. Spectral efficiency improves as vehicles slow. Under software control, this characteristic can be managed in different ways depending on the specific user needs. Bandwidth can be adjusted to a maximum while spectrum use can be held constant. Or, bandwidth can be maintained by varying the number of channels used to deliver it. Depending on the application, video quality can improve as the vehicle slows or stops, while lower quality can be delivered at higher speeds. With modern video compression techniques, this can enable a very high-performance and flexible solution to suit many demanding and increasingly important video scenarios.

Indeed many of the emergency responder market’s needs are quite different from public carrier LTE needs. While public networks evolve to handle the generic public need for more users per cell and more bandwidth per user, the private networks are striving for more spectral efficiency, flexible channel allocation, spectral fragmentation and more rapid responder support into targeted areas with increased bandwidth with limited spectrum.

Both of these sets of market drivers are being supported by all-IP networks, and both market segments are driving forward the performance of radio solutions. The more tailored approach to the special needs of the emergency responders is driving suppliers such as RF Technology and IP Mobilenet to use the software-controlled nature of these networks to their advantage in satisfying customer requirements.

Non-contiguous spectrum

Using Ngara technology, individual narrowband channels can be combined to leverage whatever spectrum is available into the bandwidth required. If, for example, a 25 kHz channel of spectrum can deliver 128 Kbps then two channels can deliver 256 Kbps. But the added flexibility comes because the two channels do not need to be adjacent - any number of channels can be combined to provide the bandwidth required. In an uncontended network with appropriate IP protocols, we would expect to see throughput equal to or better than traditional public networks with a greater degree of security and control.

Furthermore, these channels can be dynamically allocated under software control to allow the system to work around unavailable or noisy channels for example.

A prototype Ngara antenna. Courtesy CSIRO.

Ngara versus LTE

Ngara and LTE can be considered highly complementary in many ways; however, it’s also interesting to consider the core difference. A typical LTE antenna array would be configured with two 20 MHz channels, which results in an ‘ABAB’ pattern shown. This approach is often used to minimise the poor radio performance at the boundaries if the same spectrum is re-used in adjacent sectors. This results in 4 x 150 Mbps of available bandwidth using 2 x 20 MHz of spectrum.

But with Ngara’s beam forming model, a single 20 MHz of spectrum can be used with individual beams transmitted to each remote device. For example, 12 beams could all re-use the same spectrum and deliver 150 Mbps bandwidth, giving 12 x 150 Mbps of available bandwidth, and this is the case even if two adjacent beams are only 3° separated. This results in a spectral efficiency six times that of the LTE example, which would be of great benefit for agencies with limited contiguous spectrum or fewer channels.

It is envisaged that the LTE approach and the Ngara beam forming approach will be highly compatible technologies. The beam forming approach is also an excellent backhaul solution as well as a multipoint mobile solution. An added benefit is that since the beam is software controlled, there is no need for complex, time-consuming and expensive antenna alignment - the software can automatically and dynamically adjust the beam between transmitter and receiver to maximise performance, which means that lower-cost towers can be used and deployed very quickly. This would be especially valuable for special deployments for both planned and emergency events.

While public networks continue to address the needs of the mass market, private networks will play a critical role in addressing the very specific needs of the emergency services, leveraging advanced technologies such as Ngara.