Sharing The Unlicensed Spectrum
By Peter Flynn, business development manager, Communications Processors, TI
The shared access model of the unlicensed Industrial, Scientific and Medical (ISM) wireless spectrum bands has stood the communications industry in good stead for many years. However, it’s no surprise that additional bandwidth is desperately needed to meet the mounting tide of user demand. It’s time to explore the unlicensed spectrum by tapping into some of those restricted bands.
Simply stated, the current ISM bands cannot meet the market’s appetite for more data and mobility. Currently, 12 ISM frequency bands are defined by ITU, providing a theoretical total of 4.0GHz of bandwidth. WiFi (802.11xx) suffers from the greatest congestion, but only occupies three bands (.9GHz, 2.4GHz and 5.7GHz) and is limited to 276MHz of bandwidth.
A significant and often overlooked portion of the ISM bands is actually consumed by the cellular service providers who offload traffic from their licensed bands to unlicensed ISM bands via WiFi access points. A study done by Cisco found that 45 percent of the global smartphone traffic traverses the ISM shared spectrum bands¹. And global mobile phone data traffic is projected to exceed 15 Exabyte per month by 2018! Add to this the expected 50 billion Internet-of-things devices² projected to pop into the ISM bands in the next few years and there’s little wonder where that once adequate ISM bandwidth has gone.
Give and take
The overly simplistic answer would be to expand the unlicensed ISM bands, but this is not feasible. First, there is no unassigned spectrum to simply add to the ISM list. And second, re-assigning or moving incumbent users around is too expensive and time-consuming.
A more realistic solution involves leveraging underutilized spectrum. It turns out that plenty of spectrum is available. A study by the Illinois Institute of Technology³ puts total spectrum utilization in urban areas like Chicago at less than 15 percent!
So the best option, and the one that is being aggressively pursued by the Federal Communications Commission (FCC), is to share underutilized spectrum without interfering with current users. In addition, spectrum must be shared more efficiently and managed proactively as a valuable resource. These last objectives will require some higher-order management techniques that go beyond merely avoiding collisions, which is the basis of WiFi’s Carrier Sense Multiple Access (CSMA) protocol.
In fact, more robust wireless data models like LTE and WiMAX have already been developed and are in deployment today. These are generally referred to as 4G (4th generation) protocols which employ Time Division Duplex (TDD) techniques to support multiple users simultaneously without interference.
More recently, the FCC has promoted spectrum-sharing solutions based on a database approach. The International Engineering Task Force’s Protocol to Access White Spaces (PAWS) database currently provides this function for the television broadcast bands. A very similar solution, Spectrum Access System4 (SAS), has been proposed by the FCC for the new 3.5GHz unlicensed band.
Basically, these models query a database to find unused spectrum. If bandwidth is available, the Access Point (AP) notifies the database which band it intends to use and then directs users to that band. This general model is followed by cognitive radio devices which can use many different bands, dynamically switching between them. The FCC has applied a three-tiered database approach to the 3.5GHz band. The potential to use this sharing solution in other bands is already being considered.
Using the underutilized
Clearly, technology is evolving to better share underutilized spectrum. In fact, the President’s Council of Advisors on Science and Technology (PCAST) in 2012 issued a report which identified many federal bands that could be shared.5
Figure 1 – Federal and shared bands under investigation for shareduse.
The incumbent users in these bands, mostly federal agencies but also Doppler radar and satellite communications, will require assurances that they will not experience significant interference from unlicensed use. Managing this new spectrum with diverse requirements will not be trivial, but it can be accomplished by applying consistent yet flexible models.
One model that was proposed by PCAST and is gaining momentum places the value decision with consumers. They decide the value of the bandwidth at any given time and this will, in turn, ultimately drive the deployment of new capacity appropriate to that consumer value. The key point is that nothing is free when it comes to spectrum. There must be an economic model driving infrastructure deployments that are based on the value established by the user community.
Challenging OEMs and technology vendors
Dynamically accessing many bands, accommodating various management policies and adapting mobile device development strategies to the available spectrum in a certain location can be quite challenging.
Access points and base stations will need to be very flexible and able to dynamically select available bands quickly and efficiently. Of course, developing band-adaptive mobile devices would be pointless until a clear plan for deploying wireless infrastructure has emerged. The initial challenge will be to bring APs to market which can leverage the capacity of underutilized bands in the wireless backhaul infrastructure.
Devices which are able to cooperatively negotiate for access across a wide range of bands and quickly adapt to the optimal band available are referred to as cognitive radio devices. Developing this high degree of flexibility can be challenging and it invariably leads to software defined radios because of their inherent flexibility.
There is technology available today which can implement all the latest protocols and cognitive radio functionality to achieve the goal of shared unlicensed spectrum. For example, TIPot multi-band long range gateway supports both 2.4Ghz and 780-900Mhz and we also provide NB-IOT development kit by using SDR technology.
All stakeholders need to come together to achieve success in this endeavor, but it won’t happen overnight. The feasibility of sharing underutilized spectrum should first be demonstrated in several bands. Then the marketplace will drive further expansion. Shared unlicensed spectrum will have a huge impact on broadband data services around the world. The technology exists to solve the problem. Now, global policy consensus must be built.