How Better RF Sensing Can Improve Spectrum Allocation on the CBRS Band and Enable Dynamic Spectrum Sharing

Shared spectrum arrangements like the Citizens Broadband Radio Service in the United States have been a catalyst for private cellular networks across different enterprise verticals. At the same time, there is an ongoing debate between Environmental Sensing Capability (ESC) and Incumbent-Informing Capability (IIC), mainly acted out by Google and Federated Wireless: ESC is a mechanism in the CBRS that relies on sensors to detect the presence of incumbent users and dynamically adjusts spectrum allocation to protect them. IIC is an alternative approach that uses real-time coordination among CBRS devices to avoid interference with incumbents without relying on dedicated sensors. ESC relies on external sensing, while IIC uses internal coordination for spectrum sharing in the CBRS band. The fact that this fundamental debate around spectrum allocation in the CBRS is still on the table shows that even the nature of the CBRS framework is not set in stone.

At the same time, increasing demand for enterprise connectivity will increase demand for mobile network spectrum. In other words, mobile network spectrum will remain a scarce resource and careful spectrum management and dynamic spectrum sharing will be needed to satisfy this growing demand.

Spectrum and interference management in the CBRS is currently a coarse process that takes days to complete and allow new devices in the network. Moreover, networks with a General Authorized Access (GAA) license cannot guarantee the uninterrupted and interference-free operation of a network, meaning that mission-critical use cases using Ultra-Reliable Low Latency Communication (URLLC) and deterministic networking are not probable, and perhaps even possible in CBRS GAA spectrum. It is, thus, necessary to adopt a real-time spectrum management approach, which will allow more granular and detailed interference management. There are several illustrations of this:

• As the propagation model in the CBRS framework uses an Irregular Terrain Model (ITM), interference levels are overestimated, meaning that network planning and dimensioning typically lead to higher deployment and equipment costs than necessary.
• Interference calculations and planning take place over a 24-hour period, meaning that potential interferers can remain unidentified, even days after their appearance. Therefore, it is not possible to guarantee mission-critical use cases due to the lack of real-time understanding of the radio channel.

A new device appearing in the network asking for access can be permitted to transmit the following day and only after Spectrum Access System (SAS) providers have calculated potential interference levels. This means that new devices can only be permitted on the network after a day at best and SAS providers are required to perform complex, intensive calculations daily.

Because of the shortcomings of static spectrum arrangement regimes like the CBRS, enterprises wishing to deploy a private cellular network have to reduce the transmission power of their cellular networking infrastructure. This, in turn, leads to less capacity and a smaller coverage area for enterprise networks.

Ultimately, this will lead to more expensive network deployments, as infrastructure will need to be deployed a lot more densely because of the reduced coverage area.

The discussion outlined above illustrates that better spectrum management needs to take place for the CBRS and shared spectrum in general. Against this backdrop, there are two different approaches being discussed in the industry today:

1. The 3rd Generation Partnership Project’s (3GPP) approach to Dynamic Spectrum Sharing (DSS) allows Long Term Evolution (LTE) (4G) and New Radio (NR) (5G) radio networks to use the same spectrum and coexist without interfering with one another. This is deployed in the market, but inflicts significant overheads and is used only where necessary.
2. New telco software vendors come up with their own solution, which relies on extensive Radio Frequency (RF) sensing and analysis to improve spectrum allocation. This combines comprehensive RF capture, performs analysis of these data at the edge, extracts insights, and optimizes processes to improve the RF environment and other parts of the network.

Such a new spectrum sharing solution measures the channel in near-real-time and reconfigures the communication link to address potential interference and other mitigations. For example, it can detect communication anomalies and interference, factor in complete environmental data for the deployment, and consider customer rules and goals to assist in decision-making for the communication link. In addition, it can combine near-real-time data collected from the RF environment with historical data to predict how the channel will behave and perform near-real-time optimization. Furthermore, near-real-time RF awareness with rich data that include time, frequency, spatial, and signal characteristics can improve CBRS network planning in many ways. This can reduce infrastructure Capital Expenditure (CAPEX).

This means two things for the telco industry. First, static spectrum sharing arrangements like CBRS can become a lot more interesting through better spectrum management and interference prevention. Second, it can enable true DSS and, therefore, increase the efficient use of mobile network spectrum significantly if it gains enough traction in the market. To do so, it needs to be tightly integrated into existing telco infrastructure equipment.

• By integrating this new RF sensing, SAS providers can augment and improve their value proposition in many ways. For example, they can improve their ESC sensors significantly without significant expenditure, while ensuring that incumbent access is protected. Moreover, they can integrate better spectrum management processes in their systems by understanding spectrum conditions better.
• RAN vendors can directly integrate the DSMS software stack into their equipment and communication stacks. By integrating RF sensing software into their equipment, RAN vendors can offer advanced functionality to enterprises, including 5G deterministic networking and URLLC. This means that RAN vendors can start tailoring their value proposition to sell new services, rather than equipment or capacity.
• Neutral host providers and tower companies can host RF sensors themselves or partner with RAN vendors that have integrated this into their systems. Neutral hosts and tower companies can provide detailed RF environment data to tenants, who will be able to optimize their equipment and minimize equipment and deployment costs. This is an example of a new type of service that will allow new forms of monetization.