Joint Communication and Sensing in 6G Networks

In this recent blog by Ericsson, a common view of Ericsson and NXP is shared on the importance of joint communication and sensing (JCAS) in 6G networks, reflecting the collaboration between Ericsson and NXP for this to materialize. In this blog, we will elaborate on this view from NXP's standpoint and explain the concepts based on a few example use cases.

But first, let’s quickly review the JCAS topic in the blog by Ericsson. As Ericsson already indicated, the inclusion of sensing capabilities in a communication network is a very promising area that presents many opportunities and challenges. There are use cases applicable to improving the performance of the network itself and also exciting new use cases where the spatial sensing can be offered as a service to users or applications that are external to the network.

The main advantage of the communication network in terms of future sensing is that most of the infrastructure will already be in place with transmit/receive (Tx/Rx) nodes, providing full area coverage as well as a good interconnection between nodes, which facilitates a multi-static sensory mesh. Hence, the sensing can be provided almost ‘for free’.

NXP has teamed up with Ericsson to look into potential new use cases for a network where communication and sensing functionalities are fully integrated into the same transmission/reception nodes. Together we will also investigate implementation aspects of such future systems and evaluate the technical challenges and opportunities for JCAS use cases.

Sensing in a Network Setting

In their recent blog, Ericsson laid out how network sensing may be implemented:

With the evolution of 4G to 5G, the spectrum allocations have expanded towards higher frequencies. This trend will continue and communication spectra in the sub-Terahertz region will likely be available as some of the frequency bands for 6G deployments. With the introduction of these new frequencies, the potential for very accurate sensing based on radar-like technology arises. That is, reflections of transmitted signals are received in the network and processed to yield spatial knowledge of the physical surroundings.

Sensing as an integrated capability is of interest throughout the frequency range used by mobile communication networks, starting as low as 700 MHz, where the lowest time-division duplex (TDD) bands are located. Having said that, this blog post will mainly focus on the higher, sub-Terahertz bands.

At these frequencies, the communication network must employ beamforming of the transmitted signals to concentrate and direct the signal energy to a specific geographical area where the intended receiver is located. The inter-site distance (ISD) necessary to create full geographical coverage without beamforming would be prohibitively short. To create full coverage, beamforming combined with beam sweeping over time can be employed. Hence, the mechanisms for beamforming are already present in the network and can then also be utilized for sensing.

Sensing Use Cases: Intersection Safety

As this is still a rather abstract description, let’s have a look at real life in the context of smart cities. According to research by IOT Analytics, ‘Traffic Monitoring and Management’ comes second in the top 10 smart city use cases, only after ‘Connected Public Transport’¹. Also, other use cases in this top 10 relate to traffic-related sensor information utilization such as ‘Video Surveillance and Analytics’, although replacing the video in this case by 6G sensing, reducing privacy concerns. To imagine a more concrete example in the context of Traffic Monitoring and Management, consider a complex intersection. Modern cars are equipped with many sensors, such as cameras and radar. However, these sensors all have a disadvantage: they only work in line-of-sight situations. Vehicle-to-Vehicle (V2V) communication  ² partly fills this gap, but only for equipped road users. If, for example, line of sight to a vulnerable road user is occluded by another traffic participant, as by the truck in figure 1 below, in-car sensors have no way of detecting this.

As observed by Ericsson:

Broadening the scope to sensing as a feature offered by the network, all kinds of spatial monitoring is of interest. Let’s first look at the basic sensing capabilities. Depending on the frequency, the resolution of the sensing image that can be obtained varies. For frequencies around 100 GHz and their typical bandwidths it is possible to reach below 1 cm. However, the resolution also depends on the reflective properties of an object as well as the proximity to other nearby objects and their reflective properties. In comparison to a visual image from a camera, a sensing image based on the reflections from the transmitted signals is quite crude. However, this sensing method offers other attractive properties that a camera cannot provide.

By measuring the delay of the return echo in the line-of-sight path between the transmitter and the object, the distance to the object can be calculated, and therefore, its position. Similarly, by measuring the Doppler shift in the received echo, compared to the transmitted signal, the velocity of the measured object can be calculated.

Another useful feature of sensing based on radio signals is the fact that it also works in complete darkness. It can also ‘see’ in rain or fog, but with somewhat degraded performance, since the water particles in the air will attenuate the signals, especially at higher frequencies.

Using the above-mentioned features, by means of network sensing we can collect position and velocity data, and to a certain extent (based on signatures of the reflection, e.g., due to leg movement of a cyclist) even an indication of the type of object. Thereby a high-quality map with dynamic object data on an intersection can be obtained giving a spatial and temporal ‘image’ of what is happening in real time. This can be used to find potential safety and/or traffic flow issues. The required sensing can be done using just the 6G sensing capable systems without installing additional sensor systems, thereby reducing the complexity (and visual impact) of the intersection infrastructure significantly. Alternatively, the data may be fused with data available from other sensors and/or with data from traffic control centers such as real-time traffic light schedule information.

Using the obtained information, we can go two routes: vehicle-centric versus infrastructure-centric processing. In the vehicle-centric approach, the data in the vehicle can be augmented by sensor data provided by the intersection, distributed by the same 6G network infrastructure. Here the vehicle does the final fusion and determines any actions to be taken. In the infrastructure-centric approach, the mobile edge or cloud computer running the intersection traffic control could itself issue a warning to the car and its driver, using direct vehicle-to-infrastructure (V2I) vehicle-to-everything (V2X) communication or cellular communication; or in case the vehicle does not have advanced communication methods itself, alert or redirect a driver using infrastructure signs (e.g., using the traffic lights themselves).

The selected implementation will depend on many considerations, ranging from technical capabilities and limitations to assignment of liabilities in the system. In the vehicle-centric approach the vehicle manufacturer is in control, relying on NXP’s Functional Safety compliant automotive systems. In the infrastructure-centric approach, part of the responsibility can be shifted to the road infrastructure operator. A more hybrid approach, making use of best aspects of combined functional safe automotive systems with wide area network communication and sensing capabilities will allow full optimization.

In the figure above, the car turning left is warned about the bicycle on the road, obscured from sight (and thus also from in-car sensors such as radar) by the truck waiting for the traffic light. Likewise, the information can be used for a wealth of other intersection related warnings and use cases, either alerting drivers, rerouting traffic, or otherwise. This all can be done leveraging installed 6G connectivity systems enhanced with sensing capabilities. Having a 6G connection to the traffic control back-office and traffic control centers, the data can be used real-time for wider area traffic flow monitoring and management.

The described sensing system is based on millimeter wave (mmWave) radar technology which is tolerant to poor visibility situations [at night or in poor weather conditions] and does not make use of cameras. An advantage of not having to rely on cameras is also that this may even help acceptance of such management in the context of privacy regulations, as no actual video/image data is used, reducing fear of people being identified personally or feeling ‘tracked’ by the traffic management systems.

It is good to realize there is quite an ecosystem involved here, offering business potential not only to the equipment manufacturer but to multiple organizations in the traffic management chain. There is the network operator, able to monetize the gathered additional data and there is the road infrastructure operator, using the data to enable implementing advanced traffic management and safety use cases. Local municipalities see a reduction in accidents and improved traffic flow, and teh vehicle owner, with a subscription to this data or otherwise having access to the warning issues by the sensor system, may appreciate the increased safety and reduced travel time.

Sensor Infrastructure

With the continued growth of mobile wireless communication, frequency spectrum will remain a scarce resource, to be handled with care. This means there is a drive to enable more frequency bandwidth for telecoms industry use, which is a long-term trend already, but it also means an effective frequency re-use is needed especially in densely populated or typically crowded areas. With ever increasing wireless data network throughput needs (by a factor 100x in a 10-year period), we consequently see that the wireless network infrastructure becomes more and more densely populated with radio capable devices.

Today, in cities, base station inter-site distances of about 1 kilometer are common and further distance reductions are expected, making use of small cell technologies that include high-frequency beamforming radios. The high-frequency usage of mmWave domain spectrum as in 3GPP 5G NR, up to the 52 GHz band, naturally facilitates high spectral re-use coming from the close-range confined propagation characteristics. Now moving to the next generation, the expectation is that more bandwidth in the sub-Terahertz region will be made available as frequency bands for 6G deployment. Realistically, only at these higher frequencies potentially larger chunks of bandwidth can be made available for communication.

Looking at the combined trends, we see that infrastructure elements equipped with mmWave beamforming technology are brought physically closer to comms users, people, devices etc., in areas with high communication network data throughput needs. Communication and sensing now have the potential of building a strong complementary partnership. Here, situational awareness is key, as is the ability to communicate. The situational awareness coming from the 6G network-based sensing capability will be helpful here to increase safety, building further on automotive safety use cases which are supported by NXP from its automotive radar business.

In addition to this, the value of the radio infrastructure RF solutions will further increase with the additional support of sensing features in the radio. This is a strong future enhancement to NXP’s RF portfolio, providing high-performance solutions connected to the radio base-station side antenna elements. NXP is excited to work with Ericsson, bringing the strength of radio networks and mmWave system technologies together to enable many new use cases, combining the strengths of communication and sensing technology.

Conclusion

As we conclude this first blog on 6G sensing, we see this as a great development, moving towards integrated efficient sensing within the network. It offers new possibilities for smart city environments, building on the automotive and industrial strengths of NXP.

As Ericsson concluded in their blog:

The future sensing capabilities perceived in 6G networks present many opportunities. There are multiple use cases where spatial sensing can be used to improve the performance of the network itself, as well as provide new, exciting sensing services to external users and applications. There are, of course, also technical challenges to handle.

The infrastructure of Tx/Rx-nodes in an area-covering network, in combination with good interconnection between the nodes, provides an excellent setting for a multi-static sensory network. Since this infrastructure is already in place in a communication network, the addition of sensing capabilities can be provided in a very cost-efficient manner.

We, NXP together with Ericsson, are excited to explore the potential of JCAS and expect it to become an important part of future 6G systems. It will be very interesting indeed to evaluate different solutions with new use cases and provide the best possible sensing capabilities from the network.

In a next installment of this blog, we will have a look at industrial use cases for 6G sensing. Stay tuned.

1. The top 10 Smart City use cases that are being prioritized now; IOT Analytics, September 2020
2. Wikipedia Vehicle-to-everything

To learn more about Connectivity, please visit here.