By Onn Haran

V2X (Vehicle to Everything Technology) can prevent and mitigate most of the collisions on the road. But it has an inherent limitation: Protection is extended solely to road users equipped with V2X, and their number in the initial market phase is low.

Cooperative perception, also called sensor sharing, solves this. A vehicle transmits information about road users detected by all its sensors. With that, road users without V2X become part of the V2X network. A study^[1] showed that even with 10% penetration, all vehicles on the highway are known.

Although a cooperative perception message (CPM) is standardized, its deployment is planned to start only toward the end of the decade. The industry targets that CPM will use a different channel than the one currently used for Day1 (CAM/BSM) messages. However, that 2^nd channel isn’t available globally, and in some geographies unlikely to be available at all.

CPM is too valuable to depend on 2^nd channel availability. Transmitting CPM in the 1^st channel is a paradigm shift that can solve the problem. Current and planned V2X deployments could include CPM, increasing the value to the OEMs and the end customer, without adding any hardware or cost.

But is it feasible to transmit CPM on the existing channel? It is hard to tell based on the existing studies, as summarized by 5GAA^[2].

Autotalks spectrum study

Recently, Autotalks conducted a new study to determine if Day2 V2X can be expedited. More specifically, the study calculates whether CPM can be transmitted on the current channel allocated to V2X. The study is based on the following principles:

• The number of vehicles is taken from the Day1 models.
• Redundancy mitigation schemes are applied.
  • CPM excludes road users with V2X
  • Information about the same road user isn’t transmitted by more than N vehicles, where N=3 in the basic model, and N=1 in the most efficient model
• The triggering conditions are considered. The periodicity of transmission depends on the distance passed and the speed difference since the previous transmission, as applied in Day1 models.
• CPM usage percentage, also called penetration, is assumed as half of Day1 penetration since aftermarket devices are expected to be common, and those cannot support CPM

Prior studies summed the maximal values of Day1 and CPM bandwidth. This inflates the results because the maximal values of Day1 and CPM are achieved in different penetration rates.

The bandwidth required for a highway scenario is similar between the US and EU, so those are combined for clarity in the graph below:

CPM bandwidth starts to drop at high penetration because fewer non-connected vehicles can be reported by CPM when most vehicles have V2X.

The peak bandwidth is 13.1MHz when up to 3 vehicles can transmit a specific road user in a CPM. When only a single vehicle can transmit, the bandwidth requirement is further reduced to 10.7MHz.

The bandwidth requirements are calculated also for an urban scenario. It is the first study to include the bandwidth requirement of bicycles as part of the Vulnerable Road User (VRU). VRU calculation also includes pedestrians with safety relevance.

The calculations for the US and EU apply different triggering conditions. The higher result, for the US, is shown in the graph below:

With a limit of 3 vehicles transmitting a road user in a CPM, the peak bandwidth is 12.4MHz, dropping to 9.6MHz when a single vehicle transmits.

In both highway and urban scenarios, CPM can use the same 20MHz LTE-V2X channel as Day1. In Europe, where DSRC doesn’t apply automatic retransmission (HARQ), the bandwidth requirement is halved, and CPM can be added to the existing deployed 10MHz channel.

Furthermore, the initial bandwidth usage is very low. The 10MHz mark is crossed only when reaching 50% penetration. Even the optimists will not predict 50% penetration before 2035. There is no reason to add a 2^nd channel now, where the 1^st channel is barely used. Once the 2^nd channel is added, probably for other services, if the 1^st channel will be loaded, some messages can be moved to the new channel, while maintaining backward compatibility.

Practically, the deployment profile should be amended to enable CPM. CPM parameters should assure low bandwidth consumption. Redundancy mitigation should be mandated, and safety-irrelevant pedestrians should be ignored.

To conclude, the bandwidth required for Day1 and CPM services is bounded by ~13MHz, fitting comfortably inside the existing 20MHz LTE-V2X channel.

See also: DSRC vs CV2X

---

[1] Collective Perception: A Safety Perspective – https://www.mdpi.com/1424-8220/21/1/159/pdf

[2] https://5gaa.org/content/uploads/2021/10/5GAA_Day1_and_adv_Use_Cases_Spectrum_Needs_Study_V2.0.pdf