The LTE Connected Car

Cell-based connectivity for vehicles is quickly becoming a serious competitor to direct vehicle-to-vehicle communication. For many of the envisioned VANET applications that are based on multihop routing I would say that cell-based, centralized, communication architectures are a more robust, efficient and feasible alternative. But the case for direct V2V communication is not dead, these technologies will work side by side to provide both high bandwidth "bulk services" for entertainment and comfort as well as low latency, low bandwidth safety critical services.

On January 19 telecom company Alcatel Lucent demoed a Prius connected to a Long Term Evolution (LTE) basestation outside of Paris. From the press release:

"In addition to the LTE Connected Car, the event also will feature drive demonstrations, in an LTE-equipped R&D van – used to test the capabilities of LTE technology -- of mobile services on the company’s campus-wide LTE network operating at 2.6 GHz, which is expected to be one of the leading frequency bands for LTE in Europe. The demonstration, using Alcatel-Lucent’s end-to-end LTE solution and a mobile device from LG Electronics, will feature multiple, simultaneous streaming video sessions, multiplayer mobile games and photo and music downloads, as well as a smooth transition between 3G/HSPA and LTE networks, a critical requirement for operators that need to ensure a seamless quality of experience for their subscribers as they begin to deploy LTE."

In addition to the communication part, the demo also showcases how this could be integrated into several cool services in future cars. Now we just need to fill up our cars with passenger to utilize all these services that the driver can't because he's driving, the average occupancy of vehicles in the SF Bay Area in 2006 was 1.3

Crook-to-crook communication

Inter-vehicle communication enhances the ability of the vehicle and the driver to perceive and react to the environment around them. For example, wireless communication could alert drivers of each other even in situations where they are unable to see each other directly, such at a blind intersection.

Unfortunately, some traffic participants don't always want to be seen. Law enforcement setting up speed checks often want an element of surprise. The idea being that drivers will stick to the speed limits as they don't know if a speed check is coming up. A french company, Coyote, has a system based on centralized wireless communication that allows drivers to easily "tag" locations where they observe speed checks, effectively warning other drivers to slow down, or conversely reassuring them that they can speed otherwise. According to this article they already have a user base of 250.000 subscribers in France.

This is not completely new, you have been able to download point-of-interest (POI) databases of automatic traffic cameras for quite some time, but the dynamic nature of users submitting tags is a new twist. I wonder if the French police can turn this to their favor by analyzing the lag in user feedback loop. If it is possible to set up speed checks at a rate faster than the Coyote system can invalidate old warnings they could in fact appear to be at multiple locations simultaneously. Another exploit could be to inject bogus data into the Coyote system, however I guess some type of reputation system might be employed to stop such malicious users.

Update: Someone also pointed out that there are other services like this that do not rely on specialized hardware, rather take advantage of the fact that most modern cell phones contain positioning devices, e.g. Trapster.com