Adaptive Multi-Path Routing Protocol in Autonomous Vehicular Networks

Vehicular ad hoc networks consist of self-organizing nodes using multi-hop wireless links for communication without any infrastructure support. Traditionally, ad hoc routing protocols use the minimum hop count for their routing metric since a smaller number of transmissions is typically equivalent to a higher throughput, lower delay, and minimal power consumption. However, with the muti-rate capability of emerging radio interfaces, e.g., 802.11ax/be standards, the min-hop metric no longer results in high throughput. For instance, if the higher data rate links are selected for the route, it could result in a higher throughput even if the route takes more hop counts. In this paper, we propose a high throughput routing scheme, called MARV, which makes two key contributions. MARV searches for high throughput paths using an on-demand route searching algorithm so that the routing overhead is smaller compared to other multi-rate-aware routing schemes. MARV also searches for multiple paths to maintain both min-hop and high-throughput paths to select the adequate path depending on the data packet size. We conduct simulations to demonstrate that MARV outperforms not only min-hop path metrics but also previously proposed high-throughput metrics.

Automated summary

Vehicular ad hoc networks are composed of self-organizing nodes that communicate using multi-hop wireless links without any infrastructure support. Traditional ad hoc routing protocols use the minimum hop count as their routing metric, but this no longer results in high throughput due to the emergence of multi-rate capable radio interfaces. This paper proposes a high throughput routing scheme called MARV, which searches for high throughput paths using an on-demand route searching algorithm, reducing the routing overhead compared to other multi-rate-aware routing schemes.