Vehicular Communication Simulation - Group #5

V2V Communication - Simulation
(Vehicle to Vehicle Communication System)

- by : N. Gnana Sai, Pranjal Sahu, Udit Sethi, Y. Sai Akshay

When you're driving on a busy lane, you have to always be aware of the conditions around you and take in consideration even the conditions on the lanes which are quite far ahead of you. Imagine a Scenario in which the vehicles could communicate with each other and act accordingly, say - a person is trying to overtake your vehicle from one side and the required information regarding it like the speed of the vehicle, the distance and various other parameters have already been communicated to you. Then one can use the communicated information to act accordingly. All this is possible now thanks to V2V - "Vehicle to Vehicle" or even V2X - "Vehicle to Everything" communication technologies. Not only does this increase the safety feature advantages in a vehicle but also opens up new areas and business opportunities for different sectors such as Internet of Things, Robotics and many more.

Vehicle-to-vehicle communication comprises a network where automobiles communicate with each other with current state information like speed, location, the direction of motion, sudden braking, loss of stability, etc. Vehicle-to-vehicle communication uses DSRC (dedicated short-range communication) a standard set by ISO. V2V is a mesh network meaning every node of the network can receive, send and forward data. The goal of this project is to simulate the V2V communication system using appropriate software.

As the Internet of things expands and we are set on the course of attaining 5G services, the automotive industry is geared up for some monumental changes.

With the ever-increasing traffic density, the number of accidents is also going up day by day. Along with automated systems which react on the information provided through communication between vehicles, these technologies carry the potential to radically improve transportation, viz-a-viz from reduced collisions to increased fuel efficiency.

In our project, we focus on the Vehicle-to-vehicle communication aspect only.

The plan is to model and simulate a real-world-like scenario of a cluster of vehicles which interact with each other and react to the changes around them maintaining and specifically increasing the safety measures to be taken and the work up the efficiency that is achieved by manual ways.

Implementation

Now we'll look at how we implement the above criteria in terms of Hardware and Software.

Hardware Implementation of Prototype

Every vehicle must consist of (Arduino and a Wi-Fi modem) or Node MCU, GPS sensor, LM393 speed sensor module, CAN communication module.
Most vehicular positioning devices e.g. a low-cost GPS receiver has too large of position error to be used for these applications, hence we use vehicular positioning accuracy enhancement (VITAE) algorithm. It capitalizes on reducing the position error by using GPS estimate and the sensing estimate of each nearby vehicle via GPS machine and local sensors respectively.
The local sensors e.g. vehicular radar uses greedy data-association(GDA) algorithm to find a minimal-weight matching between a GPS estimate and a sensing estimate based on the Euclidean distance.
We will take the speed of the vehicle from a speed sensor module and the data of the communication module and send it to other vehicles using a suitable multicast routing protocol.

Software Implementation

We can treat each node as a vehicle. These nodes can then optimize over weight-based cost function, which penalizes on failure and delay of delivery. This algorithm can then be deployed in a virtual environment simulator as suggested below:

• OMNET++ and Sumo
• NS3 and Sumo

Both of these to be used along with SUMO but we will be using OMNET++ along with some additions which make the IDE better and add required functionalities, We will deploy Artery and GEMV2 which are to be used alongside OMNET++.

They work hand-in-hand with the Artery simulation running, SUMO is the vehicle traffic simulator which has the modelled networks in the form of XML files defining parameters such as the road network, traffic demand – defining how many vehicles are supposed to be on their way etc. And it may even consist of some miscellaneous polygons which may describe the outline of buildings, structures etc.

We observe that for different reasons high node density scenarios appear to be critical for the overall performanceof both systems. This includes end-to-end delay, packet error rates and a non-optimal channel utilization that leaves room for improvement. We find that the approach proposed by ETSI ITSG5 with Decentralized Congestion Control (DCC) may access the channel rather conservatively but still outperforms IEEE WAVE in most of the scenarios.

Artery does not provide any simulation model of wireless communication itself, but employs existing models. At present, you can choose between the IEEE 802.11 model of Veins and INET depending on your needs. If you are not sure which one to pick, we mostly use the INET model nowadays for our simulations.

Artery wraps several entities of an ETSI ITS-G5 stack as OMNeT++ modules, e.g. the router, congestion control, and security entities. Vanetza provides the implementation of these entities, however, Artery makes these accessible in OMNeT++. Each vehicle possesses its own instance of the ITS-G5 stack. Instead of using ITS-G5 we will also be trying to use some other standards such as 5G.

Results : simulating...