Claims:The scope of the invention is defined by the following claims:

Claim:
1. Channel based allocation algorithm is proposed for the VANETS comprising the following steps
a) In real-world scenarios, our model-based dynamic programming method generates better results in less time
b) Our model-free RL method, which is based on numerical data, converges to more dependable solutions.
c) In order to fully use the vehicular network's capabilities, effective channel allocation mechanisms are required due to the rapid proliferation of applications that require channel resources.
d) Two Reinforcement learning (RL)-based channel allocation algorithms for maximizing a long-term average system reward in a cognitive enabled VANET environment are proposed.

2. Channel based allocation algorithm is proposed for the VANETS as claimed in claim1, provides the vehicular network's capabilities, effective channel allocation mechanisms, the rapid proliferation of applications that require channel resources.

3. Channel based allocation algorithm is proposed for the VANETS as claimed in claim1, Reinforcement learning (RL)-based channel allocation is preferred because of maximizing a long-term average system reward in a cognitive enabled VANET environment. , Description:Field of Invention
The Vehicular Ad-Hoc Network (VANET) is a mobile network created by moving automobiles acting as nodes. Every vehicle node included in VANET is turned into a wireless router. Cars with a range of 300 to 500 meters are able to connect and share information in this type of network. As a result, a larger network is developed. As the first cars/vehicles leave the network's connection range and get disconnected, new cars will join the network, becoming a mobile Internet. Most likely, fire and police vehicles will be the first to integrate this technology and interact with one another for safety reasons.

Background of the Invention
The DSRC spectrum is split into seven 10 MHz channels: one control channel for safety and control messages, and six service channels for safety and non-safety applications (Cheng, Bin, et al. IEEE, 2017). The IEEE 802.11p and 1609 working groups have been working on DSRC standardisation in the United States since 2004. The physical layer operation, node transmission procedure, and channel accessing mechanism are all specified in IEEE 802.11p (Wang, S. Y., et al. 2009 IEEE).
The IEEE 1609 protocols are primarily concerned with upper-layer operations. The wireless access in vehicle environment (WAVE) stack is made up of the IEEE 802.11p and 1609.x standards (Hussain, SM Suhail, et al. IEEE 2018). The WAVE protocol stack's overall architecture is described in IEEE 1609.0. IEEE 1609.1 provides the control message format and data storage format in the WAVE system, as well as the resource management control flow and standard interfaces for application registration and maintenance.
Our model-based planning technique is depicted in Figure. The SMDP model's state transition probabilities and estimated time intervals between decisions epochs are first trained using data collected from the environment (CN102662764A). The best channel allocation policy is then determined using the planning approach. In general, a closed-loop model-based planning strategy is one in which the model is continuously trained while the system is running, allowing the policy to respond to changes in the environment (CN102026841B).
The objective of the invntion is to design a channel allocation algorithm for the VANET for the cognitive environment and to design a novel path planning and planning method. Two Reinforcement learning (RL)-based channel allocation algorithms for maximizing a long-term average system reward in a cognitive enabled VANET environment.
Summary of the Invention
When a vehicle reaches an RSU's coverage area and intends to request service, the RSU will recognise the vehicle user's type and decide whether to accept or refuse the request based on available channel resources. If the SU (Secondary User) request comes from a secondary user, RSU will allow it as long as there are vacant channels available. The RSU should determine the number of channels allotted to that service request. More channels equal a higher transmission rate, which means more revenue from that user's delight. Accepting that user's service request with more channels, on the other hand, means fewer channels are accessible for other incoming customers.
Brief Description of Drawings
The invention will be described in detail with reference to the exemplary embodiments shown in the figures wherein:
Figure 1 Model Based Planning
Figure 2 Path Planning
Figure 3 Reinforcement learning model
Figure 4 Path planning method
Figure 5 Flow chart of proposed method
Detailed Description of the Invention
In real time, the road accidents due to high congestion of vehicular traffic increase the mortality rate and injure numerous people worldwide. Dissemination of road safety messages significantly supports the drivers to avoid accidents and enhance the road safety. Vehicular Ad hoc Network (VANETs) are a special type of mobile ad hoc networks (MANETs) which enable effective communication among vehicles. A VANET comprises a collection of high-speed vehicles and fixed Road Side Unit (RSU). Each vehicle is equipped with an On Board Unit (OBU) to establish communication with other vehicles or RSU. The RSUs have a higher communication range than that of the vehicles, and they act only as routers. A vehicle can act as a sender, receiver, and router to ensure effective communication. The VANET offers two types of communication such as vehicle-to-vehicle (V2V) and Vehicle-to-RSU (V2R). In V2V communication, the vehicles transmit data to other vehicles in a single or multi-hop manner without relying on any infrastructure support.
In safety applications, each vehicle periodically broadcasts a safety message that comprises a vehicle position, speed, and road conditions. Vehicular networks have the potential to minimize the accidents on roads and improve the security of drivers. The V2R communication allows a vehicle to communicate with RSU and assists in obtaining toll details, parking areas, emails and traffic conditions along the road. VANET is a network that combines ad hoc, WLAN, and cellular technology into a single system. It's a mobile ad hoc network (MANET) in which cars communicate with one another and with roadside infrastructure (RSUs). VANETs are self-organizing, distributed networks that form the backbone of the Intelligent Transportation System (ITS). The major purpose of VANET is to make transportation safer by enabling efficient vehicle-to-vehicle communication, allowing for the rapid dissemination of safety and essential signals. The high mobility of vehicle nodes, the variable density of vehicle nodes, and the network's infinite range are some of the distinctive aspects of VANETs.
As a result, the RSU must make sound decisions in order to solve the channel allocation problem. If this RSU's channels are all busy, the request will be rejected. As a result, when an SU request is received, the RSU will either accept it or allocate available channels to the service, or it will refuse it. If the request comes from a primary user, the RSU will always give primary users priority. As a result, whether there are accessible channels or not, the RSU will always accept PU requests. The RSU will allocate a fixed number of channels to the PU service when there are idle channels. The RSU should also decide on the number. If no empty channels are available, the RSU will stop a certain number of SU services in order to provide channels for that PU request. The base station will host all SU services. In this instance, the RSU should also optimize the number decision. There is a compromise between transmission rate and transferring cost for some SU services.
When GPS fails, people nowadays propose a variety of techniques to tackle navigation problems. These techniques, however, have a number of flaws. The most basic way is the previously described inertial navigation systems (INS) or dead reckoning (DR). This is a method of navigating by applying a calculation, but it will result in errors and move the placement. This necessitates regular GPS correction, but if GPS fails for an extended period of time, the error will increase. It's possible that the car is going through an underground tunnel in a city and is relying on INS or DR because GPS has failed. The INS/DR inaccuracy is increasing over time, and the vehicle may miss the intersection and continue straight.
Our model-based planning technique is depicted in Figure1. The SMDP model's state transition probabilities and estimated time intervals between decision epochs are first trained using data collected from the environment depicted in Figure 2 and 4. The best channel allocation policy is then determined using the planning approach depicted in Figure 3. In general, a closed-loop model-based planning strategy is one in which the model is continuously trained while the system is running, allowing the policy to respond to changes in the environment depicted in Figure 5.
3 claims & 5 Figures