The ultimate goal of this project is to simulate obstacle avoidance behaviour of quadcopters swarm in a Gazebo. For this goal, initially the Crazyﬂie 2.0 was chosen as a quadcopter to be simulated and CrazyS extension of RotorS autopilot was selected to execute the swarm node. Next, it was realized that the RotorS could only send angular velocity command to the rotors of quadcopters while our controller would provide velocity command. Then it was decided to choose PX4 autopilot to communicate with Gazebo through ROS knowing that PX4 supports sending velocity command to Gazebo. Thus, the reference conﬁguration of the parameters of the Crazyﬂie was provided to PX4, but it was observed that based on reference parameters the Crazyﬂies even could not take oﬀ. Afterwards, required drone parameters for PX4 were either collected or calculated based on available resources. Then, each of the identiﬁed parameters were tested for the Crazyﬂie, and it was realized that the Crazyﬂie could take oﬀ. However, there were a large steady state error between the com- manded step velocity to the Crazyﬂie and its real velocity. Therefore, either the Crazyﬂie physical parameters are required to be evaluated by conducting a thorough identiﬁcation or PX4 PID gains for all its embedded controllers should be tuned for the Crazyﬂie. After- wards, it was decided to continue the project based on Iris quadcopter. Finally, obstacle avoidance maneuver of swarms of Irises were studied by utilizing the olfati-saber protocols.