The simulation results after phase 2 is completed (a) The first run (b) The second run. 

Context in source publication

Context 1

... of the most basic features of swarm robotics is network coordinate building (localization), as only after that, algorithms such as pattern formation, mapping, or collective movements can continue to be integrated into the swarm. In this paper, we use the trilateration algorithm, proposed in [ 9], which has been used extensively in wireless sensor networks. This algorithm has two phases. First, each robot builds their own local coordinate system by using information about the distance of surrounding robots. Fig. 3 depicts the simulation results on Unity after phase 1 is complete. It should be noted that, since this process is random, each time the simulation is run, a different result is obtained. In phase 2, all coordinate systems are rotated or mirrored to con- verge into one unique network coordinate. Fig. 4 describes the simulation results after this phase is completed and Fig. 5 shows the details about the coordinate of a robot (with ID is 15). As can be seen, each time the simulation is run, a different network coordinate system, with another origin ID and rotation, is obtained. Fig. 5 is also a great example of how all public properties of a script can be viewed directly in Unity without the need to manually print each of them out in the debug console. Thus, this greatly simplifies the debugging process. In this paper, we have presented and verified how Unity can be used to simulate swarm-robotic system. With the ability to switching between 2D and 3D environments, using state of the art physics engine, flexible programming model, and a good debug interface, Unity has met all of our requirements and is comparable to other robotic simulators such as V-REP, Webot, Gazebo and MORSE. Therefore, in future works, we will continue to implement and simulate the S-DASH algorithm [8] on both the simple 2D robot ...