both the transmitted and reflected waves. Then, using a localized cooperative pressure

measurement the sensors detect the direction of maximum intensity in which they move

and finally reach the object. Here the spacing between the sensors needs to be the same

for accurate pressure measurements. Although this requirement is satisfied, the group

moves in no particular formation towards the target. Moreover, the sensors need to share

their positions and pressure measurements in order to move collectively. The proposed

self-organization algorithm minimizes the sharing of information between agents.

 The most difficult situation arises when the robots have no knowledge about the

nature and position of the target and a decentralized strategy is required. Hence, some

kind of external guidance needs to be given to the robots to move them towards the

target. The above requirements suit for applications where minimal risk and high

indestructibility (ex. bombing of a target), are desired. The target tracking algorithm for

this situation is investigated in the next section.

 4.2 Decentralized Target Tracking Algorithm

 The technique used for decentralized and collective target tracking can be discussed

with the help of figure 4.1. Two external base stations are used to transmit the direction

information to the swarm. The two base stations can also be assumed as stationary IPCs

transmitting IF to the swarm. But this IF needs to be the same for all the agents in the

swarm and has to be rotated properly in order to move the swarm properly towards the

target. This can be achieved as follows. Each base station is assumed to have a simple

radar through which it detects the position of the target as well as the position of the

center of the swarm. With the above position information, each base station then decides

independently on whether the center of the swarm is to the left or right of the line of sight

(LOS) with the target. Accordingly, this direction information is transmitted to the