combined shadow area area that is not visible to any of the robots was measured for a

robot configuration using this objective function. The objective function not only takes

into account the shadow area of each robot but it also has to incorporate the line-of-sight

criteria. It was decided that line-of-sight communication was maintained as long as there

was an unobstructed path between the robot in question and at least one other robot.

 After a solid objective function was developed, a string representation of the robot

positions was developed. It can be noted that the map resolution correlates directly to the

number of bits used in the binary string. For example, for the 64x64 map, there are a

total of 4096 data points. Therefore, in order to represent all of the corresponding x and y

robot positions using a binary string there had to be a total of 12 bits (212 = 4096). The

most significant six bits represent the robot's x position and least significant six bits

represent the robot's y position (see figure 4.4). The position of each repeater robot is

represented using the binary string representation.

 X1 X2 X3 X4 X5 X6 Y1 Y2 Y3 Y4 Y5 Y6
 Figure 4.4: Binary String Representation of Robot's (x,y) Position

A series of maps were generated with greater resolution in order to demonstrate the

expandability of the method and the versatility of the genetic algorithm. These maps had

a resolution of 256x256, which corresponds to 65,536 data points a considerably larger

search space than the aforementioned. Due to the increased number of data points, the

number of bits in the string was increased to 16 bits (216 = 65,536). A similar binary

string as that in figure 4.4 was used with the addition of two more bits to both the x and y

component of the string.

 The software for this work used the GAlib genetic algorithm package, written by

Matthew Wall at the Massachusetts Institute of Technology [12]. This library required an