SenSim

This is a sensor network simulation framework written in python. This can be used to simulate different protocols on a user defined network of sensors and targets.

Environment

To handle some implementation obstacles, we had to define an environment where we constrain a 'test area'. This is because when randomly generating the sensors and targets they are generated using numpy.random.rand() which generates a number between [0-1]. And you have to select a range which will ultimately cover all targets. Therefore the default range of all the sensors has been set to 0.5 which solves this problem partially.

To simulate the protocols

Creating a protocol

Create a class, following name convention. If the protocol abbreviation is DEEPS then create a file with DEEPSProtocol.py and give the class name as DEEPSProtocol extending Protocol class given here with. Sample implementation is given at the file LBPProtocol. Three abstract functions should be implemented on_rule, off_rule and shift. These will control how the sensors are turned on and off and shift will control one schedule instance of turning on and off the sensors.

Testing a protocol

Simulation is now a class. To test a simulation, you can instantiate it and then pass the protocol to it. You also can generate a random network using this class.

from Simulation import Simulation
from MTLBProtocol import MTLBProtocol
from Simulation import Simulation

sim = Simulation()
network_items = sim.generate_network(4,7, max_battery = 5)
mtlbp = MTLBProtocol(network_items)

times, coverings = sim.simulate(mtlbp, graphs=True)

Compare created protocols

You can compare two protocol objects on how they behave on their environment. This is supported by Simulation.compare(p1,p2). Currently this implementation is not capable of running graphical simulations but it will give the computed outputs for times of the network cover and sensor alive times.

import matplotlib.pyplot as plt
from MTDEEPSProtocol import MTDEEPSProtocol
from MTLBProtocol import MTLBProtocol
from Simulation import Simulation

sim = Simulation()
network_items = sim.generate_network(4,7, max_battery = 5)
p1 = MTLBProtocol(network_items)
p2 = MTDEEPSProtocol(network_items)

times, p1_coverings, p2_coverings, p1_alive, p2_alive = sim.compare(p1, p2)
plt.plot(times, p1_alive, color='green')
plt.plot(times, p2_alive, colot='red')
plt.show()

Important notes

• These scripts assume that all sensors are initially being turned on and then put through the protocol.
• All the items in the network are placed in the rectilinear space to make the calculations easier

Protocols which are implemented

• Load Balancing Protocol, p3
• DEEPS, p4
• Moving Target Load Balancing Protocol (MTLBP)

TODO

• Generalize adding protocols
• Add animating pyplot plots to visualize
• Add functionality to take simulations out of protocols
• Simulation insights plotting
• LBP battery issue
• Range 20/15