SenseSim: An agent-based and discrete event simulator for Wireless Sensor Networks and the Internet of Things

This paper presents the concept and the main design aspects of the new sensor network simulator called SenseSim. It is also a tool which can be used for researches on the Internet of Things. In contrast to most of existing simulators in this area, SenseSim does not mainly focus on the wireless communication issues. It has been designed to simulate heterogeneous sensor network as an autonomous system which observes changing phenomena. Moreover, the behavior of the devices can be modified by macroporgramming the network. Both sensors and phenomena can be easily configured. SenseSim uses efficient simulation engine called DisSim and its modular design allows to expand its functionality and interoperate with real devices.

I. Introduction

Computer simulation is currently one of the most popular way of conducting researches on the Internet of Things and Wireless Sensor Networks. It allows researchers to verify new ideas in a virtual environment without the need to use any hardware. There are several dozen of currently available simulators. Following Musznicki and Zwierzykowski [1] they can be divided into the following groups: •

Emulators - they emulate a sensor hardware or pro-cess/interpret loaded programs in a way it would be executed on a real device. Emulators usually are platform specific and execute low level programs. Examples are: Avrora [2] TOSSIM [3] and WSim [4];

•

Topology control - those simulators allow to construct network topology and maintain it by applying appropriate algorithms and protocols. An example is Atarraya [5];

•

Wireless environment and medium simulators - those platforms focus on simulating wireless communication, taking into account hardware configuration like antenna type, signal strength etc. Examples are: Prowler [6] and WSNet [7];

•

Network and application level simulators - those tools are dedicated to simulate sensor network as a transporting structure where mainly routing algorithm can be tested. Examples are: AlgoSenSim [8], NetTopo [9], Wireless Sensor Simulator [10];

•

Cross level simulators - simulators in this category are able to simulate devices and network at various levels of abstraction. It is possible to evaluate routing algorithms, energy and resource consumption and simple applications. Examples are: COOJA [11], J-Simand Sensor Network Package [12], CupCarbon [13].