Fei Yu, [email protected] (A paper written under the guidance of Prof. Raj Jain) DownloadRecently there has been growing interest in providing fine-grained metering and control of living environments using lowpower devices. Wireless Sensor Networks (WSNs), which consist of spatially distributed self-configurable sensors, perfectlymeet the requirement. Since running real experiments is costly and time consuming, simulation is essential to study WSNs,being the common way to test new applications and protocols in the field. This survey illustrates some main-stream WSNssimulators including: NS-2, TOSSIM, EmStar, OMNeT++, J-Sim, ATEMU, and Avrora.Keywords: Wireless Sensor Network, Simulation, Simulator, NS-2, TOSSIM, EmStar, OMNeT++, J-Sim, ATEMU, Avrora1. Introduction1.1 What is WSN1.2 Comparison of wired and wireless network1.3 Why use simulation in WSNs2. Basic Concepts2.1 Discrete-Event Simulations and Trace-Driven Simulation2.2 Simulator and Emulator3. Simulation Tools3.1 NS-23.1.1 Overview3.1.2 Merits and Limitations3.2 TOSSIM3.2.1 Overview3.2.2 Merits and Limitations3.3 EmStar3.3.1 Overview3.3.2 Merits and Limitations3.4 OMNeT++3.4.1 Overview3.4.2 Merits and Limitations3.5 J-Sim3.5.1 Overview3.5.2 Merits and Limitations3.6 ATEMU3.6.1 Overview3.6.2 Merits and Limitations3.7 Avrora3.7.1 Overview3.7.2 Merits and Limitations4. Summary5. Reference6. List of Acronyms1.1 What is WSNA Survey of Wireless Sensor Network Simulation Tools http://www1.cse.wustl.edu/~jain/cse567-11/ftp/sensor/index.html1 of 10 5/4/2011 4:52 PMWith the development of embedded system and network technology, there has been growing interest in providing fine-grainedmetering and controlling of living environments using low power devices. Wireless Sensor Networks (WSNs), which consistof spatially distributed self-configurable sensors, perfectly meet the requirement. The sensors provide the ability to monitorphysical or environmental conditions, such as temperature, humidity, vibration, pressure, sound, motion and etc, with verylow energy consumption.The sensors also have the ability to transmit and forward sensing data to the base station. Most modern WSNs arebi-directional, enabling two-way communication, which could collect sensing data from sensors to the base station as well asdisseminate commands from base station to end sensors. The development of WSNs was motivated by military applicationssuch as battlefield surveillance; WSNs are widely used in industrial environments, residential environments and wildlifeenvironments. Structure health monitoring, healthcare applications, home automation, and animal tracking becomerepresentative WSNs applications.A typical Wireless Sensor Network (WSN) is built of several hundreds or even thousands of “sensor nodes”. The topology ofWSNs can vary among star network, tree network, and mesh network. Each node has the ability to communication with everyother node wirelessly, thus a typical sensor node has several components: a radio transceiver with an antenna which has theability to send or receive packets, a microcontroller which could process the data and schedule relative tasks, several kinds ofsensors sensing the environment data, and batteries providing energy supply.A sensor node might vary in size. The “Smart dust” [SmartDust] sensor node, shown in Figure 1, from Electrical Engineeringand Computer Science department at University of California Berkeley, is smaller than a coin. The cost of sensor nodes issimilarly variable depending on the quality of onboard chips. One of the main challenges in WSNs is to decrease the cost andsize. There are an increasing number of small companies producing WSN hardwires. The most popular two are TelosB sensornode, shown in Figure 2, from Crossbow Technology [TelosB] and Tmote Sky sensor node, shown in Figure 3, from Sentilla[TmoteSky].Operating systems for WSN nodes are much less complex than general-purpose operating systems. This is because WSNusually deployed with a particular purpose and low power microcontrollers cannot afford complicated computing. TinyOS isthe most popular operating system specifically designed for wireless sensor networks. TinyOS is based on an event-drivenarchitecture using NesC programming language.Figure 1: Smart dust[SmartDust] Figure 2: TelosB [TelosB] Figure 3: Tmote Sky [TmoteSky]1.2 Comparison of wired and wireless networkThe wired network has been around for decades, as long as the internet itself. Compared with wireless networks, wirednetworks are more secure and faster in transfer speeds. However, wired networks contain one of the biggest growingproblems, wires. Complicated wires and power cords are difficult to manage and hugely degrade the flexibility. Wiring andrewiring are the bottleneck of development of wired network. With the rapid development of wireless technology, more andmore people prefer to use wireless network as their end-user network.Compared with the traditional wireless network, WSN has its own features, such as low cost and low energy consumption. Toreduce cost, each sensor board has very limited onboard resource, such as computing speed, storage and energy source. ToA Survey of Wireless Sensor Network Simulation Tools http://www1.cse.wustl.edu/~jain/cse567-11/ftp/sensor/index.html2 of 10 5/4/2011 4:52 PMachieve long lifetime with limited power supply usually batteries, onboard components are designed to consume energy aslittle as possible. For instance, the transmit power of radio is 1000 times smaller than the one in Wi-Fi routers. WSN is alwaysdeployed in difficult-access areas; the ability of self-configuration is another design goal.1.3 Why use simulation in WSNsNowadays, the WSN is a hot research topic. Many network details in WSNs are not finalized and standardized. Building aWSNs testbed is very costly. Running real experiments on a testbed is costly and difficulty. Besides, repeatability is largelycompromised since many factors affect the experimental results at the same time. It is hard to isolate a single aspect.Moreover, running real experiments are always time consuming. Therefore, WSNs simulation is important for WSNsdevelopment. Protocols, schemes, even new ideas can be evaluated in a very large scale. WSNs simulators allow users toisolate different factors by tuning configurable parameters.Consequently, simulation is essential to study WSNs, being the common way to test new applications and protocols in thefield. This leads to the recent boom of
View Full Document
Unlocking...