VMNet: Realistic Emulation of Wireless Sensor Networks

Abstract

Abstract-Many research activities on Wireless Sensor Networks (WSNs) need detailed performance statistics about protocols, systems, and applications; however, current simulation tools and testbeds lack mechanisms to report these statistics realistically and conveniently. To address this need, we have developed a WSN emulator, VMNet. VMNet emulates networked sensor nodes at the level of CPU clock cycles and executes the binary code of real applications directly. It emulates the radio channel with loss and noise as well as emulates the peripherals in sufficient detail. Moreover, VMNet takes parameter values from the real world and logs detailed runtime information of emulated nodes. Consequently, the application performance, both in response time and in power consumption, is reported realistically in VMNet, as demonstrated by our comparison studies with real sensor networks.

Notos: Efficient Emulation of Wireless Sensor Networks with Binary-to-Source Translation

“…However, ATEMU does not support parallel emulation and, thus, does not benefit from multi processor or multi core systems. VMNet [27] enhances ATEMU with support for host-networking but does not address the scalability issue.…”

TEPAWSN - A tool environment for Wireless Sensor Networks

“…at the instruction level [7], control algorithm level [8], hardware level [9], etc. Nevertheless, power analysis of WSNs has been mainly and currently addressed in a simulation context [10], [11].…”

Distributed Cross-Layer Scheduling for In-Network Sensor Query Processing

“…(Figure 9). We also used VMNet [22] to emulate the MICA2 motes and evaluated our scheme on this realistic testbed. We chose VMNet because it can provide detailed runtime information of emulated nodes whereas a real WSN cannot due to the limitation of the real sensor nodes and the measuring equipment.…”

“…The difference between the power consumption measured in VMNet and that in the real WSN was within ±15%. Possible reasons for this difference include the difference in the topologies and the measurement errors of the oscilloscope [22].…”

“…Among these four sources, idle listening and overhearing are the dominating ones [17]. In our previous experience with the MICA2 [6] networks, these two factors cost more than 70% of the power during query processing [22]. Therefore, in our scheduling scheme, we focus on reducing idle listening and overhearing and identify a number of constraints for schedule construction to reduce collision.…”