A simulation testbed to jointly exploit multiple image compression techniques for wireless multimedia sensor networks

Abstract

As the demand for large-scale wireless multimedia sensor networks increases, so does the need for well-designed protocols that optimize the utilization of available networks resources. This requires experimental testing for realistic performance evaluation and design tuning. However, experimental testing of large-scale wireless networks using hardware testbeds is usually very hard to perform due to the need for collecting and monitoring the performance metrics data for multiple sensor nodes all at the same time, especially the node's energy consumption data. On the other hand, pure simulation testing may not accurately replicate the real-life scenarios, especially those parameters that are related to the wireless signal behavior in special environments. Therefore, this work attempts to close this gap between experimental and simulation testing. This paper presents a scalable simulation testbed that attempts to mimic our previously designed small-scale hardware testbed for wireless multimedia sensor networks by tuning the simulation parameters to match the real-life measurements obtained via experimental testing. The proposed simulation testbed embeds the JPEG and JPEG2000 image compression algorithms and potentially allows for network-controlled image compression and transmission decisions. The simulation results show very close match to the small-scale experimental testing as well as to the hypothetical large-scale extensions that were based on the experimental results.