How different transceiver hardware effects concurrent transmissions in WSNs

Start Page
139
End Page
146
Article Number
8510975
Citation
Proceedings - 14th Annual International Conference on Distributed Computing in Sensor Systems, DCOSS 2018: 8510975, 139-146 (2018-10-25)
Contribution to Conference
14th Annual International Conference on Distributed Computing in Sensor Systems, DCOSS 2018  
Publisher DOI
10.1109/DCOSS.2018.00030
Scopus ID
2-s2.0-85057138823
Publisher
IEEE
The efficiency and effectiveness of network flooding protocols has recently been shown by many publications as well as real world deployments. One notable protocol is Glossy which combines Concurrent Transmission (CT) with Constructive Interference (CI). A drawback of Glossy is that up to now it has only been implemented and evaluated for one type of radio transceiver. In this paper we first present the Glossy implementation for an AT86RF233 radio transceiver which simplifies the packet forwarding due to a shared receive and transmit buffer. We evaluate our implementation against the original Glossy implementation in a minimalist setup as well as in a real world testbed. However, we observed a noticeable difference in the timing accuracy of the transceiver chips but without adverse effects on the network's performance. For this reason, in the second part of this paper we provide a deeper investigation by using Software Defined Radios to emulate concurrent transmission for different types of transceivers. This work confirms the general benefit of constructive interference but also shows its limitations. Moreover, we give novel insight into hardware dependability of constructive interference in concurrent transmissions.
Subjects
Concurrent Transmission
Constructive Interference
WSN
DDC Class
600: Technik