TY - GEN
T1 - Revisiting broadcast algorithms for wireless edge networks
AU - Rosa, André
AU - Costa, Pedro Ákos
AU - Leitão, João
N1 - The work presented here was partially supported by the Lightkone European H2020 Project (under grant number 732505), NG-STORAGE (FC&T grant PTDC/CCI-INF/32038/2017), and NOVA LINCS (FC&T grant UID/CEC/04516/2013).
PY - 2019/10
Y1 - 2019/10
N2 - With the advent of Edge Computing, suitable, practical, and novel abstractions are required for applications to leverage the existing computational power at the edge. In particular, applications in the domains of smart cities and the Internet of Things (IoT) can rely on devices in the vicinity of data consumers and producers for their operation. While these devices are expected to be equipped with wireless radios, network infrastructure might be unavailable in many scenarios. In those cases, devices must rely on wireless ad hoc networks for coordination and cooperation. In this context, one of the most important primitives is the broadcast of messages, that can be leveraged as a building block to devise more complex distributed services and applications. The literature on wireless ad hoc broadcast algorithms is quite vast, with many different algorithms being proposed which explore or combine different techniques or features in their operation. While such protocols are becoming increasingly relevant, understanding how they relate among them is complicated. To address this challenge, in this paper, we introduce a novel framework that allows to abstract the operation of wireless ad hoc broadcast protocols. Leveraging on our framework, we explore a particularly interesting class of these protocols: neighbor-aware ad hoc broadcast protocols; of which we propose 4 novel protocols. Finally, we rely on a materialization of our framework to implement prototypes of these protocols and experimentally study their performance in a testbed composed of 21 Raspberry Pi 3 - model B.
AB - With the advent of Edge Computing, suitable, practical, and novel abstractions are required for applications to leverage the existing computational power at the edge. In particular, applications in the domains of smart cities and the Internet of Things (IoT) can rely on devices in the vicinity of data consumers and producers for their operation. While these devices are expected to be equipped with wireless radios, network infrastructure might be unavailable in many scenarios. In those cases, devices must rely on wireless ad hoc networks for coordination and cooperation. In this context, one of the most important primitives is the broadcast of messages, that can be leveraged as a building block to devise more complex distributed services and applications. The literature on wireless ad hoc broadcast algorithms is quite vast, with many different algorithms being proposed which explore or combine different techniques or features in their operation. While such protocols are becoming increasingly relevant, understanding how they relate among them is complicated. To address this challenge, in this paper, we introduce a novel framework that allows to abstract the operation of wireless ad hoc broadcast protocols. Leveraging on our framework, we explore a particularly interesting class of these protocols: neighbor-aware ad hoc broadcast protocols; of which we propose 4 novel protocols. Finally, we rely on a materialization of our framework to implement prototypes of these protocols and experimentally study their performance in a testbed composed of 21 Raspberry Pi 3 - model B.
KW - Broadcast algorithms
KW - Experimental evaluation
KW - Reliability
KW - Wireless ad hoc
UR - http://www.scopus.com/inward/record.url?scp=85084134153&partnerID=8YFLogxK
U2 - 10.1109/SRDS47363.2019.00033
DO - 10.1109/SRDS47363.2019.00033
M3 - Conference contribution
AN - SCOPUS:85084134153
SN - 978-1-7281-4223-4
T3 - Proceedings of the IEEE Symposium on Reliable Distributed Systems
SP - 231
EP - 240
BT - Proceedings
PB - IEEE Computer Society Press
CY - New Jersey
T2 - 38th IEEE International Symposium on Reliable Distributed Systems, SRDS 2019
Y2 - 1 October 2019 through 4 October 2019
ER -