weaver

Weaver

This folder contains the implementation of Weaver, a collection protocol leveraging CTX.

Code structure

We implement Weaver for the Decawave DW1000 radio, targeting the EVB1000 board. The code targets Contiki OS.

Weaver exploits SAMU to build a synchronised flood where the sink is the time reference.
The code for SAMU can be found at contiki-uwb/dev/dw1000/samu/.

Energy information is managed by the Statetime module, which is embedded in SAMU. The source code of Statetime can be found in contiki-uwb/dev/dw1000/dw1000-statetime.c.
More information on the Statetime module can be found in the README.

Weaver maps the set of deployed node to a bitmap in which each node is assigned a specific index. At the moment the bitmap is 64bit long.
It is possible to configure the deployed nodes creating a specific folder for your topology and defining the node addresses in contiki-uwb/examples/deployment. For a complete explanation of the required steps, please consult the README.

Weaver code is structured as follows:

• weaver.c contains the Weaver logic and an application that schedules U initiators at every epoch;

• weaver-utility contains operations concerning the bitmap used by Weaver in local and global acknwoledgemnts;

• weaver-log provides convenient functions to collect and print information to be logged at the end of the epoch;

• rrtable is responsible to define a pre-allocated bounded FIFO queue used to store incoming data.

Running experiments

We provide scripts to conveniently build Weaver from a configuration file that allows to define the sink node, the payload to be used and several other parameters.

An example of such a configuration file can be found in experiments/example/params.py. Simulations can be built executing the builder.py script in the same folder of the configuration file:

../../test_tools/builder.py

Simulations will be created in the folder specified in the configuration file. When processing logs, we assume that run logs are within the corresponding simulation folder.

Processing scripts

Test logs are processed with the log_parser2.py script as follow:

test_tools/log_parser2.py test.log

A stats folder is created within the log file parent directory, containing individual csv files related to the perfomance of the protocol and the statistics used for further processing. The summary.csv file provides the performance of the protocol at each epoch, as well as its mean performance (the very last row).

Publications

The following work was published in "One Flood to Route Them All: Ultra-fast Convergecast of Concurrent Flows over UWB":

@inproceedings{10.1145/3384419.3430715,
author = {Trobinger, Matteo and Vecchia, Davide and Lobba, Diego and Istomin, Timofei and Picco, Gian Pietro},
title = {One Flood to Route Them All: Ultra-Fast Convergecast of Concurrent Flows over UWB},
year = {2020},
isbn = {9781450375900},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3384419.3430715},
doi = {10.1145/3384419.3430715},
abstract = {Concurrent transmissions (CTX) enable low latency, high reliability, and energy efficiency. Nevertheless, existing protocols typically exploit CTX via the Glossy system, whose fixed-length network-wide floods are entirely dedicated to disseminating a single packet.In contrast, the system we present here, Weaver, enables concurrent dissemination towards a receiver of different packets from multiple senders in a single, self-terminating, network-wide flood.The protocol is generally applicable to any radio supporting CTX; the prototype targets ultra-wideband (UWB), for which a reference network stack is largely missing. Our modular design separates the low-level mechanics of CTX from their higher-level orchestration in Weaver. Other researchers can easily experiment with alternate designs via our open-source implementation, which includes a reusable component estimating UWB energy consumption.Our analytical model and testbed experiments confirm that Weaver disseminates concurrent flows significantly faster and more efficiently than state-of-the-art Glossy-based protocols while achieving higher reliability and resilience to topology changes.},
booktitle = {Proceedings of the 18th Conference on Embedded Networked Sensor Systems},
pages = {179–191},
numpages = {13},
keywords = {low-power wireless, ultra-wideband, concurrent transmissions},
location = {Virtual Event, Japan},
series = {SenSys '20}
}