CurvingLoRa_NSDI22
1.0.0
该存储库包含脚本和指令,用于在我们的NSDI'22纸上复制实验“ Curvinglora,通过并发传输来提高Lora网络容量”。
必需的软件:
运行以下命令从Github下载Curvinglora。
git clone https://github.com/liecn/CurvingLoRa_NSDI22.git
cd CurvingLoRa_NSDI22
cp config_example.m config.m
# update the HOME_DIR in the config.m
Repo Root
+-- 0_demo # A toy example for non-linear chirp generation.
+-- 1_observation # Fig 5(a)-(d), Fig 7(a)-(d).
+-- 2_simulation # Fig 6(a)-(b), Fig 8(a)-(b).
+-- 3_deployment # Evalution Part
+-- symbol_emulation # Fig 16(a)-(b), Fig 17(a)-(d), Fig 18(a)-(d).
+-- outdoor_emulation # Fig 20(a)
+-- result # Results
+-- transmitter # Matlab scripts for packet generation
+-- data # Dataset
+-- symbol # Indoor symbol dataset
+-- outdoor # Outdoor dataset
+-- groundtruth # Groundtruth for the outdoor dataset
+-- utils # Utility functions
+-- figs # Some figures in the paper
+-- config_example.m # Configuration template
我们在核心目录result/下提供性能结果。为了重现论文中的数字,您可以在Python脚本中运行以下命令。
cd CurvingLoRa_NSDI22
matlab -nodisplay %% Matlab
addpath(genpath( ' ./. ' ));
%% Observation Results
1_observation / fig_5 # Fig 5
1_observation / fig_7 # Fig 7
%% Simulation Results
2_simulation / fig_6a # Fig 5
2_simulation / fig_6b # Fig 7
2_simulation / fig_8a # Fig 8a
2_simulation / fig_8b # Fig 8b
2_simulation / fig_sir2map # SIR map
%% Evaluation Results
3_deployment / symbol_emulation / fig_16a # Fig 16a
3_deployment / symbol_emulation / fig_16b # Fig 16b
3_deployment / symbol_emulation / fig_17abcd # Fig 17abcd
3_deployment / symbol_emulation / fig_18abcd # Fig 18abcd
3_deployment / symbol_emulation / fig_17abcd # Fig 20a
3_deployment / outdoor_emulation / figs_outdoor_emulation # Fig 20a请确保配置中的所有路径对于数据集,脚本和日志都一致。
下载数据集。请下载符号和室外数据集,并将它们放在data/下,如上所述目录树所示。
请在每个目录下运行EVA _ {***}脚本,以从头开始复制结果。
如果您使用研究项目中的代码或数据,请考虑引用我们的论文。
@inproceedings { CurvingLoRa_NSDI22 ,
author = { Li, Chenning and Guo, Xiuzhen and Shuangguan, Longfei and Cao, Zhichao and Jamieson, Kyle } ,
title = { CurvingLoRa to Boost LoRa Network Throughput via Concurrent Transmission } ,
year = { 2022 } ,
booktitle = { Proceedings of USENIX NSDI } ,
}[email protected] chenning li