Information

This repository contains the code for this workshop presentation: Thanh Le, Yusheng Ji, Thanh-Trung Nguyen, John C.S. Lui , " X-armed Bandits for Optimizing Information Freshness in Robots Communication ", IEICE SeMI Vietnam Workshop, Danang, Oct. 2024.

How to use the code

Main working directory is "core":

cd core

Step 1: create adhoc wifi link

On pi01

python3 main.py --scenario=wifi_off
python3 main.py --scenario=wifi_on --ip=192.168.1.1

On pi02

python3 main.py --scenario=wifi_off
python3 main.py --scenario=wifi_on --ip=192.168.1.2

Step 2: using traffic control command to limit bandwidth

In both rasberry pi, run

python3 main.py --scenario=tc_off
python3 main.py --scenario=tc_on

Step 3: Synchronize time

Installing the "chrony" NTP server is done in a terminal/ssh session via

sudo apt install chrony

and then edit the config file with

sudo vim /etc/chrony/chrony.conf

adding, at the end, these 3 lines

server 127.127.1.0
allow
local

will create a NTP pseudo server and using the internal clock as reference. Restart, after the changes, with

sudo systemctl daemon-reload
sudo systemctl restart chronyd

Check status with

sudo systemctl status chronyd

Configure clients For other RPi:s/devices, acting as clients, in the LAN that has to be synched, edit the config file with

sudo vim /etc/systemd/timesyncd.conf

and after the "[Time]" line just add

NTP=nnn.nn.n.n

which is the IP, or hostname, of your NTP server. Then restart and check with

sudo systemctl daemon-reload
sudo systemctl restart systemd-timesyncd
sudo systemctl status systemd-timesyncd

Step 4: Learn best sampling rate

On pi02

python3 main.py --scenario=receive --receiver=<enter receiver name here> --display_mode=<enter display mode here>

Possible algorithm at the receivers are "greedy/finite_difference/vhct". Possible display modes are "image/aoi".

On pi01

python3 main.py --scenario=send --display_mode=<enter display mode here>

To plot figures:

python3 main.py --scenario=plot_metric --metric=peak_aoi
python3 main.py --scenario=plot_metric --metric=n_packet
python3 main.py --scenario=legend_metric