DCTPQ: Dynamic Cloud Gaming Traffic Prioritization Using Machine Learning and Multi-Queueing for QoE Enhancement
This paper has been submitted to the SBRC 2025 Conference and is under review!
- Install P4Pi [https://github.com/p4lang/p4pi)]
- Python3 [https://www.python.org/downloads/release/python-3123/]
To run the data plane (P4 code), controller, and classifier ...
Note: Ensure the P4Pi's Ethernet interface is connected to a network with DHCP enabled for internet access!
ssh [email protected] # default P4Pi IP address is 192.168.4.1 and Username is `pi'
pass: **** # Password
sudo su # change to root user
cd ~ # home directory
./setup_eth_wlan_bridge.sh
Output:
Note: It is recommended to be disconnected and connect again with the local IP assigned by DHCP (e.g., 200.18.102.14/25 using wireless or 200.18.102.32/25 using wired interface) and after ssh connection and entering the password, change the user to root and go to the home directory as we did! (run the previous instructions except './setup_eth_wlan_bridge.sh')
Note: Copy the DCTPQ to the P4Pi from your computer! These instruction is running on your PC!
# Run on your computer
git clone https://github.com/dcomp-leris/DCTPQ.git
scp -r ./DCTPQ [email protected]:/home/pi/ # Copy the DCTPQ folder to P4Pi
pass: **** # Enter your computer password
# Attention ==> change the IP address (200.18.102.14) to your local network address!
Note: To compile and run, keep it memorized that you copied files "/home/pi/DCTPQ"! if you change the addree be aware of changing the address!
To run our project, it is necessary to have the newest version of BMv2 and P4C, installing them from source.
First of all, unistall the current installed package versions:
sudo apt remove p4lang-bmv2 # It will remove both BMv2 and P4C
To install BMv2 from source, run:
# Clone the official repository
git clone https://github.com/p4lang/behavioral-model.git
# Install dependencies
sudo apt-get install -y automake cmake libgmp-dev \
libpcap-dev libboost-dev libboost-test-dev libboost-program-options-dev \
libboost-system-dev libboost-filesystem-dev libboost-thread-dev \
libevent-dev libtool flex bison pkg-config g++ libssl-dev
# Install BMv2
cd behavioral-model
./autogen.sh
./configure
make
sudo make install
To install P4C from source, run:
cd .. # Leave the BMv2 dir
# Clone the official repository
git clone --recursive https://github.com/p4lang/p4c.git
# Install dependencies
sudo apt-get install cmake g++ git automake libtool libgc-dev bison flex \
libfl-dev libboost-dev libboost-iostreams-dev \
libboost-graph-dev llvm pkg-config python3 python3-pip \
tcpdump
# Install P4C
mkdir build
cd build
cmake .. <optional arguments>
make -j2
make -j2 check
sudo make install
Now you have the newest versions of BMv2 and P4C installed on your computer!
You will need to edit a bash script:
nano /usr/bin/bmv2-start
In the line 19, add "-- --priority-queues 3" in the end, like this:
p4c-bm2-ss -I /usr/share/p4c/p4include --std p4-16 --p4runtime-files ${BM2_WDIR}/bin/${P4_PROG}.p4info.txt -o ${BM2_WDIR}/bin/${P4_PROG}.json ${BM2_WDIR}/examples/${P4_PROG}/${P4_PROG}.p4 -- --priority-queues 3
Save and close the file.
p4c-bm2-ss --target bmv2 --arch v1model -o /home/pi/DCTPQ/P4Pi/DCTPQ.json /home/pi/DCTPQ/P4Pi/DCTPQ.p4 # Compile the code
# Note: Be sure the forward.json was generated in the address you set!
# create the folder in bmv2 to run P4
mkdir bmv2/examples/DCTPQ
cp /home/pi/DCTPQ/P4Pi/DCTPQ.p4 ./bmv2/examples/DCTPQ # P4 file name must be "DCTPQ.p4"
cp /home/pi/DCTPQ/P4Pi/DCTPQ.json ./bmv2/examples/DCTPQ # json file name must be "DCTPQ.json"
# Set the name of the project/folder P4Pi will run
echo "DCTPQ" > /root/t4p4s-switch
# Restart the simple switch
systemctl restart bmv2.service # Run the current P4
systemctl status bmv2.service # Check the running config
Output:
# Classifier
sudo python3 /home/pi/DCTPQ/P4Pi/flow_classifier.py
Output:
Note: The classifier is listenning the ports in parallel with dataplane forwarding and the match/action table is filled by the flow_classifier.py running in the controller!
Note: Run on your computer!
# First -> INT sender
python3 send3.py
# Second -> INT Receiver
python3 receive_Showing.py # To show
python3 receive_logging.py # To store the log
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
DCTPQ is running as your Access Point (AP)!
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Note: If you use this paper, please cite it as
@inproceedings{sbrc,
author = {Alireza Shirmarz and Carlos Henrique de França Marques and Fábio Luciano Verdi and Roberto Silva Netto and Suneet Kumar Singh and Christian Esteve Rothenberg},
title = { DCTPQ: Dynamic Cloud Gaming Traffic Prioritization Using Machine Learning and Multi-Queueing for QoE Enhancement},
booktitle = {Anais do XLIII Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos},
location = {Natal/RN},
year = {2025},
keywords = {},
issn = {2177-9384},
pages = {476--489},
publisher = {SBC},
address = {Porto Alegre, RS, Brasil},
doi = {10.5753/sbrc.2025.6266},
url = {https://sol.sbc.org.br/index.php/sbrc/article/view/35152}
}