Simple Wi-Fi Optimizer

A simple Wi-Fi connection optimizer based on the geolocation. The connection's race condition can be resolved by matchmaking between the APs and the nodes. The script runs on the nodes, because the AP's commercial loadbalancing features are nonfree and not a standard.

It's useful for highly congested environment such as Office with AI Training. Currently only the static loadbalancing is supported. Dynamic loadbalancing is in progress.

Requirements

sources.csv

kind,id,x,y
desktop,(my-machine-uuid),0,0
desktop,f1ba76c3-6c5b-4f19-bf8c-f71363052a9f,2,3

targets.csv

kind,id,x,y
ap,(my-ap-mac-address),0,0
ap,12:34:56:78:9A:BC,3,4

Test on the Local Machine

With the docker

sudo docker run --privileged \
  --env DEBUG=true \
  --volume /run/dbus:/run/dbus:ro \
  --volume $(pwd):/src:ro \
  quay.io/ulagbulag/wifi-optimizer:latest

Environment Variables

• DEBUG: Whether to show debug logs (default: false)
• SRC_FILE: The sources.csv file path (default: sources.csv)
• TGT_FILE: The targets.csv file path (default: targets.csv)
• DRY_RUN: Whether to change BSSID virtaully (default: false)
• INTERVAL_SECS: The interval of updating BSSID as seconds (default: 30)

Deploy on the K8S

# Please edit examples/kubernetes/configmap.yaml to your own configuration

kubectl apply -f examples/kubernetes/namespace.yaml
kubectl apply -f examples/kubernetes/configmap.yaml
kubectl apply -f examples/kubernetes/daemonset.yaml

License

Please check our LICENSE file.