Weighted Fair Queuing Algorithm

Overview

Weighted Fair Queuing (WFQ) is a packet scheduling algorithm used in computer networking to achieve fairness and improve Quality of Service (QoS) in the transmission of data packets. WFQ ensures that each flow gets a fair share of the network resources while allowing for differentiated services based on assigned weights.

Key Concepts

Fairness: WFQ ensures that every flow, regardless of its weight or packet size, receives an equitable share of network resources over time.
Weighted Scheduling: The algorithm prioritizes packets based on predefined weights assigned to different flows or packet streams.
Virtual Time: WFQ uses the concept of virtual time to schedule packets, granting priority to flows with a higher ratio of data to weight.

How It Works

Packet Arrival: As packets arrive, WFQ establishes a virtual finish time based on the current time and the weight of the flow.
Queueing: Packets are placed into separate queues for each flow and are scheduled for transmission based on their virtual finish times.
Scheduling: The scheduler selects the packet with the earliest virtual finish time for transmission, ensuring fairness and prioritization based on the assigned weights.

How to run code

Configurations

router_port : its a port router use to listen the sources
destination_port : its a port destination use to listen the routers
source_amount : amount of sources
destination_port : its a port destination to used to listen the routers
csv_address : the csv files contain data want to send to the destination
result_address : the address of result.csv to see the order of packets them received to the destination

below you can see the example of config.ini file in root directory of project :

[CONFIGS]
router_port=9006
destination_port=9004
source_amount=3
router_amount=1
csv_address=/home/y/Desktop/iust/AdvanceNetwork/WeightedFairQueueing/client/
result_address=/home/y/Desktop/iust/AdvanceNetwork/WeightedFairQueueing/destination/

Starter Data

You have to provide data to client in a structure like below csv file in the csv_address directory that you provide later

csv files : this is the structure of named csv file data_<name>_<weight>.csv the name can be any thing but weight must be an integer number that is the sign of weight of data send to destination from a client each csv file is a represent of a client that want to send contained data to destination
data in csv files : below example of data in a csv files contian packet which is name of the packet and data is the data a packet contained and start_time is the time that a packet send to router
```
packet, data, send_time
B, sdjglr, 5
D, kdjgniegl, 8
H, gjlrogdi, 20
```

Run code

Manual

step one : after set your config.ini file first you have to run destiantion.py file in the destiantion module
step two : in this step you have to run router.py file in the router module
step two : in final step you have to run client.py file in the client module to start the flow

Automate

in windows : after make sure you have install python on your windows os just run runner.bat in cmd like .\runner.bat or open gitbash and make sure python3 is install on it then run bash runner.sh
in ubuntu : after make sure python3 is install on your local ubuntu just run bash runner.sh

Result

You can see the result in a csv file that you set the directory address of it in the config.ini file later in a file named result.csv and you can see the example of output down below .

data receive in destination => packet : A data length : 8 , data : adhvngld weight : 1 at 1707651170.1469297
data receive in destination => packet : C data length : 10 , data : djflotbbhg weight : 2 at 1707651172.1142294
data receive in destination => packet : B data length : 6 , data : sdjglr weight : 1 at 1707651173.1036284
data receive in destination => packet : E data length : 8 , data : kfogjelg weight : 2 at 1707651176.1178179
data receive in destination => packet : F data length : 6 , data : kdjbit weight : 1 at 1707651178.1129177
data receive in destination => packet : G data length : 10 , data : kfldhgirfr weight : 2 at 1707651181.13731
data receive in destination => packet : D data length : 9 , data : kdjgniegl weight : 1 at 1707651182.1453245
data receive in destination => packet : H data length : 8 , data : gjlrogdi weight : 1 at 1707651190.191596

Advantages

Fairness: All flows receive an equitable share of network capacity, preventing certain flows from monopolizing resources.
Quality of Service: WFQ can provide improved QoS for delay-sensitive or high-priority traffic by using weighted scheduling.
Differentiated Services: The algorithm allows for varying levels of service based on assigned weights, making it suitable for diverse traffic types.

Example

Consider a network with three flows having weights 1, 2, and 3, respectively. In a WFQ system, the flow with weight 3 will receive twice the bandwidth as the flow with weight 1 and thrice the bandwidth of the flow with weight 2.

Implementation

WFQ is commonly implemented in network devices such as routers and switches as part of their Quality of Service (QoS) mechanisms.
Various networking libraries and protocols, such as DiffServ and MPLS, may incorporate WFQ to achieve differentiated service levels.

Contributing

Contributions and feedback on the WFQ algorithm and its applications in computer networking are welcome.

License

This project is licensed under the MIT License.

aminhzdev / weightedfairqueueing Goto Github PK

weightedfairqueueing's Introduction