PROCESS SCHEDULER WEBAPP:

KEY CONCEPTS REPORT

Sathi Surya Prakash Reddy (1RV24IS114), Samarjeet Sujeet Bhonsle (1RV24IS107)
Department of Information Science and Engineering
R V College of Engineering, Bengaluru, Karnataka, India
Faculty In-charge: Swetha S (Assistant Professor)

Abstract—This document outlines the core concepts, algorithms, and logic

implemented in the Process Scheduler WebApp. It serves as a technical basis for
understanding the simulation and analytical capabilities ofthe tool. The system
visualizes CPU scheduling algorithms and introduces a unique "Predictor"module
that recommends optimal scheduling strategies based on workload analysis.

Burst Time: The total CPU time required for the

I. INTRODUCTION process to complete.
Priority: A numerical value indicating importance
The Process Scheduler WebApp is a simulation and
(lower number indicates higher priority in this
analytical tool designed to visualize and evaluate CPU
implementation).
scheduling algorithms. Efficient CPU scheduling is
critical for operating system performance, ensuring fair Additionally, the system tracks dynamic attributes during
resource distribution and minimizing latency. simulation:
This system allows users to define processes with specific Remaining Time: Tracks progress during preemptive
attributes and observe how different algorithms manage scheduling.
their execution. A unique feature of this system is its Waiting Time: Time spent in the ready queue.
"Predictor" module, which analyzes valid workloads and Turnaround Time: Total time from arrival to
recommends the most suitable scheduling algorithm based completion.
on statistical heuristics.
Response Time: Time from arrival until the first time
the CPU is allocated.

Predictor Terminated

exit

New admit Ready dispatch Running

User Input Viz Output
preempt

Scheduler

FIG. 2. PROCESS STATE TRANSITION DIAGRAM

FIG. 1. HIGH-LEVEL TECHNICAL ARCHITECTURE

III. SCHEDULING ALGORITHMS
II. PROCESS REPRESENTATION The system implements four fundamental CPU scheduling
algorithms, covering both preemptive and non-preemptive
At the core of the system is the Process entity. Each
paradigms.
process simulates a task requiring CPU time and is
characterized by the following attributes: [Link] Come First Serve (FCFS)
PID (Process ID): Unique identifier. Type: Non-Preemptive
Arrival Time: The timestamp when the process enters Processes are executed strictly in the order of their arrival.
the ready queue. This is the simplest algorithm to implement but suffers
from the "Convoy Effect" if a long process arrives
first.
The implementation sorts processes by arrival time and
runs them to completion.
Start

[Link] Job First (SJF)

Type: Configurable (Non-Preemptive or Preemptive) Priority Used?
Yes
Suggest Priority
For Non-Preemptive logic, the process with the smallest
Burst Time in the ready queue is selected. For Preemptive No

logic (Shortest Remaining Time First), if a new process

arrives with a Remaining Time shorter than the currently High Variance?
Yes
Suggest SJF

running process, the CPU is preempted. This algorithm

minimizes average waiting time. No

[Link] Scheduling Interactive?

Yes
Suggest RR

Type: Configurable (Non-Preemptive or Preemptive) No

The CPU is assigned to the process with the highest Suggest FCFS
priority. In the non-preemptive mode, the current process
finishes before priorities are checked. In preemptive
mode, a higher priority arrival interrupts execution. Note
FIG. 5. PREDICTION HEURISTICS FLOW
that this implementation does not currently use "Aging" to
prevent starvation.
The heuristics employed include:
[Link] Robin (RR)
1. Burst Time Variance: High variance suggests SJF to
Type: Preemptive prevent the convoy effect, while low variance implies
Designed for time-sharing systems, each process is FCFS is sufficient.
assigned a fixed time unit called a Time Quantum. A 2. Arrival Time Spread: A small spread implies batch
circular queue is used; if a process's burst time exceeds processing (FCFS), whereas a large spread implies an
the quantum, it is interrupted and placed at the back of the interactive workload suitable for Round Robin.
queue. This is innovative for interactive systems as it 3. Variance & Preemption: High variance combined
guarantees responsiveness. with scattered arrivals triggers a recommendation for
Preemptive modes.
4. Load Size: Large numbers of processes with varied
Start RR
arrivals favor Round Robin for fairness.
Initialize
Ready Queue V. PERFORMANCE METRICS
Yes Queue
The application evaluates schedules based on standard
End Empty?
Operating System metrics to quantify efficiency.
No
Avg Waiting Time = Σ(Waiting Time) / N
Execute for
Time Quantum
Avg Turnaround = Σ(Turnaround Time) / N
No Add to
Done? Queue End Avg Response Time = Σ(Response Time) / N
Where N is the total number of processes. The system also
generates a Gantt Chart, a visual timeline list of
FIG. 3. ROUND ROBIN ALGORITHM FLOW
(Process ID, Start Time, End Time) tuples,
used to visualize the exact execution order and context
switches.

P1 P2 P3 P1 VI. TECHNICAL ARCHITECTURE

0 3 5 9 10 The system is built on a modular architecture to ensure
scalability and maintainability.

Backend: Python (Flask) serves as the core runtime.

FIG. 4. EXAMPLE GANTT CHART VISUALIZATION (RR)
Logic Layer: The ProcessScheduler class

IV. INTELLIGENT ALGORITHM encapsulates all algorithms and prediction logic,

PREDICTION ensuring separation of concerns from the API.
API: RESTful endpoints ( /api/predict ,
A standout feature is the predict_best_algorithm
/api/schedule ) accept JSON payloads and return
method. Unlike standard simulators, this system analyzes
the dataset to recommend an algorithm. calculated schedules and metrics.
 VIII. CONCLUSION
Algorithm Performance Comparison
The Process Scheduler WebApp successfully
demonstrates the implementation of fundamental CPU
scheduling algorithms with an innovative predictive
recommendation system. The integration of statistical
Avg Wait Time
heuristics to suggest optimal algorithms represents a
significant enhancement over traditional scheduling
simulators.

Future enhancements could include the implementation of

FCFS SJF Priority RR
aging mechanisms to prevent starvation in priority
scheduling, adaptive time quantum adjustment for Round
Robin, and machine learning-based prediction models
FIG. 6. RELATIVE ALGORITHM PERFORMANCE (SAMPLE trained on diverse workload patterns. The modular
WORKLOAD)
architecture facilitates easy extension and integration of
additional scheduling policies.
VII. IMPLEMENTATION DETAILS
IX. REFERENCES
The implementation uses object-oriented principles where
each algorithm is encapsulated as a method within the [1]Y. Cao, et al., "Hybrid Heuristic and Reinforcement Learning
ProcessScheduler class. The predictor module for Heterogeneous Cluster Scheduling," Journal of Parallel
and Distributed Computing, 2025.
employs statistical analysis including calculation of
variance, standard deviation, and distribution metrics to [2] X. Wang, et al., "INSPIRIT: AI-Aware Scheduling for
Heterogeneous Cores," IEEE Transactions on Computers,
determine workload characteristics.
2024.
The frontend interface is built using modern web [3] Z. Li, et al., "Adaptive Time Quantum: A Reinforcement
technologies, providing an intuitive user experience for Learning Approach," in Proceedings of the 2024
defining process parameters, selecting algorithms, and International Conference on Operating Systems, 2024.
visualizing results through interactive Gantt charts and [4]A. Kumar, "Adaptive Deadline-Based Scheduling (ADBS)
performance metrics dashboards. for Dynamic Systems," International Journal of Advanced
Computer Science and Applications, 2025.
[5]S. Chen, "Genetic Algorithms for Real-Time Task
Scheduling in Cloud-Based AI Systems," Cloud Computing
Advances, 2025.