Design, Simulation, and Implementation of a Unified Network for Computer Science Department Laboratories

Abstract

This research investigates and compares two network design methods for interconnecting small-scale computer laboratories. The first method utilizes a central switch to connect three access switches, while the second  method relies on three directly interconnected switches without a central control point. Both methods were implemented and simulated using Cisco Packet Tracer and  real machine with practical testing performed using network diagnostic tools, including ping, ping -n 10, and tracert. The results indicate that the first method provides higher performance in terms of stability, latency, and centralized security management, allowing for organized traffic flow and the application of VLANs or Access Control Lists (ACLs). the second method, while simpler and offering partial fault tolerance, exhibits slightly higher latency, fluctuating response times, and limited capacity for enforcing unified security policies. These findings suggest that network architecture should be selected based on the requirements of the environment; centralized designs are preferable for medium-sized educational or organizational networks requiring efficiency, security, and scalability, whereas decentralized designs are suitable for smaller or temporary networks. This study provides practical insights into network design decisions, highlighting the trade-offs between performance, reliability, and manageability.