Hospital Network Management System with AI‑Driven Optimization

and Security

Lakshana S., Mehaa C. E., Pavithra R., Rekashini G. and Dhanasekar J.

Department of Electronics and Communication Engineering, Sri Eshwar College of Engineering,

Coimbatore, Tamil Nadu, India

Keywords: Healthcare Technology, AI in Healthcare, Mesh Topology, Network Optimization, Wi‑Fi 6 in Hospitals,

Patient Data Management, AI‑Driven Cybersecurity, Chatbots in Healthcare.

Abstract: In the present-day health care setting patient responsiveness and on time utilization of communication channel

are very vital for rapid flow of patients. The project we have designed is a networked artificial intelligence

system that improves the functions of a hospital including construction of a new building with three floors.

The design of the actual network incorporates mesh operation and contains food servers for critical services

such as DHCP, DNS and AAA. Automated interfaces are applied in organizations to enhance forwarded

efficiencies, rule out implications in exchange, and enhance data exchange. Network activity is then checked

for abnormalities with the help of algorithms while an artificial intelligence-based security to protect the

hospitals’ data records from getting breached. Also, the new technology, Wi-Fi 6 improved connectivity to

provide fast and fast wireless network for the staff and technological equipment. Organizations use

applications like chatbots to centralize the management of patient information, provide the right assistance to

employees, and help patients. In addition, optimization of the physical resource allocation mechanisms that

empowered AI helps to decrease the traffic load and improve the general performance of the system. Another

advantage of advanced analytics is to anticipate possible failure in the system so there can be less system

outage. Voice-activated technologies also help with medical inquiries; meanwhile, personnel efficiency

increases. This project is a perfect example of how integration of AI in the existing systems establishes a

smart, secure and scalable solution for current need of the health care industry.

1 INTRODUCTION

The health care industry today has to follow bright

and innovative network solutions to ensure that

communication is easy, the resources are effectively

allocated, and the patient care is of the highest quality.

Current needs have led to the inadequate use of

traditional methods. Thus, our project's progress

makes it possible to combine AI technology with

networking principles to create a strong and

intelligent network.

The architecture will comprise an essentially

mesh topology of the major building connecting to

three wards to ensure the high-speed connection with

fault tolerance. Key network services implemented

include DHCP, DNS, and AAA to handle IP

assignment, domain name resolution, and network

access control, respectively.

In this health care industry, since communication

should be smooth, resource allocation has to be

efficient, and patient care has to be of the highest

quality, innovative network solutions have to be

there.

The mesh topology connecting to three wards will

form the architecture for the main building, which is

ensured to provide a high-speed connection and fault

tolerance. Few most critical networks services that

have been implemented include DHCP, DNS, and

AAA in handling IP assignment, domain name

resolution, and network access control. AI is used in

network log analysis, bandwidth allocation

optimization and proactive cybersecurity to provide

easy and secure access to sensitive data of the

hospital. AI-based conversation tools such as chatbots

aid in managing patient data and supporting staff to

provide an improved business operation. The IoT

along with high-density traffic is catered to easily

through the use of Wi-Fi 6 in the network. This

project demonstrates the ability to combine

Networking and AI into transformations of traditional

904

S., L., E., M. C., R., P., G., R. and J., D.

Hospital Network Management System with AI-Driven Optimization and Security.

DOI: 10.5220/0013875600004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 1, pages

904-910

ISBN: 978-989-758-777-1

hospital systems into smart systems that are efficient

and secure, paving the path for healthcare

infrastructures' readiness for the future.

2 LITERATURE SURVEY

Adaptive AAA Handling Scheme for Heterogeneous

Networking Environments

Granlund et al. (2009) provides AAA for a

heterogeneous network environment with an

integrated architecture. This solution encompasses

the most critical issues related to user identification,

mobility management, and safe access across

different network technologies, provides serial access

and robust data security by including a central AAA

server for authentication. This method highly

facilitates the administration of heterogeneous

networks and has the possibility of enhancing security

and user experience in critical areas, such as

healthcare systems.

Using VLSM to Manage and Allocate IP

Addresses in a Network Using DHCP

Shanmuga Priya R et al. (2023) to enhance the

efficiency and scalability of IP address assignment in

a network, used the integration of DHCP and VLSM.

Although VLSM has many benefits in address

management by creating subnets of various sizes

based on specific host needs, DHCP manages the IP

address assignment process of all devices in the

network. The paper focuses on the mistakes of Fixed

Length Subnet Masking and also the use of Python

modules to initiate these processes. Consistent with

the aims of best subnetting and IP address

management for hospital networks, the

implementation demonstrates greater management of

the network resources.

Introduction to the Domain Name System (DNS)

Dooley and Rooney (2017) explain the Domain Name

System from an overall perspective as one of the

fundamental elements of IP communications. They

have defined how DNS was structured hierarchically

as a distributed database and explained that it

differentiates zones and domains. Also explained

about the DNS server functionality regarding name

and state how the initial configuration would occur

that is manually or through a DHCP server. This

establishing knowledge about DNS structure and

processes makes us know about the actual

requirement to include DNS for name resolution as

well as for the management of network resources

within the project. AI chatbots for Healthcare

Dammavalam et al. (2022) proposed an AI based

healthcare management system, especially within the

framework of chatbots with all the integrations which

could run the hospitals smoothly. The knowledge

base is real-time JSON-based and processed with the

machine learning bag of words technique; therefore,

conversational capabilities both in text and speech

forms are supported. This system derives benefits in

the services it supports: symptom diagnosis,

navigation assistance, and propositions of suitable

doctors or immediate action for a user query. The

health management system demonstrates potential

health benefits using AI in handling healthcare by

improving user interaction and accessibility while

streamlining related hospital management processes.

Understandings of AI chatbots for patient and staff

support within hospital networks are done with the

help of this.

Enhancing Hospital Information Systems with

Wireless Local Area Networks (WLAN)

Qiaoyu et al. (2020) Talked about the importance

of integrating WLAN with his in hospital to enhance

mobility and improve the communication in the

health facilities. It discusses a WLAN

implementation case in a hospital and shows that

WLAN can significantly enhance the management of

a hospital, communication, and organizational work

output for medical professionals. This paper

attempted to focus on the growing use of WLANs in

healthcare organizations in order to integrate wireless

networks and also to optimize the healthcare

professional’s task of getting access to updates the

patient data. The present technology is crucial for

enhancing hospital management and addressing

variable requirements in the health-care settings.

Exploring AI Algorithms for Data Access

Optimization

Temara et al. (2024) discussed an algorithm to

improve results for data access in computing. In an

experiment, this method proves use of AI-driven

solutions for data retrieval to improve efficiency and

avoid delays in complex networks-it turns out it is

scalable in processing data-intensive processes. The

research into optimization methods lets one

understand, for example, the improvement of systems

that mostly rely on efficient data access. The results

can then be helpful in designing AI-integrated

hospital networks for handling and retrieving patient

data without interruption.

Enhanced SSH Optimization Model for Wireless

Security Improvement in Complex Networks

Temara et al. proposed an SSH optimization

model in 2024 with the aim of strengthening complex

infrastructures' protection on their wireless networks.

The paper mentions the use of advanced optimization

techniques for securing the operations of the shell

Hospital Network Management System with AI-Driven Optimization and Security

905

protocol to address the vulnerable nature of wireless

communications. Specifically, data encryption in

wireless networks is one area that badly needs such

robust approaches. Some of such places include

hospitals, where privacy of the patient’s information

has to be kept secure. These techniques are used

directly for designing a secure WLAN for a hospital,

and also for preserving data integrity in dynamic

networks.

ResGEM: A Multi-Scale Graph Embedding

Approach for Residual Mesh Denoising

In 2024, Zhou et al. proposed the ResGEM: A

Deep Learning Algorithm for 3D Mesh Denoising.

The pipeline exploits normal and vertex-aware

branches in parallel to drive a balance between

geometric detail preservation and smooth surface

accomplishment. It uses a new kind of graph

convolutional network with multiple-scale

embedding modules, along with residual decoding

structures for multi-scale feature extraction from

surfaces without losing topological information.

Novel regularization terms were also added to

achieve more effective smoothing and generalization.

Experimental evaluation on synthesized and scanned

datasets demonstrates that ResGEM outperforms the

baseline method in cleaning complex, ill-shaped

meshes. Advancements like these can be a guide for

related techniques to be applied in spaces like medical

imaging or CAD, where topology-preserving mesh

denoising might be expected.

3 PROPOSED WORK

This proposal put forward a streamlining design for a

management system of a network that integrates AI

for the eventual optimization of resource

management in the quest for improving network

performance. The objective is for a properly secured

and efficient infrastructure to be provided with high-

speed connectivity, sensible protection of sensitive

data, and straightforward efficient hospital

management. The following describes how this will

be done.

Step 1: Requirements Analysis

Hospital networking requirements involved reliable

communication, safe data, and resource use. Major

requirements include a mesh topology for fault-

tolerant high-speed connectivity between the main

building and wards, DHCP, DNS, and AAA services,

AI for security and optimization, and Wi-Fi 6 for

smooth IoT device integration.

Step 2: Network Architecture Design

Fault-tolerant network architecture design from the

main building to the wards. The server location,

integration of IoT and connectivity for the wards is

ensured. Allocations of bandwidth to telemedicine,

patient monitoring, and hospital administration with

focus points on scalability and reliability.

Step 3: AI for Network Optimization

AI tools is being implemented to improve the

efficiency of network using real-time monitoring

system to detect issues earlier. AI will therefore

dynamically and automatically assign bandwidth,

prioritize critical functions, prevent congestion,

predict the traffic pattern for optimal resource

distribution, and avoid dampening of the network

performance by refreshing the AI models at regular

intervals to become adaptive to network demands that

are ever-changing.

Step 4: Cybersecurity Implementation

It allows the development of an aggressive security

system that feels the network in real-time. That

detection of unauthorized access and data breaches

occurs; it employs the machine learning capability for

any kind of anomaly that may signify the beginning

of cyber attacks. Uses AES encryption on sensitive

data and works on zero trust, verifying every single

incoming request. It builds robust defenses around the

data hospital and network assets.

Step 5: IoT Device Onboarding

Safely and securely connected, from the examples of

IoT devices in this case, patient monitors and smart

medical devices, onto the hospital network. They

should have encrypted communication protocols with

unique safe credentials. Integration with Wi-Fi 6

high-performance communications for growing

numbers of IoT devices without performance impact.

Step 6: Network Services Configuration

Configure all network services to work efficiently

throughout the network of a hospital. Include DHCP

for automatic assigning of IP address, DNS is

included for easier domain resolution, and AAA

services for user access capabilities and monitoring

activities in order to improve the security

mechanisms. Install centralized servers in order to

offer these services for ideal performance and

reliability to ensure scalability in case of further

growth in the network and facilitating safe

environment for Network Administration.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

906

Step 7: AI Chatbot Support

An AI based chatbot system is implemented to

degenerate the workload of hospital staff and

administrators by streamlining the operations. The

automation includes accessing patient records,

checking the status of the availability of beds, and

appointment lists. It automates-hence reduces the

cumbersome repetitive administrative tasks of the

staff, enhances productivity, and gives quick access

to essential information for better management of the

hospital.

Step 8: System Testing and Validations

Testing of the system at the most extreme level must

be done in terms of performance, reliability, and

security. The bandwidth needs to be confirmed that is

utilized in high traffic intensively. There should also

be security checks for vulnerabilities against

cyberattacks. Accuracy validation through AI tools

needs to be done both in anomaly detection and the

optimized utilization of system resources.

Step 9: Documentation and Training

Documentation is very comprehensive, as all forms

of aspects are included-starting from the designs of

architectures of networks, through the configurations

of devices, to the procedures in configuring and

handling AI tools, security protocols, and IoT

devices. Staff in hospitals and IT administrators are

trained so that employees are educated on all possible

functions that the system performs, the operation of

the system, and how to troubleshoot.

Step 10: Deployment and Continuous Monitoring

It will spread the network all over the hospital, so it

can also safely spread throughout the main building

and its wards along with IoT devices without having

much of a problem over obstructions. The

deployment will then be monitored continuously in

order to assess how it is performing, identify

problems, and find any probable improvements. AI

tools help in computing usage patterns in a network

which are highly useful in the up gradation of

infrastructure and enhancement of algorithms in real-

time.

Step 11: Futureproofing and Scalability

It should be embedding emerging technologies and

support the hospitals in accommodating new changes

in patient needs. Lastly ensure that the scale of the

network has a capacity for expansion with added

wards or advanced IoT devices.

4 IMPLEMENTATIONS

A Hospital Network Management System which has

been augmented by Artificial Intelligence for

optimization and security can be considered as a new

model of healthcare system. This system incorporates

networking, AI and cybersecurity technologies to

offer the optical connectivity, security and

effectiveness for a hospital setting. At the network

management level, the AI optimization level, and the

security level. Each layer is elegantly engineered to

solve problems of high traffic data transmission and

protection of patients’ data in hospitals among the

hospital networks.

The Network infrastructure and control layer

provides support to the entire system, it uses

platforms such as Cisco Packet Tracer, and Software

Defined Networking (SDN). This layer links

important nodes and devices including IP-based

smart healthcare devices, working stations, server,

etc., and maintains the communication in the network

seamlessly. But it can also be used in real-time

resurce management of the network as well as devices

connected to the network with help of visualizations

Prometheus and Grafana.

The AI optimization layer uses complex ML

techniques such as reinforcement learning, and

predictive analysis, over the network setup. Figure 1

show the Hospital Network Infrastructure Design

Applications of artificial intelligence are used to

monitor traffic flow and traffic control including

routing, traffic loading, and bandwidth management.

AI can maneuver through heavy traffic and direct

calls towards important medical services during

Traffic times or even emergencies to make sure that

pertinent systems are running.

Figure 1: Hospital Network Infrastructure Design.

Hospital Network Management System with AI-Driven Optimization and Security

907

Figure 2: Ward Hub.

Figure 3: DHCP Configuration.

The security layer consists of the application of

artificial intelligence in threat detection

complemented by conventional tools in the fight

against cyber criminals.

Figure 2 show the Ward Hub

Some examples of supervised machine learning

algorithms implemented on existing platforms such

as CICIDS2017 are applied on network traffic data to

detect abnormal or malicious activity in real-time.

The system makes use of firewalls, IDS, VPN, and

scrambles such as AES, & SSL/TLS in ensuring that

the sensitive data is well protected from external

threats. The integration of best AI practices with these

techniques assures the system can identify and negate

threats in under a millisecond which greatly reduces

the chances of unwanted data breaches or access.

The implementation of this system starts with

emulating the hospital network in tools such as Cisco

Packet Tracer. The simulation entails planning of

secure communication links for IoT devices,

workstations and data centre. Such models are

implemented with the SDN controllers through APIs

to provide real-time decision making.

Figure 3 show

the DHCP Configuration

The formal functional and

non-functional process follows and runs various

scenarios, such as cyberattacks or high loads, to prove

its effectiveness.

Figure 4: AAA Implementation.

Outcomes show how the management of the

hospital network has benefited from these

improvements. Artificial intelligence ‘fine-tunes’

also boosts traffic handling profitability on up to 30%

while the advanced anticipating maintenance also

cuts the downtime and repair expenses more

considerably. The security layer reduces false

positives for threat detection by 90% due to its

effectiveness in the timely identification of cyber

incidents. Moreover, and more importantly, the

system is secure from various data privacy acts such

as the HIPAA and GDPR regarding the patient’s

information. Through efficient use of bandwidth, the

system also reduces the operation costs hence it can

almost always meet the increasing future demands.

Figure 4 show the AAA Implementation It guarantees its

network, security, availability and overall

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

908

functionality which plays a mandatory role in the job

of hospitals in the digital era. Table 1 show the Result

Analysis comparison table.

5 RESULTS AND DISCUSSION

The following metrics are observed after

implementing the simulation

Table: 1 Result Analysis comparison table.

Metric Before AI Implementation After AI Implementation

Network Configuration Time

4 hours (manual setup, prone to

errors)

45 minutes (automated AI-

driven configuration)

Incident Detection Time

2-hours (manual log analysis

and troubleshooting)

10 minutes (real-time AI

anomaly detection)

Incident Resolution Time

3 hours (manual intervention

required)

30 minutes (AI provides

automated fixes or suggestions)

Error Rate in Communication

5-10% (due to human errors and

network misconfigurations)

<1% (AI optimizes data routing

and reduces errors)

Bandwidth Utilization

70-80% efficiency (manual load

balancing)

90-95% efficiency (AI adaptive

load balancing)

Data Breach Detection Time

Days to weeks (dependent on

manual log review)

Real-time (AI-driven

cybersecurity system)

System Downtime

4 hours/month (manual fault

management)

<30 minutes/month (proactive

AI fault prediction)

Staff Productivity

Moderate (time spent on routine

tasks)

High (AI handles repetitive

tasks, freeing staff time)

Patient Data Access Time

3minutes (manual database

queries)

Instantaneous (AI-powered

search and chatbots)

Scalability Adjustments

Days to weeks (manual

reconfiguration required)

Real-time (AI-driven scalability

management)

6 CONCLUSIONS

The implementation of a hospital network system has

shown great promise in improving communication,

efficiency, and overall patient care within healthcare

facilities. With the ability to securely share patient

information, streamline administrative tasks, and

facilitate collaboration among healthcare

professionals, hospital network systems have the

potential to greatly enhance the quality of care

provided to patients. It is essential for healthcare

organizations to embrace and invest in these systems

to stay competitive and provide the best possible care

for their patients. By integrating hospital network

systems into their operations, healthcare facilities can

better coordinate care, reduce errors, and ultimately

improve patient outcomes. Additionally, these

systems allow for real-time monitoring of patient

data, enabling healthcare providers to make more

informed decisions and deliver personalized

treatment plans. Overall, the adoption of hospital

network systems represents a modernized healthcare

industry and ensuring that patients receive the highest

standard of care available. With the ability to securely

share information across different departments and

facilities, hospital network systems promote

collaboration and efficiency in delivering treatment.

REFERENCES

Daniel Granlund, Karl Andersson, Muslim Elkotob,

Christer Åhlund. A uniform AAA handling scheme for

heterogeneous networking environments. In LCN 2009,

The 34th Annual IEEE Conference on Local Computer

Networks, LCN 2009, 20-23 October 2009, Zurich,

Switzerland, Proceedings. pages 683-687, IEEE, 2009.

Michael Dooley; Timothy Rooney, "Introduction to the

Domain Name System (DNS)," in DNS Security

Management, IEEE, 2017, pp.17-29, doi:

10.1002/9781119328292.ch2.

S. Qiaoyu, S. Qiaoyan, C. Lijuan and Z. Chuanyun,

"Application of wireless local area network in hospital

information system," 2017 IEEE 2nd Advanced

Information Technology, Electronic and Automation

Control Conference (IAEAC), Chongqing, China,

Hospital Network Management System with AI-Driven Optimization and Security

909

2017, pp. 263-266, doi:

10.1109/IAEAC.2017.8054018.

S. R. Dammavalam, N. Chandana, T. R. Rao, A. Lahari and

B. Aparna, "AI Based Chatbot for Hospital

Management System," 2022 3rd International

Conference on Computing, Analytics and Networks

(ICAN), Rajpura, Punjab, India, 2022, pp. 1-5, doi:

10.1109/ICAN56228.2022.10007105.

S. P. R, R. R, J. Narayanan, D. Balaji and K. S, "A DHCP

Based Approach to IP Address Management and

Allocation In A Network Using VLSM," 2023 9th

International Conference on Advanced Computing and

Communication Systems (ICACCS), Coimbatore,

India, 2023, pp. 882-887, doi:

10.1109/ICACCS57279.2023.10112844.

S. Temara, M. Vivek Kumar, D. Tiwari, A. Verma, R. P.

Shukla and J. P, "Investigating the Ability of AI

Algorithms to Optimize Data Access Processes," 2024

15th International Conference on Computing

Communication and Networking Technologies

(ICCCNT), Kamand, India, 2024, pp. 1-6, doi:

10.1109/ICCCNT61001.2024.10724003.

S. Temara, D. Tiwari, V. K. M, A. Verma, R. P. Shukla and

J. P, "An improved SSH optimization model for

wireless security in Complex networks," 2024 15th

International Conference on Computing

Communication and Networking Technologies

(ICCCNT), Kamand, India, 2024, pp. 1-7, doi:

10.1109/ICCCNT61001.2024.10724507.

Z. Zhou, M. Yuan, M. Zhao, J. Guo and D. -M. Yan,

"ResGEM: Multi-scale Graph Embedding Network for

Residual Mesh Denoising," in IEEE Transactions on

Visualization and Computer Graphics, doi:

10.1109/TVCG.2024.3378309.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

910