Скачать или смотреть Simulation and Analysis of Federated learning in Cloud Computing

Title: An End-to-End Security and Privacy Preserving Approach for Multi Cloud Environment Using MultiLevel Federated and Lightweight Deep Learning Assisted Homomorphic Encryption based on AI Technology
---------------------------------------------------------------------------------------------------------------------------------------------------

s/w req:
------------

1)NS-3.26

2)ubuntu-14.04 LTS(32 bit os)

3)Tool: JDK-1.8

    4)IDE: Netbeans-8.2



Execution Steps:
---------------------


Open the ternimal (ctrl + alt + t)

change the directory on terminal.

First execute the files

change the image locations in the main files.

command to execute :

./waf --run filename --vis

Note:
-----------


1) use our module else you get error.

2)Refer how to add our model.txt to add module


Implementation Plan:
-----------------------------------

Step 1: Initially, We create a Network, it consists of 50- Cloud-IoT users, 5- Distributed Edge Servers, 5- Smart Contract Agents (SMAs), 3- Multiple Decentralized Cloud Servers (MDCS) and 1- Super Aggregator (SA).

Step 2: Next, We perform Three-Factor Novel Authentication process, In this process the cloud-IoT users are initiallyregistered to the TAS with their user ID, location, IP address, bio-metrics (i.e. finger veins and eye veins), and picture tag. Once the registration is completed, the TAS provides the Authentication Token (AuthTok) as the symbol of successful registration. During authentication, the cloud-IoT users need to pass the three levels of authentication stage.

2.1: At first stage, the TAS asked the user to verify their user ID and password.

2.2: At second stage, the TAS provides a challenge to the users to select the picture tag and provide appropriate secret sentences that the user provided at that time of registration.

2.3: At the third stage, the TAS asks the user to submit any one of the bio-metrics. For successful authenticated users, the TAS generated a public and private key pair using Improved Key Generation Protocol (IKGP) for encryption and decryption respectively.

Step 3: Next, Secure MLFL Entities Selection process, In this process the optimal DEdS are selected by the TAS using Trading based Evolutionary Game Theory (TEGT) algorithm. For client selection, we used a machine learning algorithm named Improved Artificial Neural Networks (IANN).

Step 4: Next, MLFL based Secure Privacy Aware Homomorphic Data Sharing & Storage, In this process we adopt MLFL the sub-processes involved in the MLFL such as local model, global model, and super model generation & aggregation
respectively.
4.1: Local Model Generation: In this model we used a novel optimized deep learning based homomorphic encryption method named Homomorphic Encryption Responsive Lightweight Residual Network with Energy Valley Optimizer (HER-LResNet-EVO).

4.2: Global Model Aggregation: In this model the normal models are aggregated using lightweight deep learning named Lightweight Factorized Pyramidal Networks (LFPN).

4.3: Super Model Aggregation & Secure Distribution: In this model the securely aggregated super models are provided to the cloud-IoT users via MDCS, and DEdS, and also securely stored in the cloud database.

[The process is based your proposal :- An End-to-End Security and Privacy Preserving Approach for Multi Cloud Environment Using MultiLevel Federated and Lightweight Deep Learning Assisted Homomorphic Encryption based on AI Technology]

Step 5: Finally,we plot the results graph for Number of Cloud-IoT Users Vs Malicious Traffic Rate, Number of Epochs Vs Accuracy, Number of Communication Rounds Vs Accuracy, Size of the Data Vs HE encryption Time, Number of HE Operation Vs Noise Budget and Attack Rate Vs Privacy threats.
=======================================================================================================================

Software Requirements:
--------------------------------------

1) Ns-3.26
2) Java
3) Ubuntu 14.04 [32-bit]

-------------------------------------------------------------------------------------------------------------------------------------------
This video aims to provide guidance towards the resolution of your inquiry
regarding the provision of assistance in selecting a topic for cloud computing
studies.

We have an experienced research team for your cloud computing research
study.

For Research Support contact us:

Our website: https://cloudcomputingprojects.net/

Mail us at: [email protected]

Call us at : +91 94448 69228