Federated Learning Insider Threat Detection Research Topics

A federated Learning (FL) Insider Threat Detection research topic is now widely utilized in many applications. FL is one of the ways to train the Artificial Intelligence (AI) models and it is proposed in this research. In this we offer several information related to our proposed FL:

Define Federated Learning

At the initial stage we begin with a definition for FL. It is a machine learning method which practiced a method through distinct session, every one utilizing its own dataset.

What is Federated Learning?

Then at the next stage we overview look for a deep understanding of FL. It is the way to train AI techniques no one can look or touch your data, providing a path to release data to provide new AI applications.

Where Federated Learning used?

Thereafter the deep understanding of FL, we discuss where to utilize that technique. It is extensively used in places like Retail, IoT devices, Mobile Devices, Autonomous Vehicles, Healthcare, Finance and Edge Computing.

Why Federated Learning Technology proposed? , Previous technology issues

The FL technology is proposed in this research to disperse the system training, executing actual-time monitoring, maintaining data secrecy among sources and exact threat identification mechanisms to find and reduce insider threats efficiently through IoT devices. Some of the previous technology issues that it handles are Delay in Threat Detection, Privacy and Timeliness Challenges, Communication Overhead in Aggregation and Scalability and Latency Challenges.

Algorithms / Protocols

Our proposed FL technique employs the following methods to overcome the existing technology issues. The methods that we utilize are Ordering Points to Identify the Clustering Structure Algorithm, Session token (Secure Hash Algorithm), Stellar Consensus Protocol and Hybrid Rivest-shamir-Adleman with Elliptic Curve Digital Signature Algorithm.

Comparative study / Analysis

In this research the FL technique is proposed to address the existing technology issues and here we compare various methods to be proposed in this research to obtain a best finding.

To decrease the overhead and perform increased security over dynamic effect by using the Hybrid Rivest-Shamir-Adleman with an Elliptic Curve Digital Signature Algorithm (RSA-ECDSA).
For best accessibility and security, the security queries validate the users and then the outcome data is saved in the blockchain using hashing (Stellar Consensus Protocol). This verification improves the safety of the network by opposing unwanted users.
Using the Hybrid Ordering Points to Identify the Clustering Structure (OPTICS) with Centroid Refinement technique for node clustering which offers a distinct approach on cluster center points and reevaluate cluster centroids to confirm that they exactly denote data point central advancement.
Session token (Hybrid Hash-based Message Authentication Code with Secure Hash Algorithm) is utilized to improve the protection.

Simulation results / Parameters

We propose a FL technique for this research to enhance the security for threat identification. Here we compare several performance metrics to obtain the best findings. The methods that we compared are Communication rounds and Number of IoT devices with Accuracy and the Number of IoT devices with throughput and the batch size with node number and the trust value with the time period and then the test with performance.

Dataset LINKS / Important URL

Here we give some important links on the basis of our proposed research FL. These links are useful when we overview the concepts, explanations, and the details about this study or research.

Federated Learning Applications

For our proposed technique FL, we provide some of the most commonly used applications. The applications that we employed for this research are Smart Cities, Emergency Response, Smart Agriculture, Environmental Monitoring, Energy Management and Supply Chain.

Topology for Federated Learning

Let’s know about the topology to be incorporated for this research FL. The topologies to be utilized are as follows: Collaborative Monitoring, Data Sources, Security and Privacy Measures, Node Setup, Training Data Diversity, Model Aggregation, Alert Generation and Response, Threat Detection, Aggregated Threat Assessment and Continuous learning and Improvement.

Environment in Federated Learning

Environment that used for this research FL may include Compliance Considerations, Network Infrastructure, Testing and Validation, Node Resources, User Training, Server Resources, Data Pre-processing, Data Collection Mechanism, Security Protocols, Scalability Planning and Monitoring Tools.

Simulation Tools

Here we provide the software requirements that are needed for this topic. The simulation set up used for this research is as follows. The network simulator that we utilized for this research is NS 3.26. Then the operating system here we employed to implement the work is Ubuntu 14.04 LTS.

Results

The research is based on Federated Learning based threat detection and it is used to provide security. The proposed research is compared with various techniques or methods and then is contrasted with some performance metrics to obtain the best findings. The research is executed by using the operating system Ubuntu 14.04 LTS.

Federated Learning Research Ideas

The succeeding are the research topics that are related to the FL technique. The offered topics give assistance to us when we have any queries or doubts about this proposed technique.

A Review of Privacy-Preserving Federated Learning, Deep Learning, and Machine Learning IIoT and IoTs Solutions
An Enhancing Semi-Supervised Federated Learning Framework for Internet of Vehicles
Federated Learning Model for Contextual Sensitive Data Quality Applications: Healthcare Use Case
Cyber threat hunting using unsupervised federated learning and adversary emulation
Multi-Site Clinical Federated Learning Using Recursive and Attentive Models and NVFlare
Privacy-Preserving and Verifiable Decentralized Federated Learning
Analysis of Mathematical Modelling Deterministic and Stochastic Problems in Federated Learning
Poster: Raising the Temporal Misalignment in Federated Learning
Detection and Mitigation of SQL and Jamming Attacks on Switched Beam Antenna in V2V Networks Using Federated Learning
Robust Federated Learning with Local Mixed Co-teaching
Fairness and Effectiveness in Federated Learning on Non-independent and Identically Distributed Data
FedWNS: Data Distribution-Wise Node Selection in Federated Learning via Reinforcement Learning
Incentive Mechanism for Federated Learning Participants Based on Statistical Analysis Features
Entropy to Mitigate Non-IID Data Problem on Federated Learning for the Edge Intelligence Environment
Secure Federated Learning: An Evaluation of Homomorphic Encrypted Network Traffic Prediction
Enhancing Federated Learning Efficiency with Generative Model-Based Data Augmentation for Non-IID Data
Emergency Vehicle Identification for Internet of Vehicles Based on Federated Learning and Homomorphic Encryption
An Efficient Light-Weight Federated Learning Framework Implemented on Kubernetes and Docker
Efficient Federated Learning with Adaptive Client-Side Hyper-Parameter Optimization
Keep It Simple: Fault Tolerance Evaluation of Federated Learning with Unreliable Clients
Privacy-Preserving Federated Learning in Healthcare
Towards Instant Clustering Approach for Federated Learning Client Selection
Cloud Native Federated Learning for Streaming: An Experimental Demonstrator
BIFT: A federated learning System for Connected and Autonomous Vehicles Based on Blockchain
Federated Learning Differential Privacy Preservation Method Based on Differentiated Noise Addition
Racial Bias Mitigation with Federated Learning Approach
Comparison of Different Models for Federated Learning Based Parking Space Estimation
Implementing Federated Learning based on RainForest Model
Federated Learning for Online Resource Allocation in Mobile Edge Computing: A Deep Reinforcement Learning Approach
Byzantine-Resilient Federated Learning With Differential Privacy Using Online Mirror Descent
Federated Learning Framework to Decentralize Mobility Forecasting in Smart Cities
Adaptive User Scheduling and Resource Allocation in Wireless Federated Learning Networks: A Deep Reinforcement Learning Approach
AFLChain: Blockchain-enabled Asynchronous Federated Learning in Edge Computing Network
FLASH-RL: Federated Learning Addressing System and Static Heterogeneity using Reinforcement Learning
A Study of Deep Learning and Blockchain-Federated Learning Models for Covid-19 Identification Utilizing CT Imaging
Enhancing Global Model Accuracy: Federated Learning for Imbalanced Medical Image Datasets
An Android Federated Learning Framework for Emergency Management Applications
FL2: Fuzzy Logic for Device Selection in Federated Learning
Research on Dynamic Image Segmentation Method Based on Detection Federated Learning Algorithm
Fast-Convergent Federated Learning via Cyclic Aggregation
Federated Learning on Edge Devices in a Lunar Analogue Environment
Accelerating Federated Learning via Modified Local Model Update Based on Individual Performance Metric
A Novel Framework for Distributed and Collaborative Federated Learning based on Blockchain and Smart Contracts
WIP: Federated Learning for Routing in Swarm Based Distributed Multi-Hop Networks
Federated Learning With Server Learning for Non-IID Data
RingSFL: An Adaptive Federated Learning System for Heterogeneous Clients
Deep Equilibrium Models Meet Federated Learning
Federated Learning vs Edge Learning for Hot Water Demand Forecasting in Distributed Electric Water Heaters for Demand Side Flexibility Aggregation
Wireless Network Selection System Using Federated Learning Considering Applications in Use
A Simple Python Testbed for Federated Learning Algorithms