ew 19(24): e4

Research Article

Secure Group based Routing and Flawless Trust Formulation in MANET using Unsupervised Machine Learning Approach for IoT Applications

Download96 downloads
  • @ARTICLE{10.4108/eai.13-7-2018.160834,
        author={Maitreyi  Ponguwala and DR.Sreenivasa  Rao},
        title={Secure Group based Routing and Flawless Trust Formulation in MANET using Unsupervised Machine Learning Approach for IoT Applications},
        journal={EAI Endorsed Transactions on Energy Web},
        volume={6},
        number={24},
        publisher={EAI},
        journal_a={EW},
        year={2019},
        month={10},
        keywords={MANET, Recommendation Filtering, Blackhole, Grayhole, IoT applications},
        doi={10.4108/eai.13-7-2018.160834}
    }
    
  • Maitreyi Ponguwala
    DR.Sreenivasa Rao
    Year: 2019
    Secure Group based Routing and Flawless Trust Formulation in MANET using Unsupervised Machine Learning Approach for IoT Applications
    EW
    EAI
    DOI: 10.4108/eai.13-7-2018.160834
Maitreyi Ponguwala1,*, DR.Sreenivasa Rao2
  • 1: Research Scholar of JNTUH, MGIT Hyderabad, India
  • 2: Professor in Dept. Of CSE, JNTUH Hyderabad, India
*Contact email: Ponguwalamaitreyi@gmail.com

Abstract

INTRODUCTION: Mobile Adhoc Network (MANET) is integrated with Internet of Things (IoT) in many application cases due to its flexibility and scalability. The dynamic nature of MANET introduces some security threats in IoT environment. In those threats, Blackhole attack and Grayhole attack are severe routing attacks that disrupts routing algorithm to crack transmission in the entire network.

OBJECTIVES: Many security mechanisms are introduced in MANET based on trust computation schemes. However, computation of inaccurate trust value degrades the performance of mitigation schemes. Thus the major objective of this work is to design a novel security mechanism to protect the MANET-IoT from different adversaries.

METHODS: in this paper we propose a novel group based routing algorithm with recommendation filtering supported by security monitors (SMs). Unsupervised machine learning algorithm is adapted for the purpose of recommendation filtering in the network. Initially the entire network is grouped by Secure Certificate based Group Formation (SCGF) algorithm. In each group, Recommendation Filtering by K-means algorithm (RF-K means) algorithm is employed to perform trust computation. For secure and optimal route selection, hybrid optimization algorithm that combines Genetic Algorithm and Fire Fly Algorithm (GA-FFA) is proposed. Data transmission is protected by novel Hash Message Authentication Code with Advanced Encryption Standard (HMAC-AES) algorithm in which hash function is integrated with cryptography function.

RESULTS: The proposed work is validated in network simulator-3 environment and the obtained results show better performance in terms of packet delivery ratio (96.3%), throughput (135kbps), delay (3.26ms), detection rate (99%), and energy consumption (8.5%).

CONCLUSION: The MANET-IoT network is secured by group formation and trust filtering approaches. Further, involvement of cryptography function ensures data security whereas hash function ensures data integrity.