Abstract

The increasing use of wireless sensor networks, which normally comprise several very small sensor nodes, makes their security an increasingly important issue. They can be practically and efficiently secured using intrusion detection systems. Conventional security mechanisms are not usually applicable due to the sensor nodes having limitations of computational power, memory capacity, and battery power. Therefore, specific security systems should be designed to function under constraints of energy or memory. A partially distributed intrusion detection system with low memory and power demands is proposed here. It employs a Bloom filter, which allows reduced signature code size. Multiple Bloom filters can be combined to reduce the signature code for each Bloom filter array. The mechanism could then cope with potential denial of service attacks, unlike many previous detection systems with Bloom filters. The mechanism was evaluated and validated through analysis and simulation.

Highlights

  • Wireless sensor networks (WSNs) are composed of several very small sensor nodes

  • The proposal had a limitation: relay nodes participate in an Intrusion detection systems (IDSs) on the network layer, causing the distributed IDS mechanism to operate on the network layer, which leads to the module not being able to detect fragmented attack signatures

  • To construct a partially distributed IDS, we propose a classification method to distribute attack signatures among multiple Bloom filter arrays

Read more

Summary

Introduction

Wireless sensor networks (WSNs) are composed of several very small sensor nodes They are generally designed to operate in non-infrastructure environments using inexpensive hardware. While conventional IP networks can employ various intrusion detection mechanisms, the high resource requirements make some systems less suited to wireless sensor nodes, and several researchers have sought to develop various systems compatible with WSNs [5]. Most such IDSs analyze anomalous behavior; signature-based IDSs for WSNs receive much less attention.

Bloom Filter
IDSs using Bloom Filters
Problem Statement
Partially Distributed IDS Mechanisms
Insert Phase for the Partially Distributed IDS
Operation of Partially Distributed IDS
Memory Use
Energy Consumption
Detection Rate
Findings
Conclusions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call