Security in Swarm Robotics

Security in Swarm Robotics

Thalia May Laing (Royal Holloway, University of London, UK), Siaw-Lynn Ng (Royal Holloway, University of London, UK), Allan Tomlinson (Royal Holloway, University of London, UK) and Keith M. Martin (Royal Holloway, University of London, UK)
DOI: 10.4018/978-1-4666-9572-6.ch002
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Inspired by social animals, such as ants, bees and fish, which appear to exhibit what has been dubbed ‘swarm intelligence', swarm robotic systems aim to accomplish goals that are unachievable by an individual robot. Swarm robotics have a large number of potential uses, including applications in the military, monitoring, disaster relief, healthcare and commercial applications. To be able to achieve their goals, it is of utmost importance that communications between agents are secure in the presence of possibly malicious interruptions and attacks from adversaries. The authors will discuss the issues surrounding the provision of secure communications in swarm robotics: what secure communications mean, how the characteristics of swarm robotics present a security challenge, the relationship between security issues for swarm robotics and other network technologies, and how different adversarial models demand different types of solutions. It will then be discussed what the important open research questions are in secure communications in swarm robotics.
Chapter Preview
Top

Introduction

Swarm robotics is concerned with the coordination of large numbers of relatively simple robots. Although there is no universally accepted definition for swarm robotic systems, Şahin (2005) proposes the following working definition: swarm robotics is the study of how a large number of relatively simple physically embodied agents can be designed such that a desired collective behaviour emerges from the local interactions among agents and between the agents and their environment.

Swarm robotic systems aim to accomplish goals that are unachievable by an individual robot. In a number of situations, having numerous simple robots forming a swarm, rather than an individual complex robot, could be beneficial as it may be cost effective or achieve the set goal more effectively. Because of this, swarm robotics have a large number of potential uses, including applications in the military, medical scenarios, disaster relief, monitoring and commercial applications (Şahin, 2005).

Şahin (2005) explicitly puts forwards five characteristics as criteria for distinguishing swarm robotics from other multi-robot research in Section 3 of his paper. Şahin suggests these characteristics as a way to differentiate swarm robotics from other multi-robot systems. His five defining characteristics of swarm robotics are:

  • 1.

    Autonomous Robots. Robots are able to act without the direct intervention of humans and have control over their own actions and internal state.

  • 2.

    Large number of Robots. Closely linked to the idea of scalability, there should be a large number of robots, or studies should be applicable to the control of large robotic swarms.

  • 3.

    Consist of a Few Homogeneous Groups of Robots. The swarm network should consist of relatively few groups of homogeneous robots.

  • 4.

    Relatively Incapable or Inefficient Robots. On an individual level, the robots should be relatively simple and either incapable of completing tasks individually, requiring cooperation amongst the swarm to achieve the global goal, or working as a group should improve the performance and robustness of the handling of the task.

  • 5.

    Local Sensing and Communication Capabilities. The robots should have local and limited sensing and communication abilities to ensure distributed coordination amongst the swarm. A global communication channel may be used to download a common program onto the swarm, however this should not be used for coordination amongst the robots (as this is likely to be unscalable) and the communication is considered to be one way, in the direction from the channel to the swarm.

Other characteristics not explicitly listed by Şahin (although some are implicit) include:

Key Terms in this Chapter

Data Origin Authentication: A security service that identifies a specific entity as the source or origin of a given piece of data.

Internal Adversary: An internal adversary is an attacker from within the swarm that has access to any secret material established prior to deployment.

Identification: Sometimes called entity authentication , identification is a security service that identifies specific entities in isolation from any other activity the entity may want to perform.

Resilience: A swarm is resilient if the loss of individual agents has little impact on the success of the task of the swarm.

Entity Authentication: Sometimes called identification , entity authentication is a security service that identifies specific entities in isolation from any other activity the entity may want to perform.

Cryptography: Cryptography is the art (and science) of designing cryptosystems to transform messages in such a way that two entities may communicate securely over an insecure channel.

Confidentiality: The assurance that no one other than the intended recipient(s) can read the data; in other words, the data is kept secret between the sender and recipient.

Availability: Availability ensures that accessibility and usability are available upon demand by an authorised entity. The loss of availability is commonly referred to as a denial-of-service attack.

Integrity: The assurance that the data has not been altered, either maliciously or accidentally, in an unauthorised way.

External Adversary: An external adversary is an agent who, unlike an internal adversary, is assumed to originate from outside the swarm and thus is not in possession of any secret material established prior to deployment.

Complete Chapter List

Search this Book:
Reset