1st International ICST Conference on Bio Inspired Models of Network, Information and Computing Systems

Research Article

Exploring biologically-inspired evolvable network applications with the BEYOND architecture

  • @INPROCEEDINGS{10.1145/1315843.1315860,
        author={Chonho Lee and Hiroshi  Wada and Junichi Suzuki},
        title={Exploring biologically-inspired evolvable network applications with the BEYOND architecture},
        proceedings={1st International ICST Conference on Bio Inspired Models of Network, Information and Computing Systems},
        publisher={ACM},
        proceedings_a={BIONETICS},
        year={2006},
        month={12},
        keywords={},
        doi={10.1145/1315843.1315860}
    }
    
  • Chonho Lee
    Hiroshi Wada
    Junichi Suzuki
    Year: 2006
    Exploring biologically-inspired evolvable network applications with the BEYOND architecture
    BIONETICS
    ACM
    DOI: 10.1145/1315843.1315860
Chonho Lee1,2,*, Hiroshi Wada1,2,*, Junichi Suzuki1,2,*
  • 1: Department of Computer Science, University of Massachusetts, Boston
  • 2: Boston, MA 02125, USA
*Contact email: chonho@cs.umb.edu, shu@cs.umb.edu, jxs@cs.umb.edu

Abstract

The BEYOND architecture applies biological principles and mechanisms to design evolvable network applications that autonomously adapt to dynamic environmental changes in the network. This paper describes two key components in BEYOND: (1) an evolutionary adaptation engine, called iNet, for network applications and (2) an application development environment, called BEYONDwork, for the adaptation engine. iNet is designed after the mechanisms behind how the immune system works. It models a set of environment conditions (e.g., network traffic) as an antigen and a behavior of network applications (e.g., migration and reproduction) as an antibody. iNet allows each network application to autonomously sense its surrounding environment conditions (i.e., antigens) and adaptively invoke a behavior (i.e., antibody) suitable for the conditions. The configuration of antibodies evolves via genetic operations such as mutation and crossover. BEYONDwork provides visual and textual languages to configure antigens and antibodies in iNet. The languages increase the ease of specifying and modifying iNet configurations. Simulation results show that iNet allows network applications designed with BEYONDwork to evolve themselves to adapt to changing network environments.