Genetic evolution and adaptation of advanced protocols for ad hoc network hardware systems

Abstract

The diversity of future technologies requiring ad hoc networks to operate within unpredicted situations will mean an increase in the required flexibility of the actual protocols used for communicating information. A methodology is proposed to genetically evolve the optimum ad hoc network communication protocol under any given network scenario. The methodology creates and dynamically adapts the communication protocol based upon an alphabet of characteristics and performance metrics using simple protocol mapping techniques and minimisation of a fitness function via a genetic selection process. A scenario has been created to evaluate the performance of the methodology in finding the optimum solution. Preliminary results show that the methodology is able to find the global optimum within several runs. The methodology could be enhanced using Field Programmable Gate Array (FPGA) hardware nodes for real time performance and distributed control. © 2011 Springer-Verlag London Limited.