Solutions to Parallel and Distributed Computing Problems: Lessons from Biological Sciences

10.8 SUMMARY

Biologically inspired techniques, such as artificial neural networks (ANN) have been successfully applied to a variety of applications. With the rapidly changing telecommunication industry, and the increasing popularity of wireless networks, there has been a great deal of research on the application of neural networks to wireless and mobile telecommunication systems. Biocomputing techniques are about to enter the main stream.

In this chapter, we have presented several of these applications and have shown how ANN has been applied in them. They represent neural-network approaches to a variety of needs. First, we discussed a neural-network-based adaptive equalizer for digital mobile-radio channels, followed by a neural net that runs autonomously to solve channel-assignment problems in an arbitrary cellular mobile-radio system.

Pattern recognition is certainly one of the most relevant areas of neural-network application [20]. The range of concrete problems that can fall into this category is very wide. Perhaps an area of great practical interest and active research is the classification of what we can call social or economical patterns. In this chapter, we described how pattern-recognition schemes could be used in GSM mobile-radio resource management to improve the handover decision process without changing the GSM standards. Finally, we discussed our recent work on neural-fraud detection in mobile-telephone systems. In all of these applications, neural-network strategies ...