Rapid channel assignment for intelligent indoor scenarios

Deploying new wireless systems is very difficult in the current system of frequency spectrum assignments. The most prospective spectrum bands are fixed allocated for specific services and these bands are controlled by national telecommunication (governmental) organizations. Measurements of the frequency spectrum background show that a huge underutilization of the frequency spectrum exists. Overall effectiveness of spectrum utilization is widely discussed in the context of cognitive radio and dynamic spectrum allocation. An autonomous and intelligent system should improve spectrum sharing capabilities by detecting current, licensed users - primary users with established connections in unused spectrum bands for cognitive radio - and secondary users. In this paper, a machine learning algorithm is used and channels in particular bands are scored according to a weight function. Real measured data are used as frequency spectrum background. This system should decrease interference in communication channels efficiently and increase data throughput with minimal costs. A further significant reduction of radiation power should be obtained by spectrum efficient communication. Smart buildings represent a great opportunity for this type of cognitive system.