This chapter developed a machine learning approach for indoor environment classification based on real-time measurements of the RF signal. Several machine learning classification methods were contemplated, including DTs, SVM, and k-NN, using different RF features. Results obtained show that a machine learning approach using k-NN method, utilizing CTF and FCF, outperforms the other methods in identifying the type of the indoor environment with a classification accuracy of 99.3%. The predication time was obtained to be less than 10 u,s, which verifies that the embraced algorithm is successful for real-time deployment scenarios. The results of this chapter facilitate an efficient deployment of IoT applications in dynamic channels

Chapter Contents:

• 10.1 Introduction
• 10.2 Indoor radio propagation channel
• 10.2.1 Characteristics of RF indoor channel
• 10.2.2 Design considerations for the RF indoor channel
• 10.3 Data collection phase: practical measurements campaign
• 10.4 Signatures of indoor environment
• 10.4.1 Primary RF features
• 10.4.2 Hybrid RF features
• 10.5 Spatial correlation coefficient
• 10.6 Machine learning algorithms
• 10.6.1 Decision trees
• 10.6.2 Support vector machine
• 10.6.3 k-Nearest neighbor
• 10.7 Cascaded Machine Learning Approach
• 10.7.1 Machine learning for indoor environment classification
• 10.7.2 Machine learning for localization position estimation
• 10.8 Conclusion
• References

Inspec keywords: learning (artificial intelligence); feature selection; pattern classification; support vector machines; Internet of Things

Other keywords: CTF; cascaded machine learning approach; indoor environment; SVM; adaptive feature selection; k-NN method; classification accuracy; indoor environment classification; FCF; real-time deployment scenarios; IoT applications; indoor localization

Subjects: Data handling techniques; Knowledge engineering techniques