In this chapter, sparse characterization of BWCS is discussed. First of all, a novel sparse non-parametric model is proposed to characterize BWCS channels, it has been demonstrated that it is an important supplement to the existing parametric models; and then, compressive sensing technique is applied to the on-body UWB channel estimation, the impulse response of the channel is perfectly reconstructed; finally, particle swarm optimization based support vector regression technique is used to explore obesity's effect on the on-body narrowband wireless channels. This chapter provides readers a totally new angle of view of looking at the current channel modelling technique in BWCS; thus will be beneficial to the ones who aim to developnew radio channel models for BWCS.

Chapter Contents:

• 4.1 Introduction
• 4.2 Basics of sparse non-parametric technique and compressive sensing
• 4.2.1 Sparse non-parametric technique
• 4.2.1.1 Empirical distribution function
• 4.2.1.2 The coefficients for the probability approximation
• 4.2.2 Basics of compressive sensing framework
• 4.3 Results and discussions regarding non-parametric modelling and on-body impulse response estimation
• 4.3.1 Establishing sparse non-parametric propagation models and their evaluation
• 4.3.1.1 Measurement setup
• 4.3.1.2 Kernel functions
• 4.3.1.3 Sparse non-parametric model characterization results and discussions
• 4.3.2 Sparse on-body UWB channel estimation
• 4.4 Statistical learning technique and its application in BWCS
• 4.4.1 Small-sample learning and background
• 4.4.2 Example of support vector regression
• 4.5 Conclusion
• Acknowledgements
• References

Inspec keywords: regression analysis; compressed sensing; wireless channels; radio networks; particle swarm optimisation

Other keywords: compressive sensing technique; channel modelling technique; support vector regression technique; BWCS channels; impulse response; particle swarm optimization; onbody narrowband wireless channels; sparse characterization; parametric models; on-body UWB channel estimation; body centric radio channels

Subjects: Optimisation techniques; Radio links and equipment; Signal processing and detection; Other topics in statistics