Performance Analysis of Chaotic Phase Transformation for NB-IoT Uplink

This work presents an approach for using chaos sequences to assist uplink data transmission in narrowband Internet of things (NB-IoT). Along with exploring the widely known Bernoulli and Chebyshev maps, Poincare map based chaos sequences are introduced, which are obtained by solving the Duffing oscillator equation. The novelty of this approach is in utilizing chaos sequences to solely modify the phase of the modulated symbols; without altering their amplitudes. Such a transformation preserves the circularly symmetric property of the noise components, which is essential for maintaining the statistical characteristics pertinent to data detection. To support the implementation of the proposed technique, a greedy algorithm for choosing an ideal set of K chaos sequences from a bigger pool of N sequences based on minimizing the maximum cross-correlation coefficient between any two selected chaos sequences is presented. It is found that Poincare map based sequences are the most efficient for K >= 128. Simulation results show that data decoding performance obtained by applying the proposed chaotic phase transformation technique is nearly the same as that of the conventional system. Since chaos sequences and/or their initial conditions can be programmed during installation, the proposed technique can potentially enhance the security of uplink data transmission.