Efficient Optimal Linear Estimation for CPM: An Information Fusion Approach

The Internet of Things (IoT) has recently experienced considerable growth, enabling communication between a wide range of devices. Given the prevalence of mobile IoT devices, optimizing hardware resources becomes crucial, requiring power and complexity reduction strategies. Continuous phase modulation (CPM) offers attractive features for IoT, such as spectral and power efficiency. However, many CPM receiver designs are computationally intensive, which limits their practicality for this kind of applications. This article presents a new efficient CPM receiver based on information fusion techniques that takes advantage of the inherent memory of CPM models. This proposal offers a family of biased and unbiased estimators for CPM with linear computational complexity, derived from a clear optimization objective based on a mean-squared error criterion. The proposed design is shown to be optimal under such criterion and explicit expressions for its error probability are provided. The error probability of the proposed design is shown to be approximately the same as the theoretical lower bound of the optimal receiver for binary phase shift keying (BPSK) for certain CPM models. Theoretical and simulation results confirm the benefits of this contribution, highlighting its near-optimal performance for specific schemes of Gaussian frequency-shift keying (GFSK). This holds particular significance due to the role of GFSK in Bluetooth low energy (BLE) wireless communication technologies, which contributes significantly to the advancement of mobile IoT devices.

I. Introduction

The Internet of Things (IoT) stands as a cutting-edge technology, enabling the connection of various devices for applications, such as automotive infotainment, medical healthcare, home automation, and machine-to-machine communications. Its primary objective is to enable seamless data interchange without the need for human intervention [1], [2]. Given the prevalence of mobile IoT devices, optimizing hardware resources becomes crucial, requiring power and complexity reduction strategies.