Channel Estimation for Industrial Internet of Things Using Particle Filter with Adaptive Pilots

Abstract

The harsh physical conditions in industrial environments cause serious difficulties in the wireless links for the Industrial Internet of Things (IIoT). Non-Gaussian impulsive noise, doubly-selective fading and nonlinearity of electrical devices in transceivers destroy the signal. Therefore, the channel estimation for equalization is a critical operation. The channel estimation is aided by pilots in the orthogonal frequency division multiplexing (OFDM) technique used in common standards of IIoT. The pilot arrangement in the time-frequency plane is an important design factor. There is a trade-off between channel estimation accuracy and transfer data rate, which decreases by increasing pilot overhead. In this paper, a new channel estimation method is developed for IIoT wireless links using a particle filter for estimating the channel frequency response in adaptive pilots. The distance of the pilots along the frequency axis is adjusted according to the channel response. The proposed method is simulated in complicated scenarios based on real-world measurements. This method is compared with the least-squares (LS) estimator and Kalman filter. The evaluations are performed by Bit Error Rate (BER) and pilot overhead criteria. The results illustrate that the novel combination of adaptive pilots and particle filter in the proposed method works appropriately for IIoT channel estimation. The better performance (less BER) is achieved with the same pilot overhead, or from another viewpoint, the similar BER is attained with less pilot overhead.