Interference Prediction in Wireless Networks: Stochastic Geometry Meets Recursive Filtering

Abstract

This article proposes and evaluates a technique to predict the level of interference in wireless networks. We design a recursive predictor that estimates future interference values by filtering measured interference at a given location. The predictor's parameterization is done offline by translating the autocorrelation of interference into an autoregressive moving average (ARMA) representation. This ARMA model is inserted into a steady-state Kalman filter enabling nodes to predict with low computational effort. Results show a good accuracy of predicted values versus true values for relevant time horizons. Although the predictor is parameterized for Poisson-distributed nodes, Rayleigh fading, and fixed message lengths, a sensitivity analysis shows that it also tends to work well in more general network scenarios. Numerical examples for underlay device-todevice communications, a common wireless sensor technology, and coexistence scenarios of Wi-Fi and LTE illustrate its broad applicability. The predictor can be applied as part of interference management to improve medium access, scheduling, and radio resource allocation.

Statistical Tools and Methodologies for Ultrareliable Low-Latency Communication—A Tutorial

“…This requires first collecting the data on the items of interest over a period of time, including its conversion into a time-series format, and then adopting appropriate statistical processing methods. The latter includes autoregressive moving average (and its variants), Markov models (refer to Section II-C), Bayesian network [61], Kalman filter [62], and ML. The ML methods suitable for time-series modeling/prediction include deep belief networks [63], convolutional neural networks [64], recurrent neural networks [65], long short-term memory networks [35], and the novel "transformer" architecture [66], which may provide unprecedented support to tackle sophisticated and high-dimensional problems.…”

“…Then, we calculate the delay bound as in (62) and upper bound the delay violation probability, i.e., P out = Pr[W (t ) > w] as in (71). Note that when we introduce fading to the model, the rate expressions include logarithmic terms, which makes it cumbersome to find closed-form expressions for MGFs and the metrics.…”

Statistical Tools and Methodologies for URLLC- A Tutorial

“…Commonly used models in the literature include the autoregressive moving average (and its variants) [196], Markov chains [197], [198], Bayesian network [199], and Kalman filter [200].…”

A Nonlinear Autoregressive Neural Network for Interference Prediction and Resource Allocation in URLLC Scenarios

“…Interference management has always been a challenge in wireless systems design [4]. Efficient interference management is among the possible ways to ensure efficient resource allocation (and to ensure scalability) for URLLC services.…”

“…Haenggi et al [6] used stochastic geometry to derive mean interference and probability distribution in wireless networks, Schmidt et al [4] proposed the use of a recursive predictor to estimate future interference values by filtering the measured interference. Chinchali et al [7] obtained promising results segmenting corrupted wireless transmissions into desired signal and interference estimate using Deep Learning.…”