Abstract:Nonlinear time series models with exogenous regressors are essential in econometrics, queuing theory, and machine learning, though their statistical analysis remains incomplete. Key results, such as the law of large numbers and the functional central limit theorem, are known for weakly dependent variables. We demonstrate the transfer of mixing properties from the exogenous regressor to the response via coupling arguments. Additionally, we study Markov chains in random environments with drift and minorization conditions, even under non-stationary environments with favorable mixing properties, and apply this framework to single-server queuing models.

What problem does this paper attempt to address?

### Problems the paper attempts to solve This paper aims to study how the mixing properties in the iterative process of nonlinear time - series models are transmitted from exogenous covariates to response variables in the presence of exogenous covariates. Specifically, the paper focuses on the following points: 1. **Transmission of mixing properties**: - Research on how to transmit the mixing properties of exogenous covariates to response variables through coupling arguments. - Explore the mixing properties of Markov chains in random environments in non - stationary environments. 2. **Basis for statistical analysis**: - Establish the applicability of the Law of Large Numbers (LLN) and the Central Limit Theorem (CLT) in these models. - Provide the conditions and verification methods for these theorems in practical applications. 3. **Application areas**: - Apply these theories to queuing theory, especially single - server queuing models. - Demonstrate the application value of these theories in practical problems through specific examples. ### Main contributions 1. **Transmission of mixing properties**: - Through coupling inequalities and moment estimates, it is proved that the mixing properties of exogenous covariates can be transmitted to the iterative process of response variables. - Propose a novel Cramér - Rao lower bound for estimating the variance of partial sums, thereby establishing the Functional Central Limit Theorem (FCLT) under certain conditions. 2. **Tools for statistical analysis**: - Prove that under appropriate conditions, the L1 Law of Large Numbers and the Central Limit Theorem hold for nonlinear time - series models containing exogenous covariates. - Provide the theoretical basis for constructing confidence intervals. 3. **Specific applications**: - Re - examine single - server queuing models and prove the applicability of the Functional Central Limit Theorem in these models. - Demonstrate the application value of these theories in practical problems through specific examples. ### Conclusion This paper provides a solid theoretical basis for the statistical analysis of time - series models containing exogenous covariates by in - depth study of the transmission mechanism of mixing properties in nonlinear time - series models. These results are not only of great theoretical significance but also have broad application prospects in practical applications, especially in queuing theory and other related fields.

Transition of $α$-mixing in Random Iterations with Applications in Queuing Theory

Strong Approximation for Ρ-Mixing Sequences

Strong mixing properties of discrete-valued time series with exogenous covariates

Statistical properties for mixing Markov chains with applications to dynamical systems

Mixing it up: A general framework for Markovian statistics

Ergodicity and Mixing of Sublinear Expectation System and Applications

Convergence of de Finetti's mixing measure in latent structure models for observed exchangeable sequences

Uniform polynomial rates of convergence for a class of Lévy-driven controlled SDEs arising in multiclass many-server queues

Ergodicity and mixing for stochastic partial differential equations

Notas sobre Teoría de colas y algunas aplicaciones

Exact sampling for some multi-dimensional queueing models with renewal input

A Lévy input fluid queue with input and workload regulation

Queue Stability and Probability 1 Convergence via Lyapunov Optimization

Perfect simulation of Markovian load balancing queueing networks in equilibrium

Learning with little mixing

Empirical and Instance-Dependent Estimation of Markov Chain and Mixing Time

Mixing Times and Structural Inference for Bernoulli Autoregressive Processes

Exponential Single Server Queues in an Interactive Random Environment

M/M/1 queue in two alternating environments and its heavy traffic approximation

Heavy-traffic queue length behavior in a switch under Markovian arrivals

Unbiased approximation of the ergodic measure for piecewise $α$-stable Ornstein-Uhlenbeck processes arising in queueing networks