ADAPTIVE METHOD FOR OPTIMIZATION OF MAS PROTOCOLS PARAMETERS IN WIRED AND WIRELESS NETWORKS

Abstract

The article develops an adaptive method for optimizing MAC protocol parameters for wired and wireless telecommunication networks, aimed at increasing the efficiency of access to a shared environment under conditions of stochastic traffic dynamics and variable load intensity. The relevance of the study is due to the growth of data transmission volumes, the spread of real-time services, IoT applications and heterogeneous network environments, in which traditional algorithms such as CSMA/CA with fixed or semi-adaptive parameters of the competition window do not provide stable quality of service.

The proposed approach is based on the Markov description of the channel competition process and its extension in the form of a Markov decision process (MDP). Such a formalization allows taking into account both the current state of the network and the predicted evolution of traffic intensity and collision probability. Within the model, a set of states, actions and a reward function are defined, reflecting the trade-off between delay, number of retransmissions and throughput. The optimal policy for controlling the parameters of the backoff procedure is formed by solving the Bellman equation, which ensures the minimization of the average delay and the reduction of the probability of collisions depending on the load mode.

The obtained analytical relations describe the influence of the adaptive choice of the contention window size and waiting time on the QoS indicators and allow implementing the algorithm for dynamic MAC-layer tuning without changing the physical layer. The effectiveness of the method was verified by modeling in the OMNeT++ environment for various load scenarios. The experimental results demonstrated a reduction in the average packet delay by 12–14%, a reduction in the probability of collisions by up to 21.6%, and an increase in network throughput by 13.9% compared to the basic algorithms.

The proposed adaptive method provides stabilization of the MAC-layer operation under conditions of high and variable load, increases the efficiency of bandwidth use, and has prospects for application in new generation telecommunication systems, in particular in Wi-Fi infrastructure, IoT networks, and heterogeneous wired-wireless segments.