Mathematical modeling and analysis for adaptive medium access protocols in wireless networks

Abstract

With the arrival of the big data era, the demand of wireless communications has soared because of Internet of Thing (IoT) and cloud services. To meet this exponentially increasing demand, an efficient spectrum sharing rule is required. Accordingly, the role of medium access control (MAC) protocol, which determines sharing rules, has become more important. However, the IEEE 802.11 distributed coordination function (DCF), which is the de facto standard for the MAC protocol in today’s WLAN, yields poor performance in perspective of fairness. To overcome this problem, a MAC protocol called the Renewal Access Protocol (RAP) was proposed. It was shown that the RAP achieves optimal throughput, high short-term fairness, and near-optimal delay performances when the number of nodes in the network is known to every node. However, the optimal RAP requires extra communications for a node to know the number of nodes. In this work, we propose an adaptive version of the RAP (A-RAP) where a node estimates the number of nodes autonomously, so that it can access the channel with the optimized RAP. With the A-RAP, a node tracks the number of nodes and adjusts its parameter in real time even when the number of nodes is changing. The algorithm of the A-RAP is elaborately constructed through mathematical models with random walks. Numerical and simulation results demonstrate the superiority of the proposed A-RAP.

Next, we examine another important issue in the performance analysis of MAC protocol through spatial modeling. The spatial modeling of MAC protocol has received little attention due to the complex-ity in the dynamics of Wireless Local Area Network (WLAN). In this work, we analyze the performance of MAC protocol based on stochastic geometry theory. Specifically, we focus on the correlation between consecutive packet transmissions, which comes from the spatial distribution of the nodes in the net-work. We verify the conditions where the independence assumption that individual transmission results are independent holds, which is widely assumed in most of the previous works on MAC protocol. To investigate the validity of the independence assumption, we derive the correlation coefficient and the mu-tual information. Furthermore, we discuss another correlation issue not covered with these correlation measures.