Adaptive Cross-Layer Optimization Using Mimo Fuzzy Control System in Ad-hoc Networks.

Due to quick and inexpensive deployment, decentralized wireless ad-hoc networks have many potential application domains. However, stringent resource constraints, user mobility, limited channel capacity and high error rate of packets received at the wireless network interfaces make the design and optimization of these networks a challenging task. To solve these issues involving different layers of the protocol stack, we propose a multivariable Fuzzy Logic based Cross-Layer Optimization (FLCLO) Algorithm. In this algorithm, multiple parameters from various layers are given as inputs to a fuzzy logic controller. In fuzzy multi-attribute decision making framework, several distinctive parameters with inherent uncertainties in practical wireless communication scenarios are efficiently characterized by the fuzzy linguistic information. The fuzzy logic optimization technique developed here makes decisions by simultaneously considering multiple criteria that affect network performance in terms of packets per second, number of packets lost, throughput, mean SNIR and number of collisions. The performance of the proposed crosslayer fuzzy algorithm is evaluated by conducting simulations in OMNeT++ with dynamic fuzzy logic system embedded on it. The experimental results obtained from the simulation show that FLCLO used for selecting an efficient next-hop node performs better in terms of reducing the average MAC delay, energy spent on packet transmissions, and packet error rate. A comparative analysis of FLCLO with the media access algorithm used by the original IEEE 802.11 protocol shows that FLCLO is 52-68% more energy-efficient, has 45% lower MAC delay and reduces the packet error rate by 71%. In addition, it guarantees a lower collision rate and reduced packet loss ratio when compared with the conventional IEEE 802.11 ad-hoc routing model in the same set up. Finally, the simulation results demonstrate that our algorithm offers considerable performance enhancement in terms of mitigated packet loss ratio, mean delay, and energy consumption in contrast to the previous related works existing in literature. [ABSTRACT FROM AUTHOR]