Survey of routing metrics for multi channel multi interface wireless mesh networks

Abstract

WMNs (Wireless Mesh Networks) represent a new trend in wireless communications, providing greater flexibility, reliability and better network performances than standard WLANs (Wireless Local Area Networks). Having in mind a need for more efficient and faster data transmission, one of the greatest challenges in designing WMNs is to provide greater throughput. This is done by introducing networks the nodes of which have multiple interfaces that can work on one of multiple channels (the channel is chosen by a predefined algorithm), so-called Multi-Channel Multi-Interface (MCMI) WMNs. In order to provide better network performances, it is necessary to define routing protocols and metrics used to find and select the optimal route from the source to the destination node. Since WMNs are characterised by a high dynamic range of the received signal level, especially when they are realized in complex environments such as indoors, it is necessary to define routing metrics that can follow these fast changes in propagation conditions that occur in each link in the network.

This paper presents a critical review of the most well-known routing metrics for MCMI-WMNs. The paper is organised as follows:

Introduction

In this section, it is pointed that the need for wireless communications is growing in the modern society. Since wireless networks are widely used, the overall wireless communication is greatly increased. In order to provide greater throughput of user data, it is very important to select the optimal route between the source and the destination node. Therefore, a great attention is dedicated to creating an optimal routing protocol and routing metrics. Routing protocols have an important role in finding paths form the source to the destination nodes, and the routing metrics have the task to select the optimal route form multiple routes.

Criteria for creating metrics

In order to find the optimal metric, it is necessary to define criteria to be taken into account when the metric is formed. These criteria are: interference (intra-flow, inter-flow and external), the amount of information, link load, agility, stability and ability to track rapid changes at every link in the network. In this section, each of the criteria is defined and it is explained what a metric should have in order to satisfy the defined criteria.

Overview of the metrics available in the literature

In this section, the metrics available in the literature are described. Based on the criteria taken into account, metrics are divided into two categories: hop count metric and link-quality metrics. Link-quality metrics described in this paper are: ETX (Expected Transmission Count), LAETT (Load Aware Expected Transmission Time), EETT (Exclusive Expected Transmission Time), WCETT (Weighted Cumulative Expected Transmission Time), iAWARE (Interference Aware Routing), MIC (Metric of Interference and Channel-switching), powerETX, powerWCETT, and powerMIC. The rest of this section is organised into the following sub-sections:

Hop count metric

ETX (Expected Transmission Count) metric

LAETT (Load Aware Expected Transmission Time) metric

EETT (Exclusive Expected Transmission Time) metric

WCETT (Weighted Cumulative Expected Transmission Time) metric

iAWARE (Interference Aware Routing) metric

MIC (Metric of Interference and Channel-switching) metric

powerETX (power Expected Transmission Count) metric

powerWCETT (power Weighted Cumulative Expected Transmission Time), and powerMIC (power Metric of Interference and Channel-switching) metrics

In each of the sub-sections, a detailed description of a particular metric is given as well as its basic characteristics. The metric forming formulas are given as well.

Comparative analysis of the described metrics

In this section, the comparative analysis of the previously described metrics is given. Besides, the comparative analysis in the terms of throughput of user data and average end to end delay is provided, based on the results given in the literature. The analysis is performed on six metrics: ETX, WCETT, MIC, powerETX, powerWCETT and powerMIC. The first three metrics are chosen since they are most frequently used in the literature, and the other three metrics are formed based on the first three by introducing a parameter that follows fast changes in the state of each link independently.

Conclusion

In the final section, the concluding remarks are given. This section represents an overview of the entire paper, as well as the reasons why this subject was chosen as a topic of this paper.