Article Preview
Top1. Introduction
In a large number of emerging WSN applications, congestion has proved to be a major challenge, for which many solutions have been provided in the past studies (Uppal, Kumar, & Singh, 2014; Prabha, 2015; Ahmad, Jabbar, Paul, & Rho, 2014). Development of multi-channel MAC protocols has mainly been motivated by the urge to control the congestion and minimize the delay (EkbataniFard & Monsefi, 2012). A WSN node is able to have parallel and collision free transmissions with the aid of Multi-channel protocols, resulting in improved energy, delay and throughput performance (Uppal, Kumar, & Singh, 2014). Due to the recent active research in multi-channel protocols, it becomes essential to investigate the comparative performance for each protocol using techniques such as queuing analysis (Park & Hwang, 2013; Sikandar & Kumar, 2014; Nieminen, Björkbom, Jäntti, & Eriksson, 2012).
In most of the existing MAC protocols, the number of channels are divided based on the traffic priority (Donmez, Isik, & Ersoy, 2014; Min, Kim, Kwon, & Lee, 2013; Wang, Li, Sohraby, Daneshmand, & Hu, 2007; Yaghmaee & Adjeroh, 2009; Patel & Khare, 2014). The traffic priority is generally set at the time of deployment based on the data type. Similarly, setting priority of nodes based on dynamic factors such as RSS offers significant help in making the WSN suitable for real time applications.
A number of channel allocation strategies have been proposed in the past, however, majority of them focused on the wireless cellular (Epstein & Schwartz, 2000) & mesh networks (Skalli, Ghosh, Das, & Lenzini, 2007), WiMAX or Wi-Fi (Andrews, Kondareddy, & Agrawal, 2010). Also, most prioritized sharing algorithms proposed for WSN focus on offering service to the high priority traffic neglecting the low priority, even if network resources are available. For this work, a research gap is thus identified for performing the comparative analysis of channel sharing mechanisms in WSNs.
In the present study, we evaluated an existing congestion control scheme developed for multichannel WSN (Park & Hwang, 2013) which proposed to use prioritized channel sharing mechanism for the nodes based on the RSS of packets. Congestion was categorized as light and heavy, and the control mechanisms for each category and node priority were developed. The authors proposed to use prioritized sharing due to which the low priority traffic might not find a channel even if it is idle. Furthermore, the analysis was provided for the total system throughput and performance for low and high priority traffic was not evaluated separately. In contrast, we do not prioritize the nodes based on the RSS but instead allow the application layer to set the priorities as per the requirements. The schemes are mainly developed for the WSN applications where each node may have data of different priorities. For example, if in an intrusion detection system, a WSN node is embedded with multimedia and temperature sensors, the priority of the multimedia traffic will be higher despite of the large size of packets. On the other hand, if the similar sensor node has to operate in a forest-fire monitoring system, the temperature sensor would be assigned a higher priority. Furthermore, we do not only study the prioritized sharing but also the other channel sharing mechanisms.