Quality of service routing for wireless sensor network
2017-02-23T02:33:54Z (GMT) by
Over the last few years, the development of sensor networks has gained increasing importance due to their potential to support challenging research in a wide range of applications, and this has led to a new era of Multi-Objective Quality of Service Routing (MoQSR). This MoQSR is a routing algorithm that aims to satisfy multiple quality of service (QoS). MoQSR has the potential to support Wireless Sensor Network (WSN) applications that have different data which have quality and requirements. MoQSR enables the data communication protocols to provide a service differentiation mechanism that can capture the requirements of heterogeneous traffic. A key distinguishing feature of MoQSR is its ability to differentiate the required data related to each type of traffic and to provide a route selection that is based on differentiating QoS requirements according to the data type. In fact, a WSN has limitations in energy supply and capabilities to ensure real-time guarantees because it poses a low duty cycle and transmission range. Most approaches related to data communication protocols are designed to deal with energy trade-off and not much has been done to optimise quality of service (QoS). Past research and current studies of QoS routing schemes only focus on a limited number of aspects of QoS. These studies do not provide a generic approach/scheme to encompass all attributes of QoS, and have limited support for meeting application specific needs. Thus, it is critical for QoS routing to factor in different application requirements. Therefore, in this dissertation, I propose, develop and validate a new QoS based routing for WSN, named Throughput Delay Guaranteed Routing Algorithm (TeGaR). Our TeGaR is based on differentiating QoS requirements according to the data type, which provides customised QoS metrics for each traffic category. It is modular and uses geographic information, which helps the routing algorithm create the best path to the destination node (BS). Our TeGaR aims to adapt to the requirements of heterogeneous traffic by using traffic diversity, the priority based delivery mechanism with multi-queuing policy and the route differentiation based Fitness Function method, which considers the quality of nodes matching the requirements of data categories. TeGaR is able to find a best node for a certain data category, while considering energy efficiency, reliability, latency and traffic congestion to cast QoS metrics as a multi-objective attribute. However, there are some areas in which the performance of the proposed TeGaR can be enhanced. First, to address the 'energy hole problem' due to an imbalance of energy consumption in the network, according to the nature of many-to-one routing in WSNs, I propose a non-uniform transmission range strategy, namely Transmission Range Extension based on the neighbourhood Size (TReNs). TReNs allows the nodes to extend their transmission range to a certain level when the number of neighbours decreases below a certain level. This extended approach is able to distribute energy consumption evenly, thus avoiding an early network dysfunction. Second, to reduce the data flooding or unguided packet transmission, due to the characteristic of non-uniform node deployment and link instability in WSNs, I propose and develop a self-adaptation routing region approach, namely the Rule-based Learning Adaptation Approach for Routing Regions (RuLeARR) that aims to control the area (region) of routing for individual communication between nodes. In this way, packet detouring or data flooding can be reduced, thus decreasing routing overheads and the duration for data delivery, further saving energy. Third, to improve the benefits of adaptation of the routing region approach that copes with minor changes in the network QoS parameters, I integrate situation awareness into the routing region algorithm, to develop Situation Aware and Self-Adaptive Routing Regions (SASARR). The SASARR uses a fuzzy logic to represent approximate and imprecise context of the changes of the network dynamic to represent a fine-grained and gradual adaptation for routing region. SASARRR extends the RuLeARR approach in order to provide a higher level of accuracy and granularity, thus improving QoS performance in a cost-efficient manner. We validate all the approaches discussed above by conducting extensive experimental simulation. This evaluation clearly demonstrates the ability of the TeGaR scheme, with its extended approaches (TReNs, RuLeARR and SASARRR), to satisfy multiple QoS requirements according to different application requirements, while coping with the challenges due to the nature of data communication characteristics in WSNs. These approaches make a significant contribution to the overall efficiency and effectiveness of the QoS differentiating routing. The contribution of this thesis has resulted in one international peer reviewed paper and additional material for further publication.