The recent research interest in wireless sensor networks has caused the development of many new applications and subsequently, these emerging applications have ever increasing requirements. One such requirement is that of mobility, which has inspired an entirely new array of applications in the form of mobile wireless sensor networks (MWSNs). In terms of communications, MWSNs present a challenging environment due to the high rate at which the topology may be changing. As such, the motivation of this work is to investigate potential communications solutions, in order to satisfy the performance demands of new and future MWSN applications. As such this work begins by characterising and evaluating the requirement of a large variety of these emerging applications. This thesis focuses on the area of routing, which is concerned with the reliable and timely delivery of data from multiple, mobile sensor nodes to a data sink. For this purpose the technique of gradient routing was identified as a suitable solution, since data can quickly be passed down a known gradient that is anchored at the sink. However, in a mobile network, keeping the gradient up-to-date is a key issue. This work proposes the novel use of a global time division multiple access (GTDMA) MAC as a solution to this problem, which mitigates the need for regularly flooding the network. Additionally, the concept of blind forwarding is utilised for its low overhead and high reliability through its inherent route diversity. The key contribution of this thesis is in three novel routing protocols, which use the aforementioned principles. The first protocol, PHASeR, uses a hop-count metric and encapsulates data from multiple nodes in its packets. The hop-count metric was chosen because it is simple and requires no additional hardware. The inclusion of encapsulation is intended to enable the protocol to cope with network congestion. The second protocol, LASeR, utilises location awareness to maintain a gradient and performs no encapsulation. Since many applications require location awareness, the communications systems may also take advantage of this readily available information and it can be used as a gradient metric. This protocol uses no encapsulation in order to reduce delay times. The third protocol, RASeR, uses the hop-count metric as a gradient and also does not perform encapsulation. The reduced delay time and the relaxed requirement for any existing method of location awareness makes this the most widely applicable of the three protocols. In addition to analytical expressions being derived, all three protocols are thoroughly tested through simulation. Results show the protocols to improve on the state-of-the-art and yield excellent performance over varying speeds, node numbers and data generation rates. LASeR shows the lowest overhead and delay, which comes from the advantage of having available location information. Alternatively, at the expense of increased overhead, RASeR gives comparatively high performance metrics without the need for location information. Overall, RASeR can be suitably deployed in the widest range of applications, which is taken further by including four additional modes of operation. These include a supersede mode for applications in which the timely delivery of the most recent data is prioritised. A reverse flooding mechanism, to enable the sink to broadcast control messages to the sensor nodes. An energy saving mode, which uses sleep cycles to reduce the networks power consumption, and finally a pseudo acknowledgement scheme to increase the reliability of the protocol. These additions enable RASeR to satisfy the needs of some of the most demanding MWSN applications. In order to assess the practicality of implementation, RASeR was also evaluated using a small testbed of mobile nodes. The successful results display the protocols feasibility to be implemented on commercially available hardware and its potential to be deployed in the real world. Furthermore, a key issue in the real world deployment of networks, is security and for this reason a fourth routing protocol was designed called RASeR-S. RASeR-S is based on RASeR, but introduces the use of encryption and suggests a security framework that should be followed in order to significantly reduce the possibility of a security threat. Whilst the main focus of this work is routing, alternative MAC layers are assessed for LASeR. Unlike the other two protocols, LASeR uses available location information to determine its gradient and as such, it is not reliant on the GTDMA MAC. For this reason several MAC layers are tested and the novel idea of dedicated sensing slots is introduced, as well as a network division multiple access scheme. The selected and proposed MACs are simulated and the GTDMA and two proposed protocols are shown to give the best results in certain scenarios. This work demonstrates the high levels of performance that can be achieved using gradient orientated routing in a mobile network. It has also shown that the use of a GTDMA MAC is an efficient solution to the gradient maintenance problem. The high impact of this work comes from the versatility and reliability of the presented routing protocols, which means they are able to meet the requirements of a large number of MWSN applications. Additionally, given the importance of security, RASeR-S has been designed to provide a secure and adaptable routing solution for vulnerable or sensitive applications.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:701608 |
| Date | January 2016 |
| Creators | Hayes, T. |
| Publisher | University of Sussex |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://sro.sussex.ac.uk/id/eprint/63977/ |
Page generated in 0.0025 seconds