T. Hayes

Introduction

Wireless Sensor Networks (WSNs) have only recently become an area of interest to the research community due to the fact that previously the technology was not available to produce cheap nodes with the ability to communicate wirelessly whilst still being power efficient. These nodes can be distributed in an ad-hoc manner, in potentially harsh conditions, and are then expected to organise themselves in such a way that gathered data is transmitted from source to sink.

The focus of this work is in the area of Mobile Wireless Sensor Networks (MWSNs), in which the nodes are assumed to be capable of movement. Due to this added mobility many of the current protocols designed for WSNs are inadequate.

Challenges

Specifically in the routing layer the mobility of nodes presents a big challenge, this is because any path from sensor to sink may change or be lost completely at any time. The subject of routing in mobile networks has been predominately investigated by researchers looking at mobile ad-hoc networks (MANETs). Though the use of MANET protocols in MWSNs is possible it's potentially inefficient. This is because sensor networks are generally only required to maintain unidirectional communication between the sensors and the sink using relatively small packets of data.

Additionally, the MAC layer provides another set of challenges. These include trying to limit or even eliminate collisions, reduce access delay as well as suppress energy consumption.

It may be that the most efficient protocols are not the combination of separate layers but the fusion of layers. This fusion should allow every layer involved to improve its performance by utilising data made available by other layers.

Aims and Methods

The main aim of this project is to produce a set of routing protocols designed for mobile wireless sensor networks, which provide reliable and robust delivery of data in a variety of scenarios. The main areas of interest for the design of these protocols are:

• Cross-layer, specifically through the fusion of the routing layer and the MAC layer.
• Cooperative communications, in the utilisation of resources network wide rather than by a single node.
• Multipath, as this seems to be a reliable way of ensuring data successfully traverses the network.
• Flat routing (non-hierarchical), which gives each node the same status such that energy can be more evenly used over the whole network.
• Proactive routing, such that the nodes aren't required to discover a path every time data needs to be transmitted, which can significantly increase overhead.

Though these ideas are applicable to a large variety of communications systems, their use in a MWSN may be optimised beyond the general case.

Figure 1 shows the route of a packet from the sensor node, mobile_node_7, to the Sink. The blue arrows show the multipath effect and highlight the multipath diversity as the packet follows a varying number of paths to the sink. These are, in terms of hops, the fastest routes from source to destination. The red arrows show cooperative diversity, where nodes have overheard the transmission and decided to act as a relay. These routes are given secondary priority and add to the robustness of the protocol.

Figure 1. The route of a single packet through the network

These aims will be achieved by designing, simulating and then refining the protocols in the popular OPNET simulator (from which the screen shot used to generate figure 1 was taken).

Contact

For more information, contact details are available here.

Funding

This research is partly funded by: