Identifier

etd-06222007-095426

Degree

Doctor of Philosophy (PhD)

Department

Computer Science

Document Type

Dissertation

Abstract

A wireless sensor network (WSN) is a collection of sensor nodes and base stations connected via wireless medium. It sends data collected from the nodes to the base stations for generating information. The size and low cost of the sensor nodes as well as the WSN's ability to connect without wired links are its key advantages which enable it to be deployed in hostile or inaccessible environments at low cost. However, WSNs suffer from high data loss due to the inherent weaknesses in a wireless transmission medium, transmission problems in hostile environments due to human interference, etc. and node failures due to limited energy of sensor nodes. Hence ensuring data transfer with minimum loss i.e. reliable data transfer is very important in WSNs. The amount of loss tolerated is application dependent.

We present a reliable protocol for data transfer from a base station to sensor nodes for time-critical applications in WSNs with zero tolerance for data loss. The protocol is based on hop-by-hop detection and recovery of lost data packets, out-of-sequence forwarding of packets and delayed request for missing packets at each node with non-acknowledgement of packets at each receiving node. We present a detailed analysis of the advantages of the key features of our protocol over other alternatives. The superiority of our protocol over an established protocol PSFQ is demonstrated via extensive simulations, in terms of both the delivery time of the entire data (sent from the base station to the sensor nodes) and the number of messages exchanged in the network during this process.

In addition, we present two methods, one, for ensuring that at least one packet is delivered to a node in non-acknowledgement based systems and another, for sending reports from destination nodes to the base station respectively.

We explore different methods for further improvement of protocol performance: (1) use of effective degree of a node in determining the optimum delay for requesting data packets missing at nodes, (2) variation of the priority order for sending different types of messages at nodes and, (3) selective response to requests for packets at nodes.

Date

2007

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Sukhamay Kundu

DOI

10.31390/gradschool_dissertations.904

Download

Share

COinS