Date of Award

Summer 2008

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Electrical & Computer Engineering

Program/Concentration

Electrical Engineering

Committee Director

Linda L. Vahala

Committee Director

R. W. Lawrence

Committee Member

Sachin Shetty

Call Number for Print

Special Collections LD4331.E55 S5165 2008

Abstract

Wireless sensor networks consist of a group of small components with limited resources allocated to sense the environment and transmit data to the user or interpreter. This concept is technically evolved from Sensors that are basically used to sense the environment around and collect data based on the sensing application incorporated in them. An extension to functionality of the sensors leads to the concept of sensor networks, where, in addition to gathering information at individual sensors, the data is processed and transmitted to a central station or Base station. This concept of transmitting the information from one node to the other is known as networking. The transmission of data can be done either directly to the base station or by using intermediate nodes (multi­ hop) to reach the base station. As the sensors are tiny components, they are flexible enough to be deployed in space in any manner depending on the application and requirement. If data traffic is low, i.e sensing at the nodes is rarely expected, then the deployment of sensors is done sparsely; in the case of high traffic loads, i.e if the chances of expecting an event are high, the sensors are deployed densely.

During the transportation phase, when a sensor receives more packets than its buffer space can tolerate, the excess packets have to be dropped, and the energy consumed by upstream sensors on the packets is wasted. This state of buffer overflow in the network is called Congestion. Congestion in a network occurs when the offered load exceeds available capacity at any point in the network. Congestion can also deteriorate network throughput, and worst of all, it may cause the network to collapse.

In addition to data losses, it also accounts for energy waste in sensor networks, as the buffer overflow obviates the need to drop the packets destined to this node. Thus, the energy

used to forward these packets until this point in the network is undoubtedly wasted. Hence, the energy efficiency of the network is indirectly connected to congestion. The occurrence of this situation in a network needs to be eliminated for efficient data delivery to the base station, and thus technically challenges the need for a congestion control mechanism in wireless sensor networks.

The Single-hop Backpressure algorithm proposed in this thesis to control the congestion is completely dependent on the transport layer and network layer of the protocol stack. This greatly reduces packet losses in a network under varied traffic loads, thereby resulting in high packet delivery performance and, indirectly, in an energy efficient algorithm. A set of simulation experiments were performed considering a small test bed that consists of 50 nodes and results were plotted. The performance of the proposed algorithm was studied based on varying the topologies, traffic loads and a set of other system parameters.

Rights

In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).

DOI

10.25777/r7xv-n045

Download

Share

COinS