Document details

Um algoritmo de difusão para redes sem fios não infraestruturadas

Author(s): Ludovico, João Luís Lopes

Date: 2011

Persistent ID:

Origin: Repositório da Universidade de Lisboa

Subject(s): Broadcast; ad hoc networks; algorithm; context-aware


Broadcast is an operation widely used in ad hoc networks. In these networks, nodes only communicate with others within ones propagation radius, therefore these networks are characterized by the absence of an infrastructure that connects and provides services to all nodes. This way, in order for a message to be delivered to the entire network, it has to be relayed by its participants. Broadcast algorithms can be divided in two major groups: context-aware and contextoblivious. In the first group, nodes collect information about their environment and decide to retransmit or not a message according to this information. Some context-aware algorithms use geographical information in order to make these decisions. By using this information these algorithms try to achieve the maximum coverage with the minimum number of retransmissions. In the second group, the decision to retransmit or not a message is based solely on configuration parameters set at deployment time. However this information can lead to a waste in nodes resources or lower the delivery rate. The developed broadcast algorithm tries, like the ones that use geographical information,to maximize the coverage area while keeping a low number of retransmissions. In order to do this, the algorithm performs estimates, local to each node, about the proximity of each node. From these estimates, each node selects a subset of neighbors that he believes being in different regions in order to assure the broadcast of the message in all directions. However, these estimates do not require the usage of a GPS device.

Document Type Master thesis
Language Portuguese
Advisor(s) Miranda, Hugo Alexandre Tavares
Contributor(s) Ludovico, João Luís Lopes
facebook logo  linkedin logo  twitter logo 
mendeley logo

Related documents