Implementação de uma rede endereçada por interesses em nível de Kernel

• Autor(a) principal: Gondolfo, Danilo Egêa
• Data de Publicação: 2017
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Estadual de Maringá (RI-UEM)
• Texto Completo: http://repositorio.uem.br:8080/jspui/handle/1/2530
• topic: Rede Ad Hoc centrada em interesses (Radnet); Protocolo REPA; FreeBSD; Redes Ad Hoc; Roteamento em redes Ad Hoc; Radnet; REPA; Roteamento; Brasil.; Radnet; REPA; FreeBSD; Ad Hoc networks; Routing; Brazil.; Ciências Exatas e da Terra; Ciência da Computação

Ad Hoc networks can be used in a lot of situations, for example in environments where there is a precarious network infrastructure or even an inexistent infrastructure. In this scenario, an Ad Hoc network can be created amongst personal communication devices, as smartphones, so that users can interact and share information. The Radnet (Interest-centric Ad Hoc Network) is an Ad hoc network in which the messages are routed between devices based on characteristics and interests of the users. This work has as main objectives the propose and the implementation of the Radnet in the operating system's kernel level, with a simplified version with focus on the computational resources reduction and encryption support during the messages exchange. Two versions of the REPA protocol have been implemented in the core of the FreeBSD operating system and are used through the sockets API. The resources comsumption reduction is based on a simplified header version in which the interests are sent in a numeric format instead of a string. The kernel level implementation also enable messages processing with less overhead, which could drive to less energy consumption in embedded systems. The experiments showed that with the new header is possible: i)to reduce the CPU consumption, obtaining gains up to 42% on the amount of instructions performed per second; ii) to increase the message processing rate per minute in about 10%; and iii) to reduce the per message processing time in about 22%.