Um esquema de seleção de rotas para o balanceamento de segurança e desempenho em redes em malha sem fio / Um Esquema de Seleção de Rotas para o Balanceamento de Segurança e Desempenho em Redes em Malha Sem Fio

AUTOR(ES)
FONTE

IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia

DATA DE PUBLICAÇÃO

09/05/2011

RESUMO

Wireless mesh networks (WMNs) have been used to support multimedia applications on different domains, such as health care and entertainment, that require high levels of security and performance simultaneously. Those networks comprise of a wireless backbone of routers that cooperate to provide data communication among mobile devices, such as notebooks, PDAs and smartphones, through multiple hops. This infrastructure allows a device to communicate with others beyond its wireless coverage range. WMNs are also self-configuring, being capable to detect new devices and automatically connect them to the wireless backbone. Moreover, the wireless backbone can automatically establish alternative routes to data forwarding in face of failures on specific routers. However, wireless medium sharing and cooperative communication allow the action of attacks, such as eavesdropping, selective packet dropping and blackhole, that aim to violate data confidentiality or disrupt the operation of network essential services, such as routing. Even in network environments without attackers, interferences among different data flows on the shared wireless medium decrease the capacity of routes. These interferences reduce the bandwidth on wireless links, increasing collisions and packet loss of applications. Therefore, security and performance aspects must be considered simultaneously in WMNs. This work proposes a route selection scheme, called CRAB (CRoss-layer and Adaptive scheme for Balancing performance and security on WMN data routing). The scheme aims to balance security and performance on routing service in WMNs, even when the network is under attack. CRAB aims also to support Quality of Service (QoS) of applications. Hence, it defines a new routing metric following a cross-layer approach that integrates network and link layers. The value of this routing metric is calculated using a Multiple Criteria Decision Making (MCDM) method that combines multiple security and performance criteria provided by intermediate nodes on routes through a cooperative procedure. Security criteria are provided by mechanisms representing preventive, reactive and tolerant defense lines, and the performance criterion represents interferences on the wireless links of a route. Based on the value of the routing metric, the scheme ranks the available routes between a source node and a destination node of a data flow, and selects the route that provides a balance between security and performance to the data transmission even in face of attacks. CRAB has been evaluated through simulations using two scenarios. These scenarios take into account different data traffic patterns from multimedia applications that require high levels of security and performance on data forwarding. In the first scenario, it has been considered a uncompressed videos traffic generated by applications such as videophone, for example. In the second scenario, it has been examined the traffic of compressed videos following the MPEG-4 specifications used by Internet Protocol TV (IPTV) applications, for example. In each scenario, it has been investigated the security and performance provided by CRAB with the variation on the percentage of malicious nodes launching packet dropping attacks on routing. Results show that CRAB balances security and performance on routing, even in face of those attacks. It improves the packet delivery ratio independently on the percentage of malicious nodes in the network with a low cost of latency. Even when attackers launch the blackhole attack, CRAB increases the packet delivery ratio in 16% to the MPEG-4 traffic, keeping this ratio higher than 90% to the uncompressed video. Such analyses are reinforced by the reduction on the amount of data packets discarded by attackers. CRAB reduces up to 73% the data packet loss from MPEG-4 traffic caused by the blackhole attack.

ASSUNTO(S)

redes de computadores - dissertaÇÕes sistema de comunicaÇÃo sem fio - dissertaÇÕes sistemas de informacao roteamento - dissertaÇÕes

ACESSO AO ARTIGO

Documentos Relacionados