Cooperative opportunistic data delivery solutions for challenging wireless networks

Abstract

Advances in wireless communications and networking technologies are rapidly making the anywhere anytime wireless access a reality. Wireless ad hoc networks (WANETs) have been largely deployed due to its several promising benefits over the wired networks. The deployment of WANET is in various forms which include mobile ad hoc network (MANET), vehicular ad hoc network (VANET), opportunistic network or delay tolerant network (DTN), and mobile social networks. Disaster recovery communications, military/tactical communications, mobile social networks, vehicle-to-any (V2X) communications, mobile sensor networks are some major applications of WANET which have been widely deployed today. Data delivery in WANET, especially in mobile and hostile environments, poses great challenges due to the infrastructure-less environment, mobility, unreliable wireless link nature, effect of terrain, and distributed nature of wireless devices. In this thesis, we propose data delivery solutions for challenging wireless networks with major focus on opportunistic networks. We exploit adaptive data forwarding strategies, cooperation opportunities, open air wireless medium nature, communities and ties in social network nature, and multiple network interfaces to tackle such challenges and improve the system performance. Firstly, we study the mobility models in WANET. The synthetic mobility models are widely used in the WANET research for network simulations and performance evaluations. Group mobility model is of particular interest as group motion occurs commonly in scenarios where WANETs are deployed, such as disaster recovery operations, military operations, and mobile social networks. We provide a systematic and critical review of recent group mobility models proposed for WANETs. Secondly, we propose a data delivery solution which aims to solve the long delivery delay problem and the information epidemics problem in store-carry-forward routing in opportunistic networks. The proposed solution comprises two main algorithms: store-carry-cooperative forward routing and information epidemic control. The susceptible-infected-recovered model is used to study the effectiveness of the proposed mechanism. Extensive network performance evaluation is conducted under a wide range of scenarios which include fading environments, obstacle-constrained environments, and mobile social network environments. We show that 1) the information epidemics control mechanism provides higher vaccination rate and recovery rate; 2) proactive replication incurs a number of unnecessary transmissions; 3) monitoring the vicinity and exploiting the opportunity shorten the data delivery delay; and 4) with the integrated solution, a robust data delivery is achieved and a substantial amount of unnecessary transmissions are well deterred. Thirdly, we propose a data delivery solution which aims to enhance the system throughput for multi-hop wireless networks under multiple cross traffic flows and unreliable wireless channel. Cooperative forwarding has shown a substantial network performance improvement compared to traditional routing in multi-hop wireless network. To further enhance the system throughput, especially in the presence of highly congested multiple cross traffic flows, a promising way is to incorporate the multi-radio multi-channel (MRMC) capability into cooperative forwarding. In this work, we propose a new cooperative forwarding solution for multi-radio multi-channel multi-flow wireless networks. The performance is studied in highly congested network. Performance evaluation shows that the proposed solution reduces about 97% of the delay and achieves about 200% of throughput improvement compared to the single-radio single-channel case. Finally, we propose a data delivery solution for opportunistic networks from social network perspective. Given the communities and centralities nature in social network, it assures to exploit receiver diversity in such network environment, and therefore be able to maximize the given contact opportunity for the reliable data transfer. In addition, we claim that having this community nature also enables information epidemics control in such network more manageable by using efficient vaccination and recovery scheme. Motivated by these facts and the previous proposed adaptive, cooperative protocols and information epidemic control mechanism, we propose a third solution called cooperation-aided data delivery solution for opportunistic networks. The evaluation is conducted with varying network connectivity ranging from highly dense to highly sparse environments and under the presence of probabilistic propagation channels. We show that the proposed solution substantially improves the data delivery performance (i.e. higher delivery ratio while cutting down 4% to 85% of the delivery delay, 86% to 93% of the data overhead/retransmissions, and 86% to 97% of the control overhead) compared to the existing highest success rate protocol.