The current Internet architecture was designed at the 1960s and 70s, to address communication needs of that time: sharing limited, expensive, and static computer resources. Since then, the Internet usage pattern has been shifted from conventional host-centric model to a flexible content-oriented model, in which users and contents are distributed and mobile. Internet of Things is becoming a new paradigm to describe global access to services and information offered by billions of heterogeneous devices, "things", ranging from resource-constrained to powerful devices in an interoperable way.
The first part of this dissertation studies the routing in Information-Centric Networks (ICN). ICN has been recently proposed and is inspiring the design of the future Internet architecture. The goal in these architectures is to provide a cost-efficient, scalable and mobile content distribution networking by adopting a content-based model of communication. ICN not only addresses the change in the Internet usage pattern but also matches the IoT applications, since they target data regardless of the identity of the object that stores or originates them.
In this dissertation, Named-data network (NDN) and Content-Centric Networking (CCNx) are presented and routing strategies for them are evaluated. A comprehensive performance evaluation is done through the simulation experiments. We enhanced the performance of link-state routing by introducing a new protocol called LSCR, link-state content routing protocol, a loop-free name-based routing algorithm that propagates link-state information selectively and provides multi-path routing to content that may be replicated in different locations. We also introduced the first content routing protocol based on the diffusing computation, DNRP. DNRP provides multiple loop-free routes to the nearest instances of a data using only distance information and without requiring periodic updates, knowledge of the network topology, or the exchange of path information.
MANET paves the way for the development of brand new IoT communication platforms with a high potential for a wide range of applications in different domains. Each layer in the design model require redefinition or modifications to function efficiently in MANETs every is mobile and usually has limited resources on computation, storage, power, etc.
Routing in Mobile Ad-Hoc Networks is studied in the second part of this dissertation. We introduce ODVR (Ordered Distance Vector Routing), that provides loop-free routes at every instant based solely on distances to destinations maintained by nodes and reference distances included in route requests. Routing state is established on demand by means of route requests stating the reference distance and replies from nodes that satisfy this distance. To make the routing more efficient in an IoT environment with a base or gateway nodes, we introduce ADRP, a hybrid routing algorithm that takes advantage of the strengths of reactive routing algorithms as well as the benefits of proactive ones. ADRP uses the same signaling for both reactive and proactive routing.