Mobility management applied to the traditional architecture of the Internet has become a great challenge because of the exponential growth in the number of devices that can connect to the network. This article proposes a Software-Defined Networking (SDN)-based architecture, called SDN-DMM (SDN-Distributed Mobility Management), that deals with the distributed mode of mobility management in heterogeneous access networks in a simplified and efficient way, ensuring mainly the continuity of IP sessions. Intent-based mobility management with an IP mapping schema for mobile node identification offers optimized routing without tunneling techniques, hence, an efficient use of the network infrastructure. The simplified mobility control API reduces both signaling and handover latency costs and provides a better scalability and performance in comparison with traditional and SDN-based DMM approaches.An analytical evaluation of such costs demonstrated the better performance of SDN-DMM, and a proof of concept of the proposal was implemented in a real environment.The IETF IP mobility management standards, as MIPv6 6 an PMIPv6, 7 depend on central units that manage both control and data traffic, elaborated according to the traditional routing of IP packets. They also pose problems, as suboptimized routing, low scalability, processing overload in core network devices, and low granularity in the mobility management service.The suboptimized routing occurs mainly because of the use of tunneling techniques. When the mobile node (MN) performs a handover, the traffic destined to it is sent to an anchor point for encapsulation and delivered through an IP tunnel. Regardless of the MN location, the traffic always flows first to the mobility anchor and then to the MN through the IP tunnel, which decreases the goodput because of the packet overhead caused by such transportation, and increases communication latency, because of the use of a longer path.The establishment of the IP tunnel in control plane operations by a central unit, ie, mobility anchor, and forwarding of traffic to the MN in data plane operations cause low scalability and processing overload for devices, as routers, since resources are shared between packet forwarding engine and mobility management tasks.Additionally, the use of heterogeneous access networks (HetNets) imposes other difficulties to such a management, as the integrated management of resources and the soft transitions among networks.The mobility management faces a constant challenge regarding the communication network efficiency with no increase in its complexity. The addition and use of new protocols, signaling messages, or processes that cause overhead because of the encapsulation and control traffic are examples of how the mobility management directly impacts on the Operational Expenditure (OPEX) and Capital Expenditure (CAPEX) of a communication network.An alternative for dealing with the intrinsic problems of centralization and costs in the mobility management is the new concept, called distributed mobility ...

The increasing demand for data transmission resources to be handled by next generation cellular networks has led to the emergence of new technologies, such as Device-to-Device (D2D) communications and software-defined networks (SDN). D2D expands the use of resources from a location perspective and SDN enables an efficient management and control of the available resources. This article addresses a study of the influence of D2D communications management on a Long Term Evolution (LTE) network with SDN controllers regarding the obtaining of their necessary number and best location within the cellular infrastructure. The controller placement problem was modeled as an optimization problem and solved by the ant colony system with external memory (ACS-EM) algorithm. The proposed algorithm was compared with a particle swarm optimization (PSO)-based algorithm and its effectiveness has been validated.INDEX TERMS Ant colony system with external memory, controller placement problem, D2D, software defined networking.

No abstract

The development of new applications and mobile devices has increased the data traffic in cellular networks and required throughput improvements, which can be achieved through heterogeneous access networks, where multimodal user devices able to establish simultaneous connections with different access networks are considered. Such connections and data offloading between Wi‐Fi and cellular networks provide load balancing and increase the throughput of heterogeneous cellular networks. Offloading can be performed at the network layer through the move of Internet Protocol (IP) flows, and mobility management solutions can be used as a basis for this purpose. This article proposes an architecture called Seamless IP Flow and Distributed Mobility Management (SIFDMM), which is based on a distributed mobility management solution and improves heterogeneous network throughput by moving IP flows between Wi‐Fi and LTE networks. SIFDMM mitigates congestion in the networks and causes no problem of single point of failure, scalability, and reliability of centralized solutions. An analytical comparison was performed for an effective demonstration of the throughput improvement in heterogeneous networks.

Resumo-Para emprego da rede LTE-Advanced (LTE-A), a comunicação "device-to-device" é uma tecnologia candidata a permitir um adequado uso de recursos, considerando perspectiva baseada na localização do usuário e em serviços baseados em proximidade. Por outro lado, redes definidas por software favorecem a melhoria da eficiência do gerenciamento e controle dos recursos de rede. Este artigo trata o problema de alocação de controladores SDN visando o adequado gerenciamento de comunicações D2D em uma rede LTE-A. Para solução desse problema, foi adotada abordagem baseada em memória externa do algoritmo de colônia de formigas, com resultados similares aos de busca exaustiva. Palavras-Chave-D2D, SDN, colonia de formigas, alocação de controladores Abstract-For the use of the LTE-Advanced network (LTE-A), device-to-device communication is a candidate technology to allow an adequate use of resources, considering user location and proximity-based services perspective. On the other hand, software-defined networks improve the efficiency of the management and control of network resources. This article addresses the problem of allocating SDN controllers for the proper management of D2D communications over an LTE-A network. To solve this problem, the ant colony algorithm based on external memory approach was adopted, with results similar to exhaustive search.