Dynamic Resource Configurations for the Convergence of Optical and Wireless Networks

Introduction

In the current landscape of Mobile Communication Networks, evolved radio access technologies, such as Long Term Evolution (LTE) (Ekstrom, et al., 2006; Hadden, 2009), offer all the more higher bitrates to mobile end users, providing support for a range of resource-demanding applications, such as IPTV, media streaming, VoIP, social network applications and gaming (Zahariadis, Grüneberg, & Celetto, 2011; Amram, et al., 2011). In parallel to this evolution of the mobile world, we also witness the evolution of Passive Optical Networks (PONs) (Kramer & Pesavento, 2002; Byun, Nho, & Lim, 2003; Yang C., 2007) in the fixed access world. Optical networks are expanding, and their use, not just in the core and backbone parts of the communication infrastructure, but also towards the last-mile and the premises, is beginning to gain ground.

In this context, one can imagine that the full potential of the communication infrastructures can be unleashed, in a cost-effective way, by enabling a smart convergence between the evolved access of the mobile world and the PONs of the fixed world. To achieve this, the notion is clear: Utilize the optical network infrastructure not just as a backbone or solely for serving residential or business customers, but also as a backhaul for base stations of mobile network operators. Indeed, using a PON as a high-capacity medium shared among different mobile base stations (e.g., LTE eNodeBs) as well as other customers (business or residential) offers significant advantages:

• •

  Enhanced support at the backhaul for the increasing capacities of the mobile radio access interfaces;

• •

  Economies of scale, since more customers can be served through the same PON, making PON-related deployment investments more attractive.

At the same time, nonetheless, significant challenges arise: Although the PON is a medium of high capacity, its sharing (namely, at the OLT – Optical Line Terminal level (Amemiya, Imae, Fujii, Suzuyama, & Ohshima, 2005)) among multiple end-users with different and fluctuating needs implies that smart bandwidth management mechanisms should be in place, and associated Service Level Agreements (SLAs) should exist, making sure that the high –but restricted– capacity of the PON is continuously utilized in an efficient and fair manner. This becomes a necessity particularly for the case of mobile base stations acting as clients/customers of the PON. We will see that, in order to address this necessity, the introduction of cognitive technologies at the backhaul segment is in order. This comes as a natural complement to existing and future cognitive solutions that focus instead on the radio access part of wireless communications.

This chapter proposes a novel converged network infrastructure that considers a passive optical network (PON) (Lam, 2007) as the backhauling solution for multiple next generation telecommunications networks, as depicted in Figure 1. Contrary to the existing fixed capacity transmission network model, the proposed solution takes into account time varying bandwidth requirements and suggests an efficient learning-based mechanism for resource optimization and flexible service provisioning. Through an integrated management platform, suitable cognitive mechanisms are deployed towards the dynamic reconfiguration and automatic and proactive adaptation to network needs. The implementation of appropriate traffic management entities in both the wireless base station and the optical backhauling segment can guarantee high quality of service and end-to-end efficiency.

Figure 1.

Envisioned converged network infrastructure