Conflicts-free auto-configuration of Physical Cell Identities (PCIs) is a main challenge in LTE systems. This challenge is the focus of this chapter, which provides a survey of present PCI assignment as well as allocation schemes and evaluates by means of simulative studies the performance of four well-known distributed PCI assignment approaches. The simulation is achieved applying two types of scenarios, namely static and dynamic. Static scenario stands for an already deployed LTE network that is a subject to a configuration using a specific PCI assignment approach, while the dynamic scenario tries to mimic the situations LTE networks are expected to face in real-life scenarios. Our results show that network dynamics heavily affect the performance of PCI assignment approaches, and even may transit LTE networks to an unstable state. So, it is essential to consider network dynamics by concept in any PCI assignment scheme to be developed.
Top1. Introduction
Ubiquitous access to information anywhere, anytime and anyhow is a main feature of future communication networks which experience a tremendous development clearly evident from the wide variety of new services and broadband applications way beyond the classical application, i.e. telephony (Diab & Mitschele-Thiel, 2014). The Internet itself emerges towards what is today termed “the Internet of things”, which assumes that users of future communication networks are objects, not only humans. Each object is embedded with sensors and is able to communicate. In other words, the physical world itself including communication networks, humans and objects is becoming a type of huge information system. To satisfy always increasing demands of users within such communication and information systems, the better support of mobility, the provision of services at low costs, etc. are the drivers standing behind the standardization of 4th Generation (4G) mobile communication networks, termed Long Term Evolution (LTE). Researchers expect that the number of LTE subscribers will reach the base of 3G/UMTS networks (1.087 million subscribers) by the year 2015 (Garza, Ashai, Monturus & Syputa, 2010).
A major challenge in LTE networks is a cost-effective network operation, which necessitates carefully controlling the Operational Expenses (OPEX). Cost-effectiveness gets more crucial when considering the additional costs resulting from the expected affordable operation of overlaid multi-standard networks. 3G mobile communication technologies have shown that network-related OPEX costs contribute by approximately 30% to the total cost, see (Motorola, 2009). Note that OPEX costs strongly depend on the solutions implemented. This means that the reduction of OPEX costs implies the provision of adequate solutions. Self-Organization plays in this context a major role since it significantly contributes in reducing OPEX costs, see (Østerbø & Grøndalen, 2012) and (Diab & Mitschele-Thiel, 2014).
So, the release being standardized by the 3G Project Partnership (3GPP) (3GPP, 2015) with self-organization capabilities is referred to as LTE-Advanced (Dahlman, Parkvall & Skoeld, 2011). This standard aims at minimizing OPEX costs, while optimizing network performance as a whole. To further contribute to Self-Organizing Networks (SONs), the Next Generation Mobile Networks (NGMN) alliance (NGMN, 2015) was also constructed. This alliance has summarized SONs requirements in a number of operation use cases, see (Ramiro & Hamied, 2012) and (Lehser, 2008). One of the use cases is an auto-configuration of evolved NodeBs (eNBs) with conflicts-free Physical Cell Identities (PCIs). This chapter focuses on this use case, addresses the problems faced in its context and highlights the state of art. The major contribution of this chapter is the investigation of the impact of network dynamics on the performance of four well-known distributed PCI assignment schemes, namely the Stable Graph Coloring (SGC) approach, the GC approach, the LTE Standard Distributed (LSD) proposal and the Nokia-Siemens (NS) solution. This investigation is of a great importance since there are no studies until now that tried to analyze this issue, although it is accepted that LTE networks will have topologies of a dynamic nature.
The reminder of the chapter is organized as follows: section 2 addresses the problems faced in the context of PCI assignment and allocation. Present PCI assignment and allocation techniques are overviewed in sections 3 and 4, respectively. These sections finalize with qualitative comparisons between overviewed techniques. Simulation assumptions, scenario and results are provided in section 5 followed by a conclusion and an outlook in section 6.