Workshop on Small Cell and 5G Networks (SmallNets)

Cellular traffic demand has been increasing exponentially over the past years, and current traffic forecasts indicate that this trend will continue with an expected 10x increase in the demand within the next 5 years. However, average revenues per user are declining, and energy consumption of networks already accounts for 1-2% of a country's total energy consumption. As a result meeting the future capacity demand with traditional macrocellular networks is infeasible. In recent years it has been widely recognized that deploying small cells are a key component to addressing the capacity problem, and today the number of deployed small cells has already exceeded the number of macrocells worldwide. In this presentation the limits of scaling capacity further by densification and alternative means such as using more spectrum and more antennas are explored. It is shown that we are still far from reaching the limits of densification. However, with increasing the number of cells, cost-effective deployment and management becomes paramount. Therefore, self-optimization is a key enabler for large scale deployments. Joint optimization of CellID, coverage and idle modes in a multi-vendor scenario is discussed as one example. Finally the impact of small cells on the energy efficiency of networks is discussed, and it is shown that when efficient idle modes are implemented, energy consumption is not a limiting factor anymore.

The ongoing explosive increase in the demand for video content in wireless networks requires new architectures to increase capacity without excessive costs. The talk will present a new architecture for solving this problem, exploiting a special feature of video viewing, namely asynchronous reuse. The approach is based on (i) distributed caching of the content in femto-basestations with small or non-existing backhaul capacity but with considerable storage space, called helper nodes, and/or (ii) usage of the wireless terminals themselves as caching helpers, which can distribute video through device-to-device communications. The talk will discuss the fundamental principles, scaling laws for the throughput, as well as practical implementation considerations. The new architecture can improve video throughput by one to two orders-of-magnitude.

Fifth generation cellular systems are going massive in the number of antennas at the base station. But will this happen at sub-6 GHz or millimeter wave frequencies? This presentation compares massive MIMO at sub-6GHz and millimeter wave frequencies. Based on stochastic geometry, a common mathematical framework is proposed to analyze SINR and rate distributions for downlink and uplink in both systems. Differentiating features between sub-6GHz and millimeter wave systems are incorporated by particular modeling assumptions. The analysis of sub-6GHz systems shows that to maintain the same SINR and rate distribution per user, the number of base station antennas should scale superlinearly with the number of scheduled users in a cell, as a function of the path loss exponent. Numerical results for the millimeter wave studies indicate that due to the presence of building blockages, a base station deployment is required to achieve a fair SINR coverage. The comparison concludes that while the benefits of going massive are confirmed for both systems, the advantage goes to sub-6 GHz for sparse deployments and to millimeter wave for dense deployments.

The panel will discuss the technologies that are bound to become an integral part of 5G networks. In particular, it will discuss whether ultra dense networks will be the foundations of 5G while shedding light on which technologies will survive the 5G hype.