Three use cases of multi-connectivity paradigm in emerging wireless networks

Abstract

As envisioned by global network operators, the increasing trend of data traffic demand is expected to continue with exponential growth in the coming years. To cope with this rapid increase, significant efforts from the research community, industry and even regulators have been focused towards improving two main aspects of the wireless spectrum: (i) spectrum capacity and (ii) spectral efficiency. Concerning the spectrum capacity enhancement, the multi-connectivity paradigm has been seen to be fundamentally important to solve the capacity problem in the next generation networks. Multi-connectivity is a feature that allows wireless devices to establish and maintain multiple simultaneous connections across homogeneous or heterogeneous technologies. In this thesis, we focus on identifying the core issues in applying the multi-connectivity paradigm for different use cases and propose novel solutions to address them.

Specifically, this thesis studies three use cases of the multi-connectivity paradigm. First, we study the uplink/downlink decoupling problem in 4G networks. More specifically, we focus on the user association problem in the decoupling context, which is considered challenging due to the conflicting objectives of different entities (e.g., mobile users and base stations) in the system. We use a combination of matching theory and stochastic geometry to reconcile competing objectives between users in the uplink/downlink directions and also from the perspective of base stations.

Second, we tackle the spectrum aggregation problem for wireless backhauling links in unlicensed opportunistic shared spectrum bands, specifically, TV White Space (TVWS) spectrum. In relation to this, we present a DIY mobile network deployment model to accelerate the roll-out of high-end mobile services in rural and developing regions. As part of this model, we highlight the importance of low-cost and high-capacity backhaul infrastructure for which TVWS spectrum can be exploited. Building on that, we conduct a thorough analytical study to identify the characteristics of TVWS in rural areas. Our study sheds light on the nature of TVWS spectrum fragmentation for the backhauling use case, which in turn poses requirements for the design of spectrum aggregation systems for TVWS backhaul. Motivated by these findings, we design and implement WhiteHaul, a flexible platform for spectrum aggregation in TVWS. Three challenges have been tackled in this work. First, TVWS spectrum is fragmented in that the spectrum is available in non-contiguous manner. To fully utilize the available spectrum, multiple radios should be enabled to work simultaneously. However, all the radios have to share only a single antenna. The key challenge is to design a system architecture that is capable of achieving different aggregation configurations while avoiding the interference. Second, the heterogeneous nature of the available spectrum (i.e., in terms of bandwidth and link characteristics) requires a design of efficient traffic distribution algorithm that takes into account these factors. Third, TVWS is unlicensed opportunistic shared spectrum. Thus, the coordination mechanism between the two nodes of backhauling link is essential to enable seamless channel switching.

Third, we study the integration of multiple radio access technologies (RATs) in the context of 4G/5G networks. More specifically, we study the potential gain of enabling the Multi-RAT integration at the Packet Data Convergence Protocol (PDCP) layer compared with doing it at the transport layer. In this work, we consider ultra-reliable low-latency communication (URLLC) as one of the motivating services. This work tackles the different challenges that arise from enabling the Multi-RAT integration at the PDCP layer, including, packet reordering and traffic scheduling.