Advanced technologies enabling unlicensed spectrum utilization in cellular networks

As the rapid progress and pleasant experience of Internet-based services, there is an increasing demand for high data rate in wireless communications systems. Unlicensed spectrum utilization in Long Term Evolution (LTE) networks is a promising technique to meet the massive traffic demand. There are two effective methods to use unlicensed bands for delivering LTE traffic. One is offloading LTE traffic toWi-Fi. An alternative method is LTE-unlicensed (LTE-U), which aims to directly use LTE protocols and infrastructures over the unlicensed spectrum. It has also been pointed out that addressing the above two methods simultaneously could further improve the system performance. However, how to avoid severe performance degradation of the Wi-Fi network is a challenging issue of utilizing unlicensed spectrum in LTE networks. Specifically, first, the inter-system spectrum sharing, or, more specifically, the coexistence of LTE andWi-Fi in the same unlicensed spectrum is the major challenge of implementing LTE-U. Second, to use the LTE and Wi-Fi integration approach, mobile operators have to manage two disparate networks in licensed and unlicensed spectrum. Third, optimization for joint data offloading to Wi-Fi and LTE-U in multi- cell scenarios poses more challenges because inter-cell interference must be addressed. This thesis focuses on solving problems related to these challenges. First, the effect of bursty traffic in an LTE and Wi-Fi aggregation (LWA)-enabled network has been investigated. To enhance resource efficiency, the Wi-Fi access point (AP) is designed to operate in both the native mode and the LWA mode simultaneously. Specifically, the LWA-modeWi-Fi AP cooperates with the LTE base station (BS) to transmit bearers to the LWA user, which aggregates packets from both LTE and Wi-Fi. The native-mode Wi-Fi AP transmits Wi-Fi packets to those native Wi-Fi users that are not with LWA capability. This thesis proposes a priority-based Wi-Fi transmission scheme with congestion control and studied the throughput of the native Wi-Fi network, as well as the LWA user delay when the native Wi-Fi user is under heavy traffic conditions. The results provide fundamental insights in the throughput and delay behavior of the considered network. Second, the above work has been extended to larger topologies. A stochastic geometry model has been used to model and analyze the performance of an MPTCP Proxy-based LWA network with intra-tier and cross-tier dependence. Under the considered network model and the activation conditions of LWA-mode Wi-Fi, this thesis has obtained three approximations for the density of active LWA-mode Wi-Fi APs through different approaches. Tractable analysis is provided for the downlink (DL) performance evaluation of large-scale LWA networks. The impact of different parameters on the network performance have been analyzed, validating the significant gain of using LWA in terms of boosted data rate and improved spectrum reuse. Third, this thesis also takes a significant step of analyzing joint multi-cell LTE-U and Wi-Fi network, while taking into account different LTE-U and Wi-Fi inter-working schemes. In particular, two technologies enabling data offloading from LTE to Wi-Fi are considered, including LWA and Wi-Fi offloading in the context of the power gain-based user offloading scheme. The LTE cells in this work are subject to load-coupling due to inter-cell interference. New system frameworks for maximizing the demand scaling factor for all users in both Wi-Fi and multi-cell LTE networks have been proposed. The potential of networks is explored in achieving optimal capacity with arbitrary topologies, accounting for both resource limits and inter-cell interference. Theoretical analyses have been proposed for the proposed optimization problems, resulting in algorithms that achieve global optimality. Numerical results show the algorithms' effectiveness and benefits of joint use of data offloading and the direct use of LTE over the unlicensed band. All the derived results in this thesis have been validated by Monte Carlo simulations in Matlab, and the conclusions observed from the results can provide guidelines for the future unlicensed spectrum utilization in LTE networks.