Abstract/Summary

Long Term Evolution (LTE) is the Fourth Generation (4G) mobile broadband technology. Third-Generation Partnership Project (3GPP) has finalized its standardization in Release 8 technical specifications (R8). As users’ demand for higher data rate continues to rise, LTE and its ability to cost effectively provide fast, highly responsive mobile data services, a scalable bandwidth and a reduced latency will become ever more important. Nevertheless, The Evolved Packet Core (EPC) of the LTE is all-IP Packet-Switched (PS) core networks that lack a native support for Circuit-Switched (CS) services. This introduces a problem of how to provide PS voice services in LTE networks and how to support a voice service quality of at least efficient as a voice service in 3G mobile technologies and the service continuity of voice service to different radio access network toward a heterogeneous wireless network. This thesis investigates the role of the Radio Recourse Management (RRM) techniques in improving the end-user experience and the service quality of Voice over LTE (VoLTE) application service. The thesis explores the possible proposed solutions to provide voice over LTE networks and suggests the required recommendations in regards to voice over LTE deployment based on one of the proposed techniques. A baseline simulation of LTE network is designed to give a clear understanding of the state of the art of VoLTE QoS and QoE, which then improved to cover scenarios where the wireless network congested under different load conditions, different types of traffic and different LTE bandwidths. In the literature, most of the studies were focused on the single Radio Access Technology (RAT) of Evolved Universal Terrestrial Radio Access Network (E-UTRAN), where the end-to-end performance of VoLTE of multi-RAT over a Heterogeneous wireless Network (HetNet) has been less investigated. The future of wireless networks is going to be heterogeneous. The continuity of the service when LTE users with active real time application services such as VoLTE running on their smart phones move between different mobile networks is the major concern in the future of these networks. In addition, there is a common issue of a lack of considering high mobility, and frequent and unnecessary handovers in these networks. VoLTE end-to-end delay is an important service quality factor, which needs to be maintain carefully, otherwise can severely affect the voice user experience. By exploiting the RRM techniques, the impact of this delay is limited, and the overall LTE network performance is further enhanced. To begin with, this study proposes a new, highly flexible Delay-Aware QoS Scheme (DAQS) for 3GPP LTE networks which is then critically simulated, and examined. The proposed scheme uses the idea of QCI-DSCP (QoS Class Identifier-Differentiated Services Code Point) mapping in the transport layer along with the queuing, scheduling on the LTE backhaul network to manage different application services with different requirements and provide an acceptable level of fairness between these services. The results reveal that the overall VoLTE end-to-end delay is remarkably reduced by approximately 35% compared to the baseline congested network scenario. FTP download and upload response time have both decreased while video conferencing traffic received increased by 30% on the price of a slight increase in LTE transport complexity. Handover delay is another factor, which investigated in this thesis. Packet transmission is disabled during the hard handover and network faces extra delay due to the interruption time that included in the process. The quality of service provided to the end-user is temporarily impaired and handover prediction is one emerging technique to mitigate this in Femtocell Home eNBs (HeNBs) with the unavoidable frequent handovers. This study proposes a mobility prediction scheme to help reduce VoLTE end-to-end delay and handover delay in the worst-case scenario of a mixed of femtocells and macrocells environment and under different UE velocities. The mobility prediction works based on user’s path and Reference Signal Received Power and Reference Signal Received Quality (RSRP and RSRQ) measurements. The results after applying the proposed prediction scheme show that the average handover delay is reduced significantly. Handover delay reduced by around 29% and VoLTE end-to-end delays kept within the standard recommendations. Regular movements resulted in 28% VoLTE end-to-end delay reduction and helped to maintain the desired VoLTE QoS. This study further proposes a novel DL hybrid-scheduling algorithm of Semi-Persistent Scheduling (SPS) and Talk-Spurt Scheduling (TSS) to reduce delay, Packet Loss Ratio (PLR) and increase capacity of VoLTE. This scheduling includes a novel work in terms of it is one of the few studies that introduce a complete implementation of end-to-end QoS over HetNet. The performance of the proposed scheme is compared with the performance of two relevant and well-known DL packet scheduling techniques, where the proposed DL scheduling algorithm outperforms them and provides a lower packet loss ratio, higher capacity and reduces VoLTE end-to-end delay accordingly.