Spelling suggestions: "subject:"emobility, handover, MRO, CHO, 5G, 15gave"" "subject:"emobility, handover, MRO, CHO, 5G, 4have""
1 |
Mobility Management in 5G Beamformed SystemsKarabulut, Umur 24 November 2021 (has links)
The number of subscribers and use cases of mobile communication networks are expanding expeditiously with the evolution of technology. The available spectrum in lower frequency ranges does not meet the unprecedented increase in demand for user data throughput in mobile networks. Facing the problem of limited spectrum in traditional cellular bands that are below 6 GHz, Millimeter Wave (mmWave) frequency bands are being standardized for the 5th Generation (5G) mobile networks as a promising means for handling the unprecedented data traffic surge. Enabling higher carrier frequencies introduces new channel conditions. Propagating signals are exposed to higher diffraction loss and are highly susceptible to blockage caused by surrounding objects, which leads to rapid signal degradation and challenges user mobility. On the other hand, higher carrier frequencies enable the deployment of many small-sized antennas that are used for directional signal transmission, resulting in beamforming gain.
In recent studies, a conditional handover procedure has been adopted for 5G networks to enhance user mobility robustness. Besides, contention-free random access procedure has been defined for beamformed systems aiming at minimizing the signaling and service interruption time caused by the random access procedure. An improper configuration of the mobility parameters, e.g., handover preparation and execution offsets, access beam selection threshold of random access procedure, leads User Equipments (UEs) to experience Handover Failures (HOFs) and Radio Link Failures (RLFs), and causes unnecessary signaling and inefficient resource utilization in the network. Each cell border has unique propagation characteristics and user mobility pattern, and, therefore, mobility parameters should be configured for each cell border individually. Moreover, mobility parameters should be updated for dynamic propagation environment (e.g., construction of buildings, seasonal changes in the vegetation) and for temporal mobility patterns. Considering the individual cell border configuration, temporal adaptation of the mobility parameters, and ultra-dense deployment, optimization of the conditional handover and random access parameters is a complex task that cannot be carried by human interaction. Therefore, an automatic optimization of the parameters is needed where the network collects statistics of the mobility events and adjusts the parameters autonomously.
To investigate user mobility under these new propagation conditions, a proper model is needed that captures spatial and temporal characteristics of the channel in beamformed networks. Current channel models that have been developed for 5G networks are too detailed for the purpose of mobility simulations and lead to infeasible simulation time for most user mobility simulations. In this work, a simplified channel model is presented that captures the spatial and temporal characteristics of the 5G propagation channel and runs in feasible simulation time. To this end, the coherence time and path diversity originating from a fully fledged Geometry-based Stochastic Channel Model (GSCM) are analyzed and adopted in Jake’s channel model with reduced computational complexity. Furthermore, the deviation of multipath beamforming gain from single ray beamforming gain is analyzed and a regression curve is obtained to be used in the system-level simulations.
In a typical system-level mobility simulator, the average downlink signal-to-interference and noise ratio (SINR) is used for RLF detection and throughput calculation. In addition to the channel model, models of desired and interfering signals are formulated first, by considering the impact of antenna beamforming, and a closed-form expression of average downlink SINR is derived by taking into account the user and beam scheduling probabilities. Then, an accurate approximation of the average downlink SINR with low computational complexity is presented, for 5G networks where the base station forms multiple beams. In addition, an SINR model is derived for both strict and opportunistic resource-fair scheduler, where the latter targets a higher utilization of radio resources when multiple beams are scheduled simultaneously.
The mobility performance of conditional handover and contention-free random access are investigated by using the proposed channel and SINR models. Besides, a resource efficient random access procedure is proposed that aims at maximizing the utilization of contention-free random access resources. Moreover, simple, yet, effective decision tree-based supervised learning method is proposed to minimize the HOFs that are caused by the beam preparation phase of the random access procedure. Similarly, a decision-tree-based supervised learning method is proposed for automatic optimization of the conditional handover parameters. In addition, enhanced logging and emergency reporting methods are introduced first time in this study to mitigate the cell detection problems that are caused by rapid signal degradation.
Results show that the optimum operation point of random access (in terms of minimizing the HOFs and maximizing the random access resource utilization) is achievable with the proposed learning algorithm for random access procedure in conditional handover. Results also show that the mobility performance of conditional handover is improved by automatic optimization of the handover parameters. In addition, the proposed enhanced logging and emergency reporting methods mitigate the mobility problems related with cell detection and further improve the mobility performance in combination with the decision-tree-based supervised learning methods.
|
Page generated in 0.4472 seconds