Centralization of the baseband processing in radio access networks may reduce radio site operations costs, reduce capital costs, and ease implementation of multi-site coordination mechanisms such as coordinated multipoint transmission and reception (CoMP). However, the initial architecture proposals for a centralized Long Term Evolution (LTE) deployment using transport of radio samples require a high-bandwidth, low-latency interconnection network. This may be uneconomical, or it may only be cost effective for a limited number of sites. To mitigate that deficiency without sacrificing the benefits of centralized processing, we identified alternative interfacing options between central and remote units. To do so we analyzed possible splits of the Long Term Evolution (LTE) baseband processing chain for their bandwidth and latency requirements. Next, we analyzed the operational impacts of potential splits based on a number of criteria including ease of CoMP introduction, the possibility of realizing pooling gains, and the ability to update the system and introduce new features. Based on our results, we propose architectures that can leverage the benefits of centralization at a much-reduced cost.

Quantitative analysis of split base station processing and determination of advantageous architectures for LTE

The constant increase in wireless handheld devices and the prospect of billions of connected machines has compelled the research community to investigate different technologies which are able to deliver high data rates, lower latency and better reliability and quality of experience to mobile users. One of the problems, usually overlooked by the research community, is that more connected devices require proportionally more signalling overhead. Particularly, acquiring users’ channel state information is necessary in order for the base station to assign frequency resources. Estimating this channel information with full resolution in frequency and in time is generally impossible, and thus, methods have to be implemented in order to reduce the overhead. In this paper, we propose a channel quality estimation method based on the concept of Gaussian process regression to predict users’ channel states for varying user mobility profiles. Furthermore, we present a dual-control technique to determine which is the most appropriate prediction time for each user in order to keep the packet loss rate below a pre-defined threshold. The proposed method makes use of active learning and the exploration-exploitation paradigm, which allow the controller to choose autonomously the next sampling point in time so that the exploration of the control space is limited while still reaching an optimal performance. Extensive simulation results, carried out in an LTE-A simulator, show that the proposed channel prediction method is able to provide consistent gain, in terms of packet loss rate, for users with low and average mobility, while its efficacy is reduced for high-velocity users. The proposed dual-control technique is then applied, and its impact on the users’ packet loss is analysed in a multicell network with proportional fair and maximum throughput scheduling mechanisms. Remarkably, it is shown that the presented approach allows for a reduction of the overall channel quality signalling by over 90 % while keeping the packet loss below 5 % with maximum throughput schedulers, as well as signalling reduction of 60 % with proportional fair scheduling.

/pdf/adaptive-csi-and-feedback-estimation-in-lte-and-beyond-a-4g2y1od1ni.pdf

Adaptive CSI and feedback estimation in LTE and beyond: a Gaussian process regression approach

In recent years, unmanned aircraft systems (UASs) have garnered significant attention, and the demand for communication utilizing unmanned aircrafts (UAs) has increased. However, the limited available frequency for UA communication, despite the increasing utilization, poses a problem. Moreover, for the practical application of UA communication, the changes and differences in the propagation environment and communication demand of each UA communication must be considered because of the variety of supposed services, and the considerable interference caused by obstacles and reflected waves, due to high UA mobility. In this research, we construct a UA communication management system at the UAS operating frequency, with reference to the physical channel configuration of the LTE and LTE Sidelink, utilizing Q-Learning. In addition, we determine the effectiveness of the proposed method by evaluating various propagation-environment scenarios and communication demands.

Efficient Resource Allocation Utilizing Q-Learning in Multiple UA Communications

In the downlink of Long Term Evolution (LTE) systems, feedback and processing delays cause a mismatch between the current channel state and the Channel Quality Information (CQI) at the base station. This CQI aging leads to inaccurate channel adaptation and can, thus, highly degrade the cell capacity. To compensate for this performance loss, we study several CQI predictors under realistic delay and channel assumptions. Our results on cell throughput show that linear prediction with Stochastic Approximation provides at least the performance gains of the computationally more complex covariance-based linear predictors and Kalman filters. This surprising result points to Stochastic Approximation as a powerful and practical technique to increase downlink performance with limited channel knowledge.

Compensating for CQI aging by channel prediction: The LTE downlink

Adaptive Modulation and Coding (AMC) in LTE networks is commonly employed to improve system throughput by ensuring more reliable transmissions. Most of existing AMC methods select the modulation and coding scheme (MCS) using pre-computed mappings between MCS indexes and channel quality indicator (CQI) feedbacks that are periodically sent by the receivers. However, the effectiveness of this approach heavily depends on the assumed channel model. In addition CQI feedback delays may cause throughput losses. In this paper we design a new AMC scheme that exploits a reinforcement learning algorithm to adjust at run-time the MCS selection rules based on the knowledge of the effect of previous AMC decisions. The salient features of our proposed solution are: i) the lowdimensional space that the learner has to explore, and ii) the use of direct link throughput measurements to guide the decision process. Simulation results obtained using ns3 demonstrate the robustness of our AMC scheme that is capable of discovering the best MCS even if the CQI feedback provides a poor prediction of the channel performance.

/pdf/robust-adaptive-modulation-and-coding-amc-selection-in-lte-2o3giplwyw.pdf

Robust Adaptive Modulation and Coding (AMC) Selection in LTE Systems Using Reinforcement Learning

Allocating wireless resources often relies on the accurate feedback of channel state information (CSI). In this paper, we study how strongly a multiuser OFDM downlink with single-cell scheduling and channel estimation suffers from feedback delay. Unlike previous work, we study this degradation for optimal joint power and rate allocation under fairness constraints. Comparing the performance of the ideal case to delayed CSI shows that (i) adjusting the schedulers fairness cannot mitigate the strong performance loss due to feedback delay but that (ii) simple linear channel prediction is a powerful tool to do so.

/pdf/on-the-effect-of-feedback-delay-in-the-downlink-of-multiuser-4u55qr2089.pdf

Rudi Abi Akl

Papers

Compensating for CQI aging by channel prediction: The LTE downlink

On the effect of feedback delay in the downlink of multiuser OFDM systems