Sana Ben Jemaa

Improving coverage estimation for cellular networks with spatial bayesian prediction based on measurements

Cellular operators routinely use sophisticated planning tools to estimate the coverage of the network based on building and terrain data combined with detailed propagation modeling. Nevertheless, coverage holes still emerge due to equipment failures, or unforeseen changes in the propagation environment. For detecting these coverage holes, drive tests are typically used. Since carrying out drive tests is expensive and time consuming, there is significant interest in both improving the quality of the coverage estimates obtained from a limited number of drive test measurements, as well as enabling the incorporation of measurements from mobile terminals. In this paper we introduce a spatial Bayesian prediction framework that can be used for both of these purposes. We show that using techniques from modern spatial statistics we can significantly increase the accuracy of coverage predictions from drive test data. Further, we carry out a detailed evaluation of our framework in urban and rural environments, using realistic coverage data obtained from an operator planning tool for an operational cellular network. Our results indicate that using spatial prediction techniques can more than double the likelihood of detecting coverage holes, while retaining a highly acceptable false alarm probability.

A REM enabled soft frequency reuse scheme

Optimizing the network performance is a key objective for cellular operators as over-provisioning a network is definitely not a cost-effective solution. However the trend for the upcoming 4G networks is to simplify the planning procedure and to improve the spectral efficiency by using agressive frequency reuse schemes, e.g reuse-1, in the downlink. In this paper, we show that this is not a suitable configuration in terms of network performance and describe a simple Radio Environment Map (REM) enabled soft frequency reuse (SFR) scheme which improves the user throughput of about 14%.

Bayesian spatial interpolation as an emerging cognitive radio application for coverage analysis in cellular networks

Coverage prediction is one of the most important aspects of cellular network optimisation for mobile operators due to the highly competitive market conditions and the obligations towards the regulatory authorities. Although wireless communications research is considerably diversifying, including new areas, problems and concepts with a notable pace, cellular coverage still remains as a core subject of research for operators because of the obligation to adapt to the continuously evolving wireless ecosystem, with new radio access technologies/architectures, emerging applications and innovative cellular concepts. This paper presents the application of a powerful mathematical tool coming from spatial statistics, Bayesian kriging, to construct a radio environment map (REM) for the purpose of cellular coverage prediction as an emerging application of cognitive radio techniques in cellular networks. The proposed approach provides an efficient alternative to the conventional manual coverage prediction on the basis of drive tests, which are expensive, polluting and slow solutions for obtaining the ‘ground-truth’ information. Bayesian kriging-based REMs allow to estimate the coverage situation in those regions where the operator lacks direct information. Our approach can be directly used by operators for the cellular network coverage optimisation. We evaluate the accuracy of the proposed REM construction approach for a long term evolution network with two mesh sizes using highly realistic data sets. Results show that the Bayesian kriging interpolation technique has a good accuracy for cellular coverage prediction, and this accuracy is directly related with the mesh size. Copyright

SON Coordination for parameter conflict resolution: A reinforcement learning framework

Self Organizing Network (SON) functions are meant to automate the network tuning, providing responses to the network state evolution. An instance of a SON function can run on one cell (distributed architecture) or can be built to govern a cluster of cells (centralized/hybrid architecture). From the operator point of view, SON functions are seen as black boxes. Several independent instances of one or multiple SON functions running in parallel are likely to generate conflicts and unstable network behavior. At a higher level, the SON-COordinator (SONCO) seeks to solve these conflicts. This paper addresses the design of a SONCO. We focus on coordinating two distributed SON functions: Mobility Load Balancing (MLB) and Mobility Robustness Optimization (MRO). Thus on each cell we will have an MLB and an MRO instance. The MLB instances will tune the Cell Individual Offset (CIO) parameter and the MRO instances will tune the HandOver (HO) Hysteresis parameter together with the CIO parameter. The task of the SONCO is to solve the conflicts that will appear on the CIO parameter. We propose a Reinforcement Learning (RL) framework as it offers the possibility to improve the decisions based on past experiences. We outline the tradeoff between configurations through numeric results.

Low complexity spatial interpolation for cellular coverage analysis

During the last decade a lot of effort has been spent on cellular network optimization to improve network capacity and end-user Quality of Service (QoS). Coverage analysis remains as one of the essential topics on which mobile operators still need innovation in terms of performance and cost. Manual coverage analysis is an inefficient and costly task. Radio Environment Maps (REMs) is an efficient coverage analysis solution for present-day cellular networks. REM concept consists of spatially interpolating geo-located measurements to build the whole coverage map using a spatial interpolation technique originating from geo-statistics. Kriging is such a powerful technique which results in high performance in terms of prediction quality. However, this method is costly in terms of computational complexity especially for large datasets: computational complexity of Kriging is O(n3) where n is the number of measurements. This paper proposes the application of a variant of Kriging, Fixed Rank Kriging (FRK), to coverage analysis in order to reduce the computational complexity of the spatial interpolation while keeping an acceptable prediction error.

A Heuristic Coordination Framework for Self-Optimizing Mechanisms in LTE HetNets

A self-organizing network (SON) is effectively realized by means of specific SON mechanisms (SONm), which relate to particular SON use cases (SONuc), these being defined by the Third-Generation Partnership Project (3GPP) and Next Generation Mobile Networks (NGMN). Typically, SONuc are grouped into self-configuration, self-optimization, and self-healing functions. Focusing on self-optimization, SONm therein aim at maintaining relevant key performance indicators (KPIs) above/below a specific target by actuating over an appropriate set of input parameters to achieve predefined network objectives. Typically, this problem is represented by several control loops where controllable input parameters are dynamically adjusted according to output measurable metrics and their corresponding target requirements. From an implementation viewpoint, the concept of a single SON “black box,” integrating multiple SONm, is very appealing. However, this approach may threaten the control that network operators (NOPs) have over their own network. As a result, a coordinated framework involving stand-alone SONm is proposed. Here, a SON controller (SONc) may be implemented responding to strategies, deciding at a given time which SONm actions have higher priority with respect to others. In the context of a heterogeneous network (HetNet) scenario, we propose and develop a simplified framework where stand-alone SONm react to either overshot or undershot KPI events by deciding to either increase or decrease corresponding influential parameter(s). By inspecting the arising interactions and possible conflicts between several SONm, we provide, for a specific evaluation scenario, a heuristic strategy-based solution for SON coordination, which may eventually trade off high-level NOP goals.

SON Coordination in Heterogeneous Networks: A Reinforcement Learning Framework

An important problem of todays mobile network operators is to bring down the capital expenditures and operational expenditures. One strategy is to automate the parameter tuning on the small cells through the so-called self-organizing network (SON) functionalities, such as cell range expansion, mobility robustness optimization, or enhanced Inter-Cell Interference Coordination. Having several of these functionalities in the network will surely create conflicts, as, for example, they may try to change the same parameter in the opposite directions. This raises that the need for an SON COordinator (SONCO) meant to arbitrate the parameter change requests of the SON functions, ensuring some degree of fairness. It is difficult to anticipate the impact of accepting several simultaneous requests. In this paper, we provide a SONCO design based on reinforcement learning (RL) as it allows us to learn from previous experiences and improve our future decisions. Typically, RL algorithms are complex. To reduce this complexity, we employ two flavors of function approximation and provide a study-case. Results show that the proposed SONCO design is capable of biasing this fairness among the SON functions by means of weights attributed to the SON functions. Also, we evaluate the tracking capability of the algorithms.

SON conflict resolution using reinforcement learning with state aggregation

In future generation networks one of the main focuses is on automating the network optimization. This is done through so called Self Organizing Network (SON) functions. A SON instance is a realization of a SON function that governs one or several cells. Several independent SON instances of one or multiple SON functions are likely to generate conflicts. This raises the need for a SON COordinator (SONCO) meant to solve these conflicts. In this paper we consider that each SON function has one SON instance on every cell and we present the design of a SONCO function for coordinating all these instances. The SONCO solves the conflicts that appear on the update requests arbitrating (i.e. accepting/denying the requests) so that it minimizes a predefined regret. This regret takes into account the weights associated to the SON functions that rank their importance according to the operator policies. We solve the problem in a Reinforcement Learning (RL) framework as it offers the possibility to improve the decisions based on past experiences. We employ a state-aggregation technique to make the state space of our solution scale linearly with the number of cells. We provide a study case for two SON functions: Mobility Load Balancing (MLB) tuning the Cell Individual Offset(CIO) and Mobility Robustness Optimization (MRO) tuning the CIO together with the handover hysteresis. The proposed SONCO function solves the conflicts on the CIO update requests. Numerical results show how the proposed SONCO is able to favor either MLB or MRO requests according to their associated weights.

Best Sensor Selection for an Iterative REM Construction

Abstract-In this paper we propose a Bayesian approach for estimating parameters of the radio propagation model, and an iterative Kriging interpolation algorithm for choosing the best candidate measurement to be retrieved into the Radio Environment Map (REM). We compare the performance with a random choice of the candidate measurement and show that our algorithm reduces the amount of measurement needed by 33%. The proposed algorithm has also the merit of being fast enough to be implemented in an online fashion for REMs with a grid size of 25m and for pedestrian mobile speeds.

Coverage mapping using spatial interpolation with field measurements

Coverage optimization is a crucial task for a radio network operator. An accurate coverage estimation is a key prerequisite for efficient coverage analysis and optimization. In this paper, we propose a coverage prediction method based on statistical modeling of the wireless environment. We build a Radio Environment Map by interpolating geo-located measurements using the Kriging spatial prediction technique. Moreover, as we perform Kriging on massive observation datasets obtained through field measurement campaigns, we use Fixed Rank Kriging, to reduce the complexity of the Kriging algorithm. We apply the FRK algorithm for Long Term Evolution (LTE) network coverage prediction. We consider as observation data, the coverage measurements obtained by operational drive tests in a rural area. Numerical results show that by using the FRK algorithm, we fulfill a good trade-off between computational complexity and prediction accuracy.