Massive-MIMO (M-MIMO) allows to schedule users with high gain narrow beams and to reduce intra-cell inter-beam interference. Close to cell edge, users may experience high interference from beams of neighboring cells which degrades their performance. This paper introduces low complexity inter-cell interference coordination between neighboring cells for M-MIMO systems implementing Grid of Beams (GoB) using Radio Environment Maps (REMs). The REMs are designed using the Kriging with the Covariance Tapering spatial interpolation technique and are created for the serving and the interfering beams. The interference coordination is introduced as constraints to the schedulers that avoid simultaneous scheduling of users served by highly interfering beams from the neighboring cells. The coordination decision is based on information retrieved from REMs and its quality depends on the REMs’ precision. Around 72 percent performance gain in terms of mean user throughput is achieved by the REM based coordination with respect to a baseline solution without coordination, and around 40 percent gain with respect to a state of the art solution that implements coordination using information at beam level resolution.

Quality of Service in mobile networks is a vigorous necessity that depends on the traffic demand growth and the complex emergence of several new services and technologies. It can be improved by reducing the network failures and avoiding the congestion. As a result, a hybrid model can be used for proactive traffic congestion avoidance to alert the operator thus enhancing the end user perceived QoS. This model consists of a co-clustering algorithm to group cells that have similar behaviour based on key performance indicators and a logistic regression model to predict congestion. The hybrid model is compared to most known deep learning models presented in the literature. We consider a Long Short-Term Memory based on recurrent neural network approach and a Temporal Convolutional Network approach for comparison. The different models are compared using real field data from operational Long Term Evolution networks.

Self-Organizing Network (SON) is a crucial technology characterizing the behaviour of future mobile networks. Complex cellular networks’ deployment, operation and maintenance are managed autonomously by multiple SON functions (SFs) with dedicated objectives. Designing appropriate configurations of SON is challenging as it requires comprehensive modelling of all their interactions. Previous works addressed this point and proposed cognitive management solutions to learn the optimal SON configurations through direct communication with the network. Some of these works have overcome SON’s shortcomings, such as limited flexibility and adaptability to changing environments. This paper discusses a context-specific policy that uses machine learning (ML) to learn the network environment. This way, the cognitive management solution finds the convenient SON configuration for each network context that is identified with a higher value of automation.

This paper studies the distribution of Signal-to-Interference-plus-Noise Ratio (SINR) for the downlink of cellular networks taking into account the joint impact of path-loss, shadowing and fast-fading. We show that the calculation of the probability density function (PDF) of the SINR is not tractable. We propose two approaches for deriving the analytical expressions of the statistics of the SINR using two different approximations and we compare the results to Monte Carlo simulations. In the first approach, the interfering signals are approximated to their means, the PDF of SINR approximates well the one obtained by simulation only when the variances of the shadowing signals are small. In the second approach, only the fast-fading is averaged for the interfering signals. We give a closed form expression of the distribution of the SINR. We show, by comparing to simulations, that this approximation is more accurate than the first one for any value of the shadowing variance. We also deduce a closed form for the average data rate expression.

The management of future cellular networks can be automatically adapted by machine learning models. In this article, we present the portability of a proposed hybrid machine learning model to forecast the network congestion allowing, operators to pro-actively monitor and control future congestion before occurring. To overcome the quality of service degradation of the radio access network, the spatio-temporal hybrid model must involve automatic quick response to forecast future congestion and alert network operators to correct future congested cells to provide a continuous quality of service for the subscribers. Based on real field data of key performance indicators from operational long term evolution networks of Orange, we apply the hybrid machine learning model to ensure spatio-temporal congestion forecasting. In this paper, we present the portability of a proposed hybrid model in space and time. The idea of portability is that knowledge extracted from a specific city environment can be applied directly to another different cities environment. The hybrid model is a combination of a Latent Block Model (LBM) co-clustering technique that succeeds in grouping cells located in distant cities that have a similarity in their behaviors and a logistic regression technique to forecast future congestion in radio access networks. The experiments presented in this article confirm that the LBM co-clustering technique allows the model’s portability from a city to another one. In addition, it is applied for the same cells for different months to validate its stability over time.

Congestion avoidance in radio access networks enhances considerably the end-user Quality of Service (QoS). Congestion should be predicted in advance to allow Self Organizing Networks (SON) algorithms to perform appropriate parameter adjustments (such as handover parameters for mobility load balancing). For this purpose, a novel hybrid model efficient congestion prediction mechanism is proposed in this paper. This hybrid learning model combines unsupervised and supervised learning algorithms. The unsupervised learning consists of a co-clustering algorithm based on Latent Block Model (LBM) that groups similar cells according to their KPIs behaviour over time. Following the co-clustering model, a logistic regression approach is applied on each cluster to predict congestion and alert operators to avoid congestion occurrence in mobile networks. The applicability of the hybrid model is validated for a real data represented by Key Performance Indicators (KPIs) collected periodically for 12 days in a live Long-Term Evolution (LTE) network. The hybrid proposed model has proven its efficiency in congestion prediction in terms of accuracy, precision, recall and F-measure.

In this paper, we consider resource allocation for vehicular safety traffic. In 5G, this traffic is carried by using the Ultra-Reliable and Low-Latency Communications (URLLC) service as it needs stringent Quality of Service (QoS) requirements in terms of latency and reliability. Since URLLC services may require specific numerology and/or channel access and retransmission strategies, network slicing has been proposed as a solution for QoS requirements and its coexistence with other services such as enhanced Mobile Broad-Band (eMBB). In order to accommodate URLLC traffic, one can opt for static resource reservation, however this is not optimal as it does not follow the real URLLC traffic present in the cell and can impact negatively eMBB traffic. Reactive, on-demand resource reservation is not feasible either as it requires reconfiguration which introduces extra delay that makes it prohibitive to meet URLLC delay requirements. This paper proposes proactive resource reservation schemes that anticipate slice demand. Resource reservation is computed per gNodeB based on the expected traffic and radio conditions. We show how field measurements and trajectory predictions can be used to achieve URLLC objectives with low impact on eMBB performance.

Signal-to-interference-plus-noise ratio (SINR) evaluates the quality of the link between the base station (BS) and the user equipment (UE), taking into account the interference coming from neighboring cells and the thermal noise. Predicting the SINR with good accuracy ensures proper planning and optimization of radio coverage and good estimation of end-user throughput. In this paper, SINR is predicted at any user location based on UE geo-located measurements, using the Kriging technique and assuming that the received signals from serving and interfering cells undergo shadowing fading. Unlike previous works where only the prediction of the received powers is studied, the current functionality also considers the prediction error on the received signals. We consider both spatial correlation of shadowing signals for each cell and also inter-cell correlation and we propose a new model to generate multiple links shadowing signals for the downlink of cellular networks. For validation, we perform Monte-Carlo network simulations in which the shadowing samples are generated based on the proposed model. Results show a good prediction accuracy of the SINR at a new location resulting in a good estimation of the end-user perceived average data rate.

Ultra-Reliable Low Latency Communications (URLLC) is a key service in fifth generation (5G) networks, that requires stringent Quality of Service (QoS) in terms of latency and reliability. As URLLC services may require specific numerology and/or specific channel access and re-transmission strategies, network slicing has been proposed as a solution for multiplexing them with other services such as enhanced Mobile Broadband (eMBB). Once the URLLC slice is configured and resources are dimensioned and allocated to it, URLLC performance targets should be attained thanks to the 5G New Radio (NR) low latency and high reliability features. However, in vehicular services such as safety message exchange, URLLC slice resource dimensioning cannot be static due to the varying number of vehicles in the cell. We show in this paper how the delay for slice reconfiguration alters the URLLC performance and propose a proactive resource reservation scheme that anticipates slice needs and allows ensuring URLLC targets. In order to reduce the impact of this proactive reservation on eMBB performance, we make use of vehicle trajectory prediction and show that limiting anticipated reservation to fewer cells allows reaching the target URLLC QoS with a limited degradation of the network capacity.

In this paper, we propose a model for proactive anomaly detection in mobile networks. We show that when Key Performance Indicators (KPIs) are highly correlated, the linear regression gives a good accuracy in anomaly detection for a short prediction horizon. When the prediction horizon is far, KPIs become weakly correlated. We propose to transform discrete measurements into functional data and apply a functional data regression method to perform the prediction. We compare pro-posed models using KPI measurements obtained from a real Long Term Evolution (LTE) network. We show that an improvement in prediction performance can be obtained by using functional data analysis.

Signal-to-Interference-plus-Noise Ratio (SINR) measures the quality of the radio link between a base station and a User Equipment (UE). This quality indicator is important for the operator as it is used to measure and to optimize network coverage and the maximum user bitrate (which can be deduced from the SINR using the Shannon formula). In this paper, we study the SINR distribution for the downlink assuming that the received signals from serving and interfering cells undergo a path-loss and log-normal shadowing fading. We show that SINR can be approximated by a log-normal random variable and we give analytical expressions of characteristics parameters of its distribution. We then derive the average SINR expression. Also, we give a closed-expression of the rate probability density function and the outage probability and we obtain closed expressions of the average rate and lower and upper bounds of the average rate. For validation, we first compare theoretical results to Monte-Carlo network simulations and show that all the theoretical curves approximate well those obtained by simulations. Then using real measurements obtained by a measurement campaign, we validate the SINR distribution assumption. Based on a test of normality, we find that the recorded SINR samples are normally distributed in the logarithmic domain. We also show that the theoretical probability density function curves and the histogram of the measured samples of SINR are close.

Self-Organizing Networks (SON) functions have already proven to be useful for network operations. However, a higher automation level is required to make a network enabled with SON capabilities respond as a whole to the operator's objectives. For this purpose, a Policy Based SON Management (PBSM) layer has been proposed to manage the deployed SON functions. In this paper, we propose to empower the PBSM with cognition capability in order to manage efficiently SON enabled networks. We focus particularly on the implementation of such a Cognitive PBSM (C-PBSM) on a large scale network and propose a scalable approach based on distributed Reinforcement Learning (RL): RL agents are deployed on different clusters of the network. These clusters should be defined in such a way that the RL agents can learn independently. As the interaction between these clusters may evolve in time due for instance to traffic dynamics, we propose a flexible implementation of this C-PBSM framework with dynamic clustering to adapt to network's evolutions. We show how this flexible implementation is rendered possible under Software Defined Networks (SDN) framework. We also assess the performance of the proposed distributed learning approach on an LTE-A simulator.

Policy Based SON Management (PBSM) is the process of orchestrating the deployed Self-Organized Network (SON) functions, so that the network responds as a whole to the operator objectives. This process is based on the configuration of the SON functions in order to steer their actions in the network towards certain operator objectives. The PBSM ensures an automated translation of these objectives into configurations of the SON functions. An approach has been recently proposed to empower the PBSM with cognitive capabilities (C-PBSM), using a Multi-Armed Bandit algorithm, namely the UCB1. The C-PBSM learns the optimal SON configurations based on network feedback. In this paper we propose an alternative approach, based on the LinUCB algorithm, that is able to learn the optimal SON configuration much faster than the previous approach. The speed of convergence is a critical factor that has to be thoroughly considered in the deployment of online learning processes. Results are shown using an LTE-A simulator that considers real-like network topology and parameters, and accurate ray tracing based propagation.

As video accounts for larger wireless traffic, improving users' quality of experience becomes important for network service providers. In this paper, we apply supervised machine learning technique to predict one objective QoE metric, video starvation, with the users' features, recorded at the beginning of each video session. We show that static users and adaptive streaming users have less starvation events and that mobile users are more difficult to predict their video starvation. In terms of users' features, we show that system parameters such as channel conditions and number of active users are two important features which contribute to better prediction performance. Prediction with the two features can provide sufficient accuracy for static users but not sufficient for mobile users. We also demonstrate that the two information, number of users served in a cell and the number of users experiencing video starvation, provide similar prediction accuracy.

Self organized networks has been one of the first concrete implementations of autonomic network management concept. Currently, several Self-Organizing-Network (SON) functions are developed by Radio Access Network (RAN) vendors and already deployed in many networks all around the world. These functions have been designed independently to replace different operational tasks. The concern of making these functions work together in a coherent manner has been studied later in particular in SEMAFOUR project where a Policy Based SON Management (PBSM) framework has been proposed to holistically manage a SON enabled network, namely a network with several individual SON functions. Enriching this PBSM framework with cognition capability is the next step towards the realization of the initial promise of SON concept: a unique self-managed network that responds autonomously and efficiently to the operator high level requirements and objectives. This paper proposes a cognitive PBSM system that enhances the SON management decisions by learning from past experience using Q-learning approach. Our approach is evaluated by simulation on a SON enabled Long-Term Evolution Advanced (LTE-A) network with several SON functions. The paper shows that the decisions are enhanced during the learning process and discusses the implementation options of this solution.

Many tasks in current mobile networks are automated through Self-Organizing Networks (SON) functions. The actual implementation consists in a network with several SON functions deployed and operating independently. A Policy Based SON Manager (PBSM) has been introduced to configure these functions in a manner that makes the overall network fulfill the operator objectives. Given the large number of possible configurations (for each SON function instance in the network), we propose to empower the PBSM with learning capability. This Cognitive PBSM (C-PBSM) learns the most appropriate mapping between SON configurations and operator objectives based on past experience and network feedback. The proposed learning algorithm is a stochastic multi-armed bandit, namely the UCB1. We evaluate the performances of the proposed C-PBSM on an LTE-A simulator. We show that it is able to learn the optimal SON configuration and quickly adapts to objective changes.

Optimization of mobile networks is a crucial task for operators aiming to offer a high quality of service. In this paper, we focus on the handover (HO) problem. An HO is the procedure through which a user changes its serving cell, ensuring service continuity for mobile users. The HO procedure is typically governed by several variables: the received signals powers and the HO parameters (offsets and hysteresis) corresponding to the different cells. HO parameters allow to modify the cell borders in order to balance loads, and prevent radio link failures and HO ping-pongs. There are two approaches for defining HO parameters: one is to use per-cell parameters and the other is to use per-cell-pair parameters. Per-cell-pair parameters provide more degrees of freedom in HO parameter optimization as we benefit from independent parameters for each neighbour. However, we show that it can lead to instabilities in the form of continuous chained HO, i.e., several HOs that follow one shortly after another, among three or more cells. Then, we show that using the same HO parameters toward all neighbouring cells, i.e., per-cell parameters, is risk free of such problems. We provide numerical results outlining the tendencies of this risk.

Coverage optimization is an important process for the operator, as it is a crucial prerequisite toward offering a satisfactory quality of service to the end users. The first step of this process is coverage prediction, which can be performed by interpolating geo-located measurements reported to the network by mobile user's equipments. In the previous works, we proposed a low complexity coverage prediction algorithm based on the adaptation of the geo-statistics fixed rank kriging (FRK) algorithm. We supposed that the geo-location information reported with the radio measurements was perfect, which is not the case in reality. In this paper, we study the impact of location uncertainty on the coverage prediction accuracy and we extend the previously proposed algorithm to include geo-location error in the prediction model. We validate the proposed algorithm using both simulated and real-field measurements. The FRK is extended to take into account that the location uncertainty proves to enhance the prediction accuracy while keeping a reasonable computational complexity.

An important problem of today's 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.

In trying to meet the demands of traffic hungry users, mobile network operators are faced with increased CAPital EXpenditures (CAPEX) and OPerational EXpenditures (OPEX). The Self Organizing Network (SON) functions have been introduced by 3GPP as a means to cut down these costs. There are mainly 3 categories of such functions: self-configuration, self-optimization and self-healing. In this paper we focus on the second which represents the SON functions performing a runtime optimization of the network. We center our attention on LTE heterogeneous networks. Having several SON functions in a network may lead to conflicts and potentially to bad network Key Performance Indicators (KPIs). Thus a troubleshooting mechanism has to be envisaged. Such a mechanism typically contains 3 steps: fault detection, cause diagnosis and solution deployment. In this paper we tackle the first two and we study the feasibility of using the Naive Bayes Classifier (NBC) in order to build a framework for SON Conflict Diagnosis (SONCD). We provide numeric results proving the feasibility of the framework.

The continuously increasing traffic demand faces us with increased CAPital Expenditures (CAPEX) and Operational Expenditure (OPEX). Self Organizing Network (SON) functions aim to lower these costs by automating the network tuning. A SON instance is a realization of a SON function which can tune one or a set of cells. Having several uncoordinated SON functions in the network creates a risk for conflicts and instability. This raises the need for a SON Coordinator (SONCO) meant to deal with these issues. In this work we consider that on each cell we have one SON instance of each SON function. We present the design of a SONCO which arbitrates conflicts based on weights attributed to the SON functions. The design makes use of Reinforcement Learning (RL) with function approximation. We provide a low complexity approximation of the action-value function based on a number of parameters that scales linearly with the number of cells. We present a study case with the Mobility Load Balancing (tuning the Cell Individual Offset (CIO)) and Mobility Robustness Optimization (tuning the CIO and the handover hysteresis) functions, where the SONCO deals with the conflicts on the CIOs. Numerical results prove that we can orchestrate the SON functions through SONCO configurations that reflect different operator policies.

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.

With the emergence of Self-Organizing Network (SON) functions network operators are faced with a practical problem: coordination of SON instances. The SON functions are usually designed in a standalone manner, i.e. they do not take into account the possibility that other instances of the same or different SON functions may be running in the network. This creates the risk of conflicts and network instability. Therefore a SON COordinator (SONCO) is needed. In this paper we design an operator centric SONCO that sees the SON instances as black-boxes, i.e. it does not know the algorithm inside the SON functions. Our aim is to improve the network stability (i.e. number of parameter changes) for SON instances of the same SON function. We employ Reinforcement Learning (RL) in order to profit from the information on the past SONCO decisions. We simplify the expression of the action-value function and we use state aggregation to further reduce the required state space, making it scale linearly with the number of coordinated cells. We provide a study case with the Mobility Load Balancing (MLB) function independently instantiated on every cell. The results show that the proposed SONCO improves the network stability.

In Long Term Evolution(LTE) networks one of the main focuses is on automating the network optimization. This is done through so called Self Organizing Network (SON) functions like Mobility Load Balancing (MLB), Mobility Robustness Optimization(MRO) and others. A SON instance is a realization of a SON function that governs (optimizes) one or a cluster of eNBs. The SON functions are built in a standalone manner, i.e. without considering the existence of other SON instances. So they do not necessarily operate in a coordinated fashion, especially in a network where different SON instances may come from different vendors. Thus we face a risk of generating conflicts and instabilities in the network and so this raises the need for a SON COordinator (SONCO) . The SONCO, built from an operator point of view, sees the SON instances as black boxes and has a very limited amount of information on them. In these conditions the SONCO has to solve conflicts and improve the network stability. In this paper we propose a Reinforcement Learning (RL) based solution for coordinating SON instances that run independently on neighboring eNBs and we provide results for a case study with MLB instances. We analyze the scalability of our solution and we provide numerical results showing how improvements in network stability can be obtained.

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.

The Self Organizing Network (SON) concept is a promising paradigm aiming to reduce both the capital and operational expenditures (CAPEX/OPEX) in future wireless networks through efficient self-configuration, self-optimization and self-healing functionalities. As for self-optimization, much effort has been devoted to standalone SON control loops, which, upon network metric observation, dynamically adjust the appropriate network parameters to enhance the overall network performance. However, less explored simultaneous operation of several SON control loops may cause conflicts between parameter adjustment and/or metric performance outcome. This paper focuses on SON mechanism interactions, providing a framework for further understanding and analysis. A basic two-SON mechanism interaction scenario will be provided along with an initial assessment of a simple SON interaction control.

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.

This paper deals with mobile-centered decision making in heterogeneous networks, where intelligent mobile terminals take autonomous decisions about the JRRM actions, consisting to connect to one of the available systems. This distributed decision-making is possible due to Q-learning algorithms implemented within the mobile terminals that enable them to profit from their past experience in order to enhance their subsequent decisions. We develop an original Markovian model that allows analyzing analytically the evolution of the Q-learning process and show how the the performance is enhanced until convergence.

In this paper, we focus on mobile-centered decision making in heterogeneous networks. We study the case where the JRRM decision is completely distributed so that mobile users have to decide to which of the available systems it is best to connect. To optimize their decision over the time, the mobiles implement a Q-learning algorithm that enables them to profit from their past experience. We study the performance of this decision-making framework in the case of a WiMAX/HSDPA heterogeneous network and show that the mobile decision is progressively enhanced until convergence.

In this paper, we focus on the joint radio resource management (RRM) for HSDPA/WiMAX cooperation. We use a generic Markovian model to model the evolution of data calls in a multi-cell and multi-access context. We apply this model to a cooperative WiMAX/HSDPA network considering both inter-cell and vertical handovers and calculate the steady-state probabilities and the performance indicators taking into account intra-cell and inter-cell mobility. Then, we detail two RRM approaches: a game-theory based RRM approach and a RRM approach based on the maximization of the global network utility. In the first approach, the RRM decision is taken by the mobile terminals (MTs) but the network takes over if necessary, while the second approach is completely network-centric. Finally, we compare both RRM approaches on the basis of the defined performance indicators. The main results show that the game-theoretic method achieves better performances than the global-utility maximization method.

This paper introduces interference cartography, a simple and effective concept that helps detect, identify and use spectrum opportunities in a secondary spectrum usage context. Interference cartography combines measurements performed by different network entities (mobile terminals, base stations, access points) with the geo-location information and applies simple and effective spatial interpolation techniques to achieve a map which indicates the level of interference experienced at each mesh over the area of interest. Using this information, a secondary network can detect the presence of a primary network (or of other secondary networks) and can use spectrum opportunities without causing harmful interference to them. As an example, a reliable spatial interpolation technique, kriging, is applied to interference data obtained from a radio network simulator. Obtained results demonstrate that interference cartography is a promising concept that can enhance the performance of secondary spectrum usage.

In this paper, we develop a generic Markovian model to study the dynamics of data service in a multi-cell context, and apply it to cells served by HSDPA and WiMAX access networks. We first present analytical models for interference and throughputs in WiMAX and HSDPA. Then, we consider a cooperative WiMAX/HSDPA network considering both inter-cell and vertical handovers and show how to calculate the steady-state probabilities and the performance indicators. Our calculations take into account mobility within the cell and between different cells and make us able to calculate intra and inter-cell dropping. We compare the performances of three joint Radio Resource Management (RRM) policies: HSDPA/WiMAX traffic distribution policy, HSDPA filling policy and WiMAX filling policy. Our numerical results show that the policy favoring WiMAX access and considering HSDPA as a rescue system (WiMAX filling policy) offers the best performances.

This paper presents a policy-based radio resource management model for decision sharing between the network and the mobile terminals in composite networks. One of the key issues discussed in this paper is the distribution of RRM decision making between the network and the mobile terminal. IEEE P1900.4 standard project proposes a policy-based radio resource usage scheme where the network orients the mobile terminals decisions by sending rules and constraints. Mobile terminals choose the most appropriate action to their own QoS objectives, within the action set allowed by the network. We propose a joint RRM framework based on this standard project and we discuss both policies definition on the network side and decision making on the mobile terminal side. We propose a model where policies are chosen automatically according to the variations of the environment conditions and the operator strategy. On the mobile terminal side, we propose to introduce a time constraint imposed by the network for policy execution in order to coordinate the resulting terminals' actions execution over time. Moreover, as final decisions are made by MTs in a distributed way, we discuss the modeling of this problem using cooperative and non cooperative games approaches.

Mobile terminals measurements capabilities are strong limitations that may impede seamless handover from UMTS to WLAN. In this article, we propose an inter-system handover method that does not require in-line measurements on WLAN. It identifies the most probable WLAN Access Point for handover, from coverage point of view, depending on UMTS signal measurements and on a handover probability database. This method can be implemented in the RNC, or in a B3G common controller. Simulation results in a WLAN hot spot deployment show that handover based on a priori knowledge of the networks increases handover success rate, compared to classical blind handover. Its performance can be improved by optimizing the handover probability database. Besides, the proposed method enables to forecast handover's success, and adapt accordingly. Handover based on a priori knowledge of the networks can consequently be integrated into several B3G handover algorithms, to decrease handover delay and complexity.

The global beyond 3G system consists of several coexisting and cooperating access technologies. One of the key concepts of this global technology is the reconfigurability, that allows different network elements to dynamically adapt their configuration to the new conditions encountered in specific service areas and time. Reconfigurability may comprise dynamic spectrum allocation: a technique that varies spectrum allocation of different systems in order to meet changing demands. In this context of multiple access techniques and changing spectrum allocation, when a mobile is switched on, it has no information about the available systems in its area nor on the current spectrum allocation to these systems. In order to avoid the scanning of all the spectrum range and to facilitate the initial connection to the network, this paper proposes that the mobile listens first to a broadcast radio channel containing the necessary information to initiate its connection. The paper defines the content of this broadcast channel, denoted common pilot channel, and proposes a technical implementation

This paper presents indicator matrices for assessing the quality of UMTS networks, the interference and the macrodiversity matrices. A design methodology based on these indicators is developed to guide the network designer in complex design tasks. The indicator matrices are of particular interest for both assessing of network performance and for design purposes since they allow one to identify the interaction between any couple of stations in the network. Hence, the identification of problematic sectors as well as possible curative solutions is made much simpler. An example of network design illustrates the effectiveness of the indicator matrix approach.