Apollinaire Nadembega

A Destination and Mobility Path Prediction Scheme for Mobile Networks

Mobile multimedia services are gaining great momentum among subscribers of mobile networks (MNs). An understanding of the network traffic behavior is essential in the evolution of todays MNs and, thus, leads to more efficient planning and management of the networks scarce bandwidth resources. The communication efficiency can be largely improved (i.e, optimizing the allocation of the networks limited resources and sustaining a desirable quality of service) if the network anticipates the needs of its users on the move and, thus, performs reservation of radio resources at cells along the path to the destination. In this vein, we propose a mobility prediction scheme for MNs; more specifically, we first apply probability and Dempster-Shafer processes for predicting the likelihood of the next destination, for an arbitrary user in an MN, based on the users habits (e.g, frequently visited locations). Then, at each road junction, we apply the second-order Markov chain process for predicting the likelihood of the next road segment transition, given the path from the trip origin to that specific road junction and the direction to the destination. We evaluate our proposed scheme using real-life mobility traces; the simulation results show that the proposed scheme outperforms traditional schemes.

Mobility-Prediction-Aware Bandwidth Reservation Scheme for Mobile Networks

Bandwidth is an extremely valuable and scarce resource in mobile networks; therefore, efficient mobility-aware bandwidth reservation is necessary to support multimedia applications (e.g., video streaming) that require quality of service (QoS). In this paper, we propose a distributed bandwidth reservation scheme called the mobility-prediction-aware bandwidth reservation (MPBR) scheme. The objective of MPBR is to reduce handoff call dropping rate and maintain acceptable new call blocking rate while providing efficient bandwidth utilization. MPBR consists of a handoff time estimation (HTE) scheme that aims to estimate the time windows when a user will perform handoffs along the path to his destination, an available bandwidth estimation (ABE) scheme that aims to estimate in advance available bandwidth, during the computed time windows in the cells to be traversed by the user to his destination, and an efficient call admission control (ECaC) scheme that aims to control bandwidth allocation in the network cells. The simulation results show that MPBR outperforms existing schemes in terms of reducing handoff call dropping rate.

A Destination Prediction Model based on historical data, contextual knowledge and spatial conceptual maps

Mobile Wireless Network technology has enabled the development of increasingly diverse applications and devices resulting in an exponential growth in usage and services. One challenge in mobility management is the movement prediction. Prediction of the users longer-term movement (e.g., 10 min in advance) with reasonable accuracy is very important to a broad range of services. To cope with this challenge, this paper proposes a new method to estimate a users future destination, called Destination Prediction Model (DPM). This method combines two types of approaches: one based on the use of filtered historical movement pattern and another based on contextual knowledge; both approaches use spatial conceptual maps. The filter is based on the day and the time of the day to increase accuracy. The current movement direction, that takes into account the recent data, is used by the proposed method to reduce historical and contextual knowledge mistakes. Simulations are conducted using real-life data to evaluate the performance of the proposed model. For subjects with low predictability degree, DPM reaches an average prediction accuracy of 79%; it reaches 91% for subjects with high predictability and 86% for other subjects. Simulation results also indicate that DPM significantly reduces the impact of learning period and the remaining distance to reach the destination on prediction performance. In the future, we plan to extend our research work by proposing a full Path Prediction Model (PPM) based on the Destination Prediction Model (DPM).

An Integrated Predictive Mobile-Oriented Bandwidth-Reservation Framework to Support Mobile Multimedia Streaming

Bandwidth is an extremely valuable and scarce resource in wireless networks. Therefore, efficient bandwidth management is necessary to support service continuity, guarantee acceptable quality of service and ensure steady quality of experience for users of mobile multimedia streaming services. Indeed, the support of uniform streaming rate during the entire course of a streaming service, whereas the user is on the move is a challenging issue. In this paper, we propose a framework, together with schemes, which integrates user mobility prediction models with bandwidth availability prediction models to support the requirements of mobile multimedia services. More specifically, we propose schemes that predict paths to destinations, times when users will enter/exit cells along predicted paths, and available bandwidth in cells along predicted paths. With these predictions, a request for a mobile streaming service is accepted only when there is enough (predicted) available bandwidth, which is along the path to destination, to support the service. Simulation results show that the proposed approach outperforms existing bandwidth management schemes in better supporting mobile multimedia services.

A path prediction model to support mobile multimedia streaming

Along with the recent and ongoing advances in the wireless and mobile access technologies, a wide plethora of mobile multimedia services have emerged. Ensuring an acceptable level of Quality of Service (QoS) is a crucial requirement to allow users enjoy these mobile multimedia services. One means to ensure QoS is to minimize the frequency and magnitude of fluctuations in the mobile multimedia streaming rates during the multimedia service and while users are on the move. For this purpose, there is need for tools to predict a users long-term movement. In this vein, this paper proposes a Path Prediction Model (PPM) to predict a users movement path. PPM is based on historical movement trace, current movement data and spatial conceptual maps; it assumes a priori knowledge of the destination. At each road intersection, the probability of selecting the next road segment is evaluated, based on historical data, towards the destination. These probabilities are computed via (a) filtering historical data according to the day of the week (e.g., weekend, holiday) and the time of the day; and (b) applying conditional probability rules taking into account the path used between the origin of movement, current position, and the destination. Simulations are conducted using real-life data to evaluate the performance of the proposed model. Encouraging results are obtained in terms of average prediction accuracy and mitigation of the impact of learning period and the remaining distance to reach the destination on the path prediction performance.

Handoff time estimation model for vehicular communications

A good understanding of the behaviour of the traffic of a mobile network is essential for an efficient planning and management of the mobile networks scarce bandwidth resources. In this paper, we propose a probabilistic approach, called Handoff Time Estimation MODel (HTEMOD), to estimate the time window when a user will perform handoffs along his/her movement/path to a destination. We derive the probability distribution function of time taken to transit each road segment along the path, using a sample of users that is selected according to navigation zone characteristics, current data on road segments, and current behaviour of users on the road segment. We evaluate our model via simulations, and compare it with the model proposed in [1]. Regardless of the given probability value to obtain a time window, the road segment density and the number of road segments to handoff, HTEMOD provides a better accuracy and good duration of predicted time window when handoff will occur. Whilst the proposed HTEMOD model can be applied to any type of user equipment, its efficiency becomes more appealing in the context of vehicles (i.e., for the support of road to vehicles communications - RVC) or highly mobile nodes travelling in urban areas constrained by predefined roads and whose velocities are also restricted according to speed limits, level of congestion in roads, and traffic control mechanisms (e.g., stop signs and traffic lights).

Mobility-Aware Streaming Rate Recommendation System

In mobile multimedia streaming services, important requirements consist of the support of service continuity, the guarantee of acceptable Quality of Service (QoS) and insurance of steady Quality of Experience (QoE). How to get a uniform data exchange rate during the entire (or partial) course of a streaming service while a user is on the move is an important challenge. Generally speaking, the streaming rate of a multimedia service may heavily fluctuate due to the unavailability or deficiency of resources along the movement path of a user. To cope with this challenge, this paper proposes a framework that integrates user mobility prediction models with resource availability prediction models to keep a constant or less fluctuating streaming rate and to ultimately ensure steady QoE. Simulations are conducted to evaluate the performance of the proposed framework in achieving its design objectives and encouraging results are obtained.

Mobility prediction model-based service migration procedure for follow me cloud to support QoS and QoE

Follow me cloud (FMC) is one of the solutions to the limitations of the service quality and type of functionality due to the inherent constraints of mobility management. In this paper, we address the tradeoff between the overhead and Quality of Experience (QoE) by proposing a mobility-based services migration prediction (MSMP). The aim of MSMP is to plan, for a given service, the sequence of data to transfer from distinct micro data centers (MDC) to the user according to his mobility pattern and the estimated load of data centers. MSMP consists of (1) a data transfer throughput estimation scheme, called DTT, that aims to estimate, in advance, the throughput that the user could receive; (2) MDC service area handoff time estimation scheme, called AHT, that aims to estimate the time windows when the user will perform MDC service area handoffs; and (3) service migration management scheme, called SMM, that aims to select optimal MDCs in terms of offered throughput and to split user requested service into several portions to be processed by these MDCs. We evaluate MSMP and compare it against the most recent related approaches; the simulation results show that MSMP outperforms existing approaches in terms of data latency.