Xavier Gelabert received the Telecommunications Engineering degree from the Universitat Politècnica de Catalunya (UPC), Barcelona, in 2004 and the MS degree in electrical engineering, with a major in wireless communications, from the Royal Institute of Technology (KTH), Stockholm, in 2003. In 2004, he joined the Radio Communication Research Group in the Department of Signal Theory and Communications, UPC, where he is pursuing his PhD. From August to December 2008 he was a visiting researcher at the Broadband Wireless Networking Laboratory (BWN-Lab) at Georgia Institute of Technology. Currently, he is a visiting researcher at the Instituto de Telecomunicaciones y Aplicaciones Multimedia (iTEAM), Universidad Politécnica de Valencia (UPV). His current research interests include the field of mobile radio communication systems, with a special emphasis on Common Radio Resource Management (CRRM) strategies in multiaccess networks, quality of service provisioning, and opportunistic/cognitive spectrum management. He has been actively involved in European-funded projects EVEREST, AROMA, and E3 along with Spanish projects COSMOS and COGNOS. He is a student member of the IEEE.
A list of publications by X. Gelabert is provided in the following:
Abstract: This work addresses the problem of radio access congestion control and resource allocation in scenarios where multiple available radio access technologies (RATs) support a wide range of services over a given coverage area. A key issue in these networks is selecting the most appropriate RAT at a call/session establishment according to some specified user/operator criteria. In this sense, a wide range of high-level policies can be defined, providing the most favorable resource allocation. Regardless of having efficient RAT selection policies, which may ensure some initial quality of service (QoS) requirements, intrinsic network dynamics (e.g., mobility, user activity, and interference rise) can cause potential QoS failures, leading to a degraded network performance and, hence, radio access congestion. This work is devoted to the study of the impact of radio access congestion on a number of RAT-selection policies. Consequently, a congestion probability (CP) model is developed to capture the statistical behavior of radio access congestion events. In addition, a general Markovian framework is adopted to evaluate the allocation of multiple services into multiple RATs by means of high-level policy definitions. Specific RAT-selection policies are defined according to several criteria, and their performances are evaluated in a time-division multiple-access (TDMA)/wideband code-division multiple-access (WCDMA) multi-RAT scenario supporting voice and data services. Moreover, the use of CP information as a possible allocation principle for RAT selection is also evaluated, which, in the assumed scenario, results in the most favorable allocation policy.
Abstract: This paper addresses the problem of opportunistic access of secondary users to licensed spectrum in cognitive radio networks. In order to avoid interference to the licensed primary users, efficient spectrum detection methods need to be developed. For this purpose, in recent years several sensing techniques have been proposed to monitor and regulate the spectrum access to the shared spectrum resources. However, spectrum sensing may be affected by errors in the form of missed-detections (i.e., an occupied spectrum is erroneously detected as free) or false-alarms (i.e., a free spectrum is erroneously detected as occupied). These two magnitudes pose a tradeoff on the design of the spectrum sensing mechanisms meaning that low missed-detection can only be achieved at the expense of high false-alarm and vice versa. Thus, the network designers should adaptively tune the sensing techniques such that the highest perceived Quality of Service (QoS) is achieved by both primary and secondary users. In this paper, a framework is introduced for determining the sensing operating points. Also the definition of Grade-of-Service (GoS) metrics is adopted to the case of primary/secondary users spectrum sharing. It is shown that the operating points of the sensing mechanisms can be easily adjusted according to the current traffic load of both primary and secondary users so that the perceived GoS is maximized. In addition, the Erlang Capacity of the spectrum sharing system for both primary and secondary users is also evaluated considering the effects of erroneous sensing.
Notes: Performance Modeling of Computer Networks: Special Issue in Memory of Dr. Gunter Bolch
Abstract: This paper addresses the problem of Radio Access Technology (RAT) selection in heterogeneous multi-access/multi-service scenarios. For such purpose, a Markov model is proposed to compare the performance of various RAT selection policies within these scenarios. The novelty of the approach resides in the embedded definition of the aforementioned RAT selection policies within the Markov chain. In addition, the model also considers the constraints imposed by those users with terminals that only support a subset of all the available RATs (i.e. multi-mode terminal capabilities). Furthermore, several performance metrics may be measured to evaluate the behaviour of the proposed RAT selection policies under varying offered traffic conditions. In order to illustrate the validation and suitability of the proposed model, some examples of operative radio access networks are provided, including the GSM/EDGE Radio Access Network (GERAN) and the UMTS Radio Access Network (UTRAN), as well as several service-based, load-balancing and terminal-driven RAT selection strategies. The flexibility exhibited by the presented model enables to extend these RAT selection policies to others responding to diverse criteria. The model is successfully validated by means of comparing the Markov model results with those of system-level simulations.
Abstract: Spectrum sensing is one key enabler towards opportunistic spectrum access in cognitive radio networks. Such scenarios allow cognitive users (a.k.a. secondary users) to access some licensed spectrum band as long as they do not interfere with the licensed (or primary) users. The main goal is to achieve an efficient and utmost access to the otherwise underutilized spectrum resources while still guaranteeing primary users a non-harmful operation. Spectrum sensing can be then used by secondary users to detect spectrum holes that may be accessed in a non-interfering manner. However, spectrum sensing may be subject to errors in the form of false-alarm and misdetection. False-alarm causes spectrum under-use while misdetection leads to spectrum interference between primary and secondary users. Unfortunately, these two magnitudes pose a trade-off on the sensing mechanism: low misdetection is achieved at the expense of high false alarm and vice versa. Consequently, an adequate operating point of the sensing mechanism should be determined. In this work we evaluate the impact of false-alarm and misdetection errors on the performance of a spectrum sensing scenario. We use a discrete time Markov chain (DTMC) model and we determine the suitable operating point for the sensing mechanism under different traffic load conditions such that some quality of service is attained by both primary and secondary users. Performance results reveal that by effectively choosing the operation point bearing in mind the traffic load levels will lead to enhanced perceived quality of service of both primary and secondary users.
Abstract: The benefits of jointly managing the combined radio resources offered by heterogeneous networks consisting of several Radio Access Technologies (RATs) have been profusely studied and assessed in recent years. Nevertheless, most of the existing work assumes scenarios where all RATs are accessible (provided the RAT is not at full capacity) to all users demanding service. If this is so, the obtained benefits become rather optimistic given that we neglect the fact that the deployed RATs may have different coverage overlapping conditions among them and that users may not have terminals that support all RATs (i.e. multimode terminals). In this paper we extend a previously developed Markov framework in order to capture the effect of having different coverage overlapping conditions along with the capability of certain terminals to support all or a subset of available RATs. As a result, we assess the degradation, in terms of Erlang capacity, that a heterogeneous network undergoes in scenarios with limited terminal and coverage conditions and compare it to the ideal case of full coverage and full terminal availability.
Abstract: An efficient and utmost utilization of radio spectrum resources has stimulated the introduction of what has been termed dynamic spectrum access methodologies and implementations. While the traditional approach has been based on licensed (or primary) spectrum access, this new communication paradigm enables an opportunistic secondary access to shared spectrum resources provided mutual interference is kept below predefined margins. In this paper we propose a novel and flexible framework so as to account for primary-secondary spectrum sharing scenarios. In this sense, the use of a Discrete Time Markov Chain (DTMC) model is suggested and further justified. Some illustrative results are provided and validated against a system-level simulator thus confirming the suitability of the proposed approach.
Abstract: Common Radio Resource Management (CRRM) represents a key functionality in heterogeneous Beyond 3G scenarios where several radio access technologies coexist. In this context, this paper presents a general CRRM functional model, accompanied by different splits of functionalities between the involved entities. This paper also introduces several principles for the Radio Access Technology (RAT) selection problem and presents a specific algorithm that is able to combine in a smart way the service, load-balancing and interference principles. Throughput improvements of up to 24% compared to other reference approaches are obtained with the proposed algorithm.
Abstract: Among the common radio resource management (CRRM) functions that are responsible for the proper allocation of resources in a multi-access network, congestion control is the one devoted to overcome potential QoS failures due to the inherent dynamics of the network. In this paper we address the problem of congestion control in a scenario considering the GSM/EDGE radio access network (GERAN) and the UMTS terrestrial radio access network (UTRAN). In particular, we face the problem where the two available radio access technologies (RATs) undergo simultaneous congestion situations. For this case, a congestion resolution scheme based on vertical (inter-system) handover (VHO) jointly with a bit-rate reduction (BRR) scheme is proposed and evaluated for a mixed services scenario considering voice and data users.
Abstract: In order to support the conceptual development of Common Radio Resource Management (CRRM) algorithms, this paper provides an analytical approach to the performance evaluation of Radio Access Technology (RAT) selection procedures in a multi-RAT/multiservice environment. In particular, a 4-Dimensional (4D) Markovian model is devised so as to consider the allocation of voice and data services in a GERAN/UTRAN system. Through the analytical definition of well-established Key Performance Indicators (KPIs) we provide numerical results on the evaluation of a load balancing RAT allocation policy.
Abstract: This paper addresses the problematic of congestion control in the radio access interface when considering the allocation of voice and data services over several radio access technologies (RATs). In particular, the GSM/EDGE radio access network (GERAN) and the UMTS terrestrial radio access network (UTRAN) are considered for the evaluation of congestion control strategies. After a congestion situation in the radio access is detected, congestion resolution mechanisms are triggered in order to reduce the overload in the congested RAT(s). In this paper, a framework for the detection and resolution of congestion conditions in a multi-access network is presented. Moreover, three approaches intending to solve congestion situations are proposed and the evaluation of an inter-RAT handover algorithm for solving congestion events in GERAN is also presented.
Abstract: Beyond 3G (B3G) networks will encompass, among other features, a wide range of radio access technologies (RATs) providing users with a flexible and efficient access to the increasing pool of demanding services. This will allow users to get connected using the access technology that is most suitable according to some specified criteria. Consequently, to take full advantage of B3G networks, mobile terminals will need to support a larger set of capabilities. Among those, the support of different RATs, i.e. multi-mode capacity, is a must. This paper addresses the impact of multi-mode terminals in an EDGE/UMTS heterogeneous network with multi-service provisioning. Results indicate that multi-mode terminal availability should be considered when designing common radio resource management strategies in heterogeneous wireless access networks. Specifically, a service-based initial RAT selection policy is evaluated, revealing different behaviors for different multi-mode terminal availabilities and service-class mixings. In order to compensate the limitations imposed by non-multi-mode terminals, it is suggested to actuate over GERAN by using a resource reservation scheme for interactive users. By doing so, we tradeoff the QoS between multi-service/multi-mode users.
Abstract: Initial RAT selection is a key Common RRM strategy, where users requiring service are to be efficiently allocated in the existing and available RATs. Although load balancing is a possible guiding allocation principle, sometimes it may not be convenient. This paper evaluates and compares a load balancing and a service-class RAT selection policy in order to discuss the suitability of the former in situations where different service-class mixings are present. Results indicate a tight dependency between this appropriateness and the mixing of demanding services.
Abstract: Future military tactical communication networks must be highly mobile, survivable and recongurable. Distributed multi-hop packet radio architectures have been adopted, combining robustness and autonomous operation. In order to meet the requirements for Quality of Service (QoS) and delay guarantees, Spatial Time Division Multiple Access (STDMA) schemes have been proposed. Key problems in STDMA are: to nd ecient distributed algorithms for STDMA scheduling, slot synchronisation and to handle mobility. In this thesis focus is on slot synchronisation. Time synchronisation is a critical piece of infrastructure for any distributed system. Moreover, slotted TDMA schemes are of special interest because of the natural mechanism it provides for refereeing the access to the medium: time. To consider timing inaccuracies and propagation delay effects a guard band is usually introduced and, therefore, perfect synchronisation is not required. Hence, good time synchronisation is important not only because it enables Time Division Multiple Access to the data link, but because it shortens the guard time allowing bigger packets to be sent. In this thesis the performance in terms of synchronisation convergence and timing accuracy will be evaluated for a STDMA scheme in Tactical Radio Area Networks (TRAN). A network set-up environment will be considered and a description of how the synchronisation algorithm ts in the initialization process will be made. We also investigate some parameters related to the synchronisation algorithm and the effects when different topology congurations are used.
