Celso Maciel da Costa

Bio: Celso Maciel da Costa is an academic researcher from Pontifícia Universidade Católica do Rio Grande do Sul. The author has contributed to research in topics: Middleware & Mobile computing. The author has an hindex of 3, co-authored 5 publications receiving 14 citations.

DEVOpT: a distributed architecture supporting heuristic and metaheuristic optimization methods

Abstract: This paper presents a distributed software architecture that allows the cooperation among research institutions in the field of Combinatorial Optimization --- DEVOpT: Distributed Evolutionary Optimization Centers. It has as main aims to share existing algorithms for optimization problems, to allow the easy testing of these algorithms with existing instances, to provide fast and better ways to design new algorithms, and to share computational power among the cooperating institutions. This is achieved respecting the autonomy and heterogeneity of the cooperating institutions. The distributed architecture is discussed here and also a case study of a Parallel Memetic Algorithm to solve the Asymmetric Traveling Salesman Problem (ATSP) running on this environment is analyzed.

A reflective middleware architecture to support adaptive mobile applications

Abstract: Mobile applications are required to operate in environments in which the availability for resources and services may change significantly during system operation. As a result, mobile applications need to be capable of adapting to these changes to offer the best possible level of service to their users. However conventional middleware is limited in its capability of adapting to the environment changes and different users requirements. Computational reflection applied to middleware design has introduced a new research field, reflective middleware. In this paper we propose a reflective middleware architecture which can be used to develop adaptive mobile applications. In order to validate the architecture proposed we developed a prototype using the Web Services technology which focuses on the problem of adapting on a set of attributes in a coordinated manner.

A Reflective Middleware Architecture for Adaptive Mobile Computing Applications

Abstract: Mobile computing applications are required to operate in environments in which the availability for resources and services may change significantly during system operation. As a result, mobile computing applications need to be capable of adapting to these changes to offer the best possible level of service to their users. However, traditional middleware is limited in its capability of adapting to the environment changes and different users requirements. Computational Reflection paradigm has been used in the design and implementation of adaptive middleware architectures. In this paper, we propose an adaptive middleware architecture based on reflection, which can be used to develop adaptive mobile applications. The reflection-based architecture is compared to a component-based architecture from a quantitative perspective. The results suggest that middleware based on Computational Reflection can be used to build mobile adaptive applications that require only a very small overhead in terms of running time as well as memory space.

A Comparative Analysis of Adaptive Middleware Architectures Based on Computational Reflection and Aspect Oriented Programming to Support Mobile Computing Applications

Abstract: Mobile computing applications are required to operate in environments in which the availability for resources and services may change significantly during system operation. As a result, mobile computing applications need to be capable of adapting to these changes to offer the best possible level of service to their users. However, traditional middleware is limited in its capability of adapting to environment changes and different users requirements. Computational Reflection and Aspect Oriented Programming paradigms have been used in the design and implementation of adaptive middleware architectures. In this paper, we propose two adaptive middleware architectures, one based on reflection and other based on aspects, which can be used to develop adaptive mobile applications. The reflection based architecture is compared to an aspect oriented based architecture from a quantitative perspective. The results suggest that middleware based on Aspect Oriented Programming can be used to build mobile adaptive applications that require less processor running time and more memory space than Computational Reflection while producing code that is easier to comprehend and modify.

Dynamic context bindings : infrastructural support for context-aware applications

Abstract: The world is increasingly equipped with high-capacity, interconnected, mobile and embedded computing devices. Context-awareness provides an attractive approach to personalize applications such that they better suit the user’s needs in this rich computing environment. Context-aware applications use context information, offered by context sources, to adapt their behavior to the situation at hand. The exchange of context information requires an association between a context consuming context-aware application and suitable context producing context sources. We call these associations ‘context bindings’. Developing context-aware applications is complex due to some intrinsic characteristics of context sources. Firstly, context sources are distributed. Consequently, creating a context binding requires some form of discovery and selection of context sources. Secondly, context sources are arbitrary available during the life-span of the application. This makes a binding hard to maintain. Finally, context sources offer context information with a fluctuating quality. This makes a binding possibly unsuitable for the application. Currently, developers need to spend considerable effort to develop application code to create and maintain required context bindings, which can deal with these complexities. This thesis provides insights in the generic characteristics of contextaware applications and their development process. We propose an abstraction, called the Context Binding Transparency. This transparency has as goal to mask the complexities of creating and maintaining context bindings for the application developer. In this way, we facilitate the development process of context-aware applications. The responsibility for creating and maintaining context bindings is relieved from the application developer and is shifted to a context binding infrastructure. This enables application developers to focus on the development of primary application logic rather than the logic needed to create and maintain context bindings. The application developer interacts with the context binding infrastructure using context retrieval and publishing services, and a context requirement specification language. This language enables application developers to specify their requirement at a high level of abstraction rather than in programming code. In this thesis, we propose a realization of such a context requirement specification language, coined the Context Binding Description Language (CBDL). This language is developed to be generic for a broad range of context-aware applications. Additionally, we propose a realization of a context binding infrastructure called the Context-Aware Component Infrastructure (CACI). This infrastructure realizes a context binding transparency and is composed of a context binding mechanism and a context discovery interoperability mechanism. The context binding mechanism uses CBDL documents, specified by the application developers, to create and maintain context bindings on behalf of the application. The process of creating a binding consists of discovery of context sources at available context discovery mechanisms, selection of suitable context sources, establishment of a binding of the application to these context sources, and maintenance of this binding. Maintenance of a context binding includes re-binding to other suitable context sources in case of lost or (re-)appearing context sources and fluctuating quality of context. This thesis gives an example of a possible rebinding algorithm. The context discovery interoperability mechanism enables contextaware applications to interoperate transparently with different context discovery mechanisms available in the application environment. The goal of the interoperability mechanism is to hide the heterogeneity and fluctuating availability of context discovery mechanisms for context-aware applications. The context discovery interoperability mechanism is a supporting mechanism for the context binding mechanism. It can also be used independently by context-aware applications that do not leverage from the context binding mechanism. We have created a proof-of-concept prototype of CACI, using the OSGi component framework. The prototype includes implementations of the context binding mechanism and the context discovery interoperability mechanism. Evaluation of the proposed context binding transparency and infrastructure consists of a user survey and a comparison on the development effort and software quality of a Telemedicine case implementation with and without CACI. The survey indicated a general interest of possible users in the features of the context binding transparency. The case implementations indicated a possible improvement in the development process of higher quality context-aware applications when using a context binding infrastructure. This research stresses that the availability of context information and the quality of this information highly influences the development of contextaware applications. By using a middleware infrastructure to support the creation and maintenance of context bindings, the development of higher quality context-aware applications can be simplified.

Position Management für ortsbezogene Community-Dienste

Abstract: In Location-based Community Services (LBCSs) mobile users interchange and correlate their spatial positions, for example, in order to find out which other community members are currently staying nearby. The so-called position management is responsible for the transmission, analysis, processing and access control of position information, which is directed along a corresponding supply chain. The supply chain spans from the mobile device of the target person, where the position is derived, for example, by GPS, via intermediaries like the location or LBS provider, to the domain of the user. Community services pose special requirements on position management, which can be coarsely divided into the fields privacy protection and efficiency: First, the target person must be able to control by who and under which circumstances her position information is accessed. To guarantee that, it must be possible to anonymize the position data with respect to the location and LBS provider, for which so far no technique exists that is suited for community services. Also, the target person must be able to authorize requests to access her position in an easy and socially acceptable fashion. Second, concepts for efficiently realizing so-called proactive multi-target LBCSs are needed. These services are automatically triggered as soon as two or more target person have entered into a certain pre-defined spatial constellation. An example is buddy tracking, which automatically detects when two persons have approached each other below a certain proximity distance. The technical problem to solve is the frequent transmission of position information over the scarce air-interface and the associated energy consumption at the mobile terminal of the target person. This dissertation develops new concepts in both of the sketched fields and shows their feasibility based on numerous simulations and analytical reflection. Also the TraX-platform is presented, which practically implements the developed concepts.

Analytical study of the "master-worker" framework scalability on multiprocessors with distributed memory.

Abstract: The paper is devoted to an analytical study of the "master-worker" framework scalability on multiprocessors with distributed memory. A new model of parallel computations called BSF is proposed. The BSF model is based on BSP and SPMD models. The scope of BSF model is the compute-intensive applications. The architecture of BSF-computer is defined. The structure of BSF-program is described. The Using this metric, the upper scalability bounds of BSF programs on distributed memory multiprocessors are evaluated. The formulas for estimating the parallel efficiency of BSF programs also proposed.