With multiprocessors and multicore processors becoming ubiquitous, focus has shifted from research on uniprocessors to that on multiprocessors. Results derived for the uniprocessor case unfortunately do not always directly extend to the multiprocessor case in a straightforward manner. This necessitates a paradigm shift in the approach used to design and analyse the behaviour of such processors. In the case of Real-time systems, that is, systems characterised by explicit timing constraints, analysis and performance guarantees are more important, as failure is unacceptable. Scheduling algorithms used in Real-time systems have to be carefully designed as the performance of the system depends critically on them. Efficient tests for determining if a set of tasks can be feasibly scheduled on such a computing system using a particular scheduling algorithm thus assumes importance. Traditionally, the ‘task utilization’ parameter has been used for devising such tests. Utilization based tests for Earliest Deadline First(EDF) and Rate Monotonic(RM) scheduling algorithms are known and are well understood for uniprocessor systems. In our work, we derive limits on similar bounds for the multiprocessor case. Our work diners from previous literature in that we explore the case when the individual processors constituting the multiprocessor need not be identical. Each processor in such a system is characterised by a capacity, or speed, and the time taken by a task to execute on a processor is inversely proportional to its speed. Such instances may arise during system upgradation, when faster processors may be added to the system, making it a non-identical multiprocessor, or during processor design, when the different cores on the chip may have different processing power to handle dynamic workloads. We derive results for the partitioned paradigm of multiprocessor scheduling, that is, when tasks are partitioned among the processors, and interprocessor migration after a part of execution is completed is not allowed. Results are derived for both fixed priority algorithms(RM)and dynamic priority algorithms (EDF) used on individual processors. A maximum and minimum limit on the bounds for a ‘greedy’ class of algorithms are established, since the actual value of the bound depends on the algorithm that allocates the tasks. We also derive the utilization bound of an algorithm whose bound is close to the upper limit in both cases. We find that an expression for the utilization bound can be obtained when EDF is used as the uniprocessor scheduling algorithm, but when RM is the uniprocessor scheduling algorithm,an O(mn) algorithm is required to find the utilization bound, where m is the number of tasks in the system and n is the number of processors. Knowledge of such bounds allows us to carry out very fast schedulability tests, although we have the limitation that the tests are sufficient but not necessary to ensure schedulability. We also compare the value of the bounds with those achievable in ‘equivalent’ identical multiprocessor systems and find that the performance guarantees provided by the non-identical multiprocessor system are far higher than those offered by the equivalent identical system.

Firm real time system with Poisson arrival process, iid exponential service times and iid deadlines till the end of service of a job, operated under the First Come First Served (FCFS) scheduling policy is well studied. In this thesis, we present an exact theoretical analysis of a similar (M/M/1 + G queue) system with exact admission control (EAC). We provide an explicit expression for the steady state workload distribution. We use this solution to derive explicit expressions for the loss ratio and the sojourn time distribution. An exact theoretical analysis of the performance of an M/M/1 + G queue with preemptive deadlines till the end of service, operating under the Earliest Deadline First (EDF) scheduling policy, appears to be difficult, and only approximate formulas for the loss ratio are available in the literature. We present in this thesis similar approximate formulas for the loss ratio in the present of an exit control mechanism, which discards a job at the epoch of its getting the server if there is no chance of completing it. We refer to this exit control mechanism as the Early job Discarding Technique (EDT). Monte Carlo simulations of performance indicate that the maximum approximation error is reasonably small for a wide range of arrival rates and mean deadlines. Finally, we compare the loss ratios of the First Come First Served and the Earliest Deadline First scheduling policies with or without admission or exit control mechanism, as well as their counterparts with deterministic deadlines. The results include some formal equalities, inequalities and some counter-examples to establish non-existence of an order. A few relations involving loss ratios are posed as conjectures, and simulation results in support of these are reported. These results lead to a complete picture of dominance and non-dominance relations between pairs of scheduling policies, in terms of loss ratios.

Dans ce travail, nous nous focalisons sur l’ajout de contraintes temporelles dans les Bases de Données de Capteurs Sans Fil (BDCSF). La cohérence temporelle d’une BDCSF doit être assurée en respectant les contraintes temporelles des transactions et la validité temporelle des données, pour que les données prélevées par les capteurs reflètent fidèlement l’état réel de l’environnement. Cependant, les retards de transmission et/ou de réception pendant la collecte des données peuvent conduire au non-respect de la validité temporelle des données. Une solution de type bases de données s'avère la plus adéquate. Il faudrait pour cela faire coïncider les aspects BD traditionnelles avec les capteurs et leur environnement. À cette fin, les capteurs déployés au sein d'un réseau sans fils sont considérés comme une table d'une base de données distribuée, à laquelle sont appliquées des transactions (interrogations, mises à jour, etc.). Les transactions sur une BD de capteurs nécessitent des modifications pour prendre en compte l'aspect continu des données et l'aspect temps réel. Les travaux réalisés dans cette thèse portent principalement sur trois contributions : (i) une étude comparative des propriétés temporelles entre une collecte périodique des données avec une base de données classique et une approche de traitement des requêtes avec une BDCSF, (ii) la proposition d’un modèle de traitement des requêtes temps réel, (iii) la mise en œuvre d’une BDCSF temps réel, basée sur les techniques décrites dans la deuxième contribution. / In this thesis, we are interested in adding real-time constraints in the Wireless Sensor Networks Database (WSNDB). Temporal consistency in WSNDB must be ensured by respecting the transaction deadlines and data temporal validity, so that sensor data reflect the current state of the environment. However, delays of transmission and/or reception in a data collection process can lead to not respect the data temporal validity. A database solution is most appropriate, which should coincide with the traditional database aspects with sensors and their environment. For this purpose, the sensor in WSN is considered as a table in a distributed database, which applied transactions (queries, updates, etc.). Transactions in a WSNDB require modifications to take into account of the continuous datastream and real-time aspects. Our contribution in this thesis focus on three parts: (i) a comparative study of temporal properties between a periodic data collection based on a remote database and query processing approach with WSNDB, (ii) the proposition of a real-time query processing model, (iii) the implementation of a real time WSNDB, based on the techniques described in the second contribution.