When Moshe Vardi asked me to write this piece for CACM, my first reaction was the article could be written in two words Still open.

The goal of this paper is to lay a foundation for rigorous empirical software engineering. I do this by introducing several theories and their models. I first present an abstract theory TM about theories and models and related processes. I then apply TM to itself, yielding MTM, a theory and model about TM. These two theories provide a unified approach to design disciplines. I then provide 4 models of them relevant to empirical software engineering: products, developments, instruments and experiments. I introduce theory E and apply E to TM and MTM yielding ETM and EMTM. These latter two theories provide a taxonomy of empirical studies for design disciplines. I discuss briefly models of such empirical studies. I then apply E to itself yielding EE that provides a taxonomy of evaluations of empirical studies. Finally, I present a list of challenges for empirical software engineering research.

Abstract—In this paper, we introduce a high-level simulation methodology for the modeling of multicore video processing architectures. This method allows design space explorations of parallel video processing applications (VPAs). It is used to test the performance of running a VPA on arbitrary virtual hardware and software configurations. The method represents an alternative to performing a “complete ” decoder implementation on a field-programmable gate array or an applicationspecific integrated circuit. The use of our method, therefore, yields the advantage of being considerably more time, labor, and cost efficient. As an application, we use our method for designing a parallel H.264 decoder targeting 720p25 resolution at bit-rates up to 50 Mb/s. Starting from a single-core decoder implementation, we use our simulator for estimating the performance gain when using a multicore architecture. We then detect the major performance bottlenecks in this multicore system and perform additional decoder splittings accordingly until we reach the targeted requirements. The use of functional splitting (i.e., pipelining) and data-parallel processing is demonstrated. The final H.264 decoder architecture is capable of fulfilling our performance requirements. Index Terms—Design exploration, H.264 decoder, parallel architectures, video processing.

Abstract — The reliability of distributed systems in which the communication links are considered reliable while the computing nodes may fail with certain probabilities have been modeled by a probabilistic network or a graph G. Computing the residual connectedness reliability (RCR), denoted by R(G), of probabilistic networks with unreliable nodes is very useful, but is an NP-hard problem. To derive the exact R(G) expressions for large networks can become rather complex. As network size increases, the reliability bounds could be used to estimate the reliability of the networks. In this paper, we present an efficient algorithm for computing the reliability upper bound of distributed systems with unreliable nodes. We also apply our algorithm to some typical classes of graphs in order to evaluate the upper bound and show the effectiveness and the efficiency of the new algorithm. Numerical results are presented. Keywords- distributed systems; residual connectedness reliability; probabilistic network; upper bound I.