ElMoustapha Ould-Ahmed-Vall

The identification of performance issues on specific computer architectures has a variety of important benefits such as tuning software to improve performance, comparing the performance of various platforms and assisting in the design of new platforms. In order to enable this analysis, most modern micro-processors provide access to hardware-based event counters. Unfortunately, features such as out-of-order execution, pre-fetching and speculation complicate the interpretation of the raw data. Thus, the traditional approach of assigning a uniform estimated penalty to each event does not accurately identify and quantify performance limiters. This paper presents a novel method employing a statistical regression-modeling approach to better achieve this goal. Specifically, a model-tree based approach based on the M5' algorithm is implemented and validated that accounts for event interactions and workload characteristics. Data from a subset of the SPEC CPU2006 suite is used by the algorithm to automatically build a performance-model tree, identifying the unique performance classes (phases) found in the suite and associating with each class a unique, explanatory linear model of performance events. These models can be used to identify performance problems for a given workload and estimate the potential gain from addressing each problem. This information can help orient the performance optimization efforts to focus available time and resources on techniques most likely to impact performance problems with highest potential gain. The model tree exhibits high correlation (more than 0.98) and low relative absolute error (less than 8 %) between predicted and measured performance, attesting it as a sound approach for performance analysis of modern superscalar machines

Distributed event detection using wireless sensor networks has received growing interest in recent years. In such applications, a large number of inexpensive and unreliable sensor nodes are distributed in a geographical region to make firm and accurate local decisions about the presence or absence of specific events based on their sensor readings. However, sensor readings can be unreliable, due to either noise in the sensor readings or hardware failures in the devices, and may cause nodes to make erroneous local decisions. We present a general fault-tolerant event detection scheme that allows nodes to detect erroneous local decisions based on the local decisions reported by their neighbors. This detection scheme does not assume homogeneity of sensor nodes and can handle cases where nodes have different accuracy levels. We prove analytically that the derived fault-tolerant estimator is optimal under the maximum a posteriori (MAP) criterion. An equivalent weighted voting scheme is also derived. Further, we describe two new error models that take into account the neighbor distance and the geographical distributions of the two decision quorums. These models are particularly suitable for detection applications where the event under consideration is highly localized. Our fault-tolerant estimator is simulated using a network of 1024 nodes deployed randomly in a square region and assigned random probability of failures