Sharif Digital Repository

This paper presents transient and permanent fault injection into Verilog models of digital systems during the design phase by a developed simulation-based fault injection tool called INJECT. With this fault injection tool, it is possible to inject crucial fault models in all abstraction levels (such as swith-level) supported by Verilog HDL. Several fault models for injecting into Verilog models are specified and described. Analyzing the results obtained from the fault injections, using INJECT enables system designers to inform from dependable parameters, such as fault latency, propagation and coverage. As a case study, a 32-bit processor, namely DP32, has been evaluated and effects of faults... 

Caches, which are comprised much of a CPU chip area and transistor counts, are reasonable targets for transient single and multiple faults induced from energetic particles. This paper presents: (1) a new fault detection scheme for tag arrays of cache memories and (2) an architectural cache to improve performance as well as dependability. In this architecture, cache space is divided into sets of different sizes and different tag lengths. Using the proposed fault detection scheme, i.e., GParity, when single and multiple errors are detected in a word, the word is rewritten by its correct data from memory and its GParity code is recomputed. The error detection scheme and the cache architecture... 

Reconfigurability for a component of a distributed component-based system means ability to replace the component with a new one, in presence of the other components. In this context, substitutability and compatibility of components should be analyzed. In this paper, we use I/O-Constraint Automata to model behavior of components and propose enough formalism for substitutability and compatibility analysis. We show that in the context of distributed non-blocking components, there is a relation weaker than usual simulation relation in automata theory for notion of substitutability. We illustrate the relation between substitutability and compatibility to reason about the reconfigurability of a... 

The capability of placing "comments" on the posts makes the blogspaces rather a complex environment. One of the interesting phenomena in blogspace is "blogger failure" when a blogger stops writing after a certain amount of time and will not return to blogspace for a long time, or when a blogger does not get any comment from her audience. In this paper we illustrate our observation on bloggers failure in a unique blogspace. First, we introduce PersianBlog blogspace briefly along with our observations of behaviors of bloggers on placing comments. Next, we will provide our definition of failure, and give a broad future research path to model failure in blogspace  

Due to the necessity for methods that are able to handle complexities of systems, use of Method Engineering (ME) to construct more adaptable and flexible methods has become of importance. Assemblybased approach is the most well-known approach of ME that performs method construction by reusing and assembling different method fragments. This approach, like other approaches of ME, suffers from the lack of suitable guidelines. In this way, architecture-centric method engineering (ArCME) is a new approach which aims at performing ME processes in a more disciplined and cohesive way. In this paper, process of assemblybased ME is enhanced by ArCME. Furthermore, benefits of implementing the process... 

Dealing with virtual channels has always been a critical issue in developing analytical performance models for interconnection networks. Almost all previous studies relied on a method proposed by Dally to capture the effect of virtual channels multiplexing  

The important problem of content location in an unstructured distributed network (such as P2P, with no central searching system, graph construction control or file placement regulation) is addressed. The existing methods that are based on either flooding or random walk impose a large traffic load on the network or require a long time to search. In this work, we propose a method that outperforms the existing methods in both of these measures. These advantages are achieved by publishing traces of the contents of each node in its nearby nodes, so that the closer nodes to the target have a stronger trace of it. Such traces are used to rapidly locate the desired content. The superb performance of... 

This paper presents a simulation-based environment to study fault effects in message missing failures in CAN-based systems. A CAN controller is modeled by VHDL at behavioral level and is exploited to set up a network composed of several nodes. A total of 27,000 transient faults of seven types are injected into five critical portions of the system including the bus and four portions of the CAN controller. The experimental results show that the faults affect the message sending in which more than 20% of faults cause the failure. Besides, with a heavy workload, faults that are occurred into the CAN controller and on the bus cause about 90% and 10% of all failures, respectively  

We find exact solutions for a universal set of quantum gates on a scalable candidate for quantum computers, namely an array of two-level systems. The gates are constructed by a combination of dynamical and geometrical (non-Abelian) phases. Previously these gates have been constructed mostly on nonscalable systems and by numerical searches among the loops in the manifold of control parameters of the Hamiltonian. © 2005 The American Physical Society  

Dealing with deadlock has always been a crucial issue for any routing algorithms proposed for wormhole-switched networks. Adaptive routing algorithms, based on deadlock avoidance strategies, usually dedicate resources specifically to ensure deadlock freedom. However, a number of recent studies have demonstrated that deadlocks are quite rare in the network. This fact has motivated researchers to introduce adaptive routing algorithms based on deadlock recovery strategies. In an effort to gain a deep understanding of the factors that affect routing performance, a comparison of the performance merits of deadlock recovery against deadlock avoidance routing algorithms is performed. While most... 

Combination of three-dimensional (3D) IC technology and network on chip (NoC) is an effective solution to increase system scalability and also alleviate the interconnect problem in large-scale integrated circuits. However, because of the increased power density in 3D NoC systems and the destructive effect of high temperatures on chip reliability, applying thermal management solutions becomes crucial in such circuits. In this study, the authors propose a runtime distributed migration algorithm based on game theory to balance the heat dissipation among processing elements (PEs) in a 3D NoC chip multiprocessor. The objective of this algorithm is to minimise the 3D NoC system's peak temperature,... 

Packet loss recovery is an important part of P2P video streaming networks due to inevitable packet loss in today's internet and interdependency of data units in compressed video streams. In addition, the architecture of P2P streaming networks, in which the data delivered to the receivers through chain of peers, can intensify the impact of the internet packet loss on the quality of perceived video at the receivers. FEC and ARQ are the two most important techniques that can be used to overcome the side effect of the internet packet loss in P2P video streaming networks. Based on these two techniques, different packet loss recovery strategies can be applied in different overlay hops of a given... 

Combination of three-dimensional (3D) IC technology and network on chip (NoC) is an effective solution to increase system scalability and also alleviate the interconnect problem in large-scale integrated circuits. However, because of the increased power density in 3D NoC systems and the destructive effect of high temperatures on chip reliability, applying thermal management solutions becomes crucial in such circuits. In this study, the authors propose a runtime distributed migration algorithm based on game theory to balance the heat dissipation among processing elements (PEs) in a 3D NoC chip multiprocessor. The objective of this algorithm is to minimise the 3D NoC system's peak temperature,... 

In this paper, a novel neuro-fuzzy computing system is proposed where its learning is based on the creation of fuzzy relations by using a new implication method without utilizing any exact mathematical techniques. Then, a simple memristor cross-bar-based analog circuit is designed to implement this neuro-fuzzy system which offers very interesting properties. In addition to high connectivity between neurons and being fault tolerant, all synaptic weights in our proposed method are always non-negative, and there is no need to adjust them precisely. Finally, this structure is hierarchically expandable, and it can do fuzzy operations in real time since it is implemented through analog circuits.... 

Inter-wire coupling capacitance may lead to crosstalk faults that significantly limits the reliability of NoCs. In this paper, we propose a numerical-based crosstalk avoidance code that can omit the Triplet Opposite Direction (TOD) transitions produced by crosstalk faults. The proposed coding does not have ambiguity and uses all of the codeword space. Simulations using VHDL for different channel widths show that the proposed method can reduce crosstalk fault in the NoC links with negligible power and area overheads  

This paper proposes an efficient mechanism that mitigates crosstalk faults in Network-on-Chips (NoCs). This is done by using a Packet Manipulating mechanism called PAM for reliable data transfer of NoCs. PAM investigates the transitions of a packet to minimize the forbidden transition patterns appearing during the flit traversal in NoCs. To do this, the content of a packet is manipulated using three different manipulating mechanisms. In other words, PAM manipulates the content of packet in three manipulating modes including: vertical, horizontal and diagonal modes. Then, comparing the transitions of these manipulating mechanisms, a packet with minimum numbers of transitions is selected to be... 

We consider a wireless distributed computing system, in which multiple mobile users, connected wirelessly through an access point, collaborate to perform a computation task. In particular, users communicate with each other via the access point to exchange their locally computed intermediate computation results, which is known as data shuffling. We propose a scalable framework for this system, in which the required communication bandwidth for data shuffling does not increase with the number of users in the network. The key idea is to utilize a particular repetitive pattern of placing the data set (thus a particular repetitive pattern of intermediate computations), in order to provide the... 

We consider the problem of massive matrix multiplication, which underlies many data analytic applications, in a large-scale distributed system comprising a group of worker nodes. We target the stragglers' delay performance bottleneck, which is due to the unpredictable latency in waiting for slowest nodes (or stragglers) to finish their tasks. We propose a novel coding strategy, named entangled polynomial code, for designing the intermediate computations at the worker nodes in order to minimize the recovery threshold (i.e., the number of workers that we need to wait for in order to compute the final output). We demonstrate the optimality of entangled polynomial code in several cases, and show... 

The rule based scheduling algorithms are a new trend in grid scheduling algorithms; the combination of rule based algorithms for resource selection with various dispatching rules for queuing of jobs can improve or deteriorate their performance. Thus, choosing a proper queuing strategy for each algorithm is a prominent issue in scheduling. In this paper, we introduce two new dispatching rules for resource selection and three new dispatching rules for queuing of jobs; we evaluate the performance of various combinations of these new rule based scheduling algorithms and queuing strategies. Also, we use some major combination of rule based scheduling algorithms with some important queuing... 

Peer-to-peer networks are beginning to form the infrastructure of future applications. The use of semantic information to organize nodes in a P2P network is a prosperous approach. In this paper we present a novel method, called Semantic Partition Tree, which organizes nodes in a semantic space, using an upper ontology. In this approach each node can find a target in a time proportioned to the depth of the upper ontology, which is significantly less than the number of nodes in the network. Also we simulated our method with several configurations and measured its performance. The results of the simulation show that our scheme is highly scalable and can be used by real P2P networks. © 2007 IEEE...