نشریه محاسبات نرم
پیاپی 1 (بهار و تابستان 1391)

At the present، energy aware routing is one of the most important research topics in wireless sensor networks. Increasing network lifetime is the most challenging need in these networks. In this study، in order to increase network lifetime، we aim to introduce a harmony search algorithm as a successful meta-heuristic algorithm for routing in the wireless sensor networks. In customizing the algorithm for routing، two following criteria، which result in increasing network lifetime، have been considered: reducing energy consumption and proper distribution of energy consumption between sensor nodes. Simulation results clearly showed the ability of this algorithm in finding the optimum route and proper balance between the abovementioned criteria. Moreover، to compare the recommended method with other methods in identical conditions، an implementation of a genetic algorithm has been carried out. Results showed the better performance of harmony search algorithm in increasing network lifetime in compared with genetic algorithm

In wireless sensor networks، the collected data are stored in the network nodes for a long time until the receivers take them. Since some network nodes may become accidentally corrupt or some mobile adversary intrudes to some node، data survival and secure and complete data transmission to some receiver is a concern. Moreover، we should consider communication efficiency and the needed store for the survival. This specific feature distinguishes the unattended wireless sensor networks from others. In 2011، Ren et al. presented a proposal for maximizing survival duration and stability of the collected data by sensor network nodes. In this paper، we first show two main weaknesses of the Ren et al proposal: (1) the propagated data in the network is not authenticated; this weakness causes the network becomes more susceptible to attack and (2) data are not identified when they become corrupt during transmission. These weaknesses cause more traffic in the network and result in the network inefficiency. Having shown the weaknesses، we present a method to resolve them using the homomorphic signature.

University course timetabling problem is a complicated problem and finding a computer-aided solution for it was a subject to work for many years. To solve this problem، we must assign courses to timeslots with respect to hard and soft constraints. Hard constraints are those which must be necessarily met (some of them could be neglected with high costs). Our aim is to meet as many soft constraints as possible. In this paper we present local search algorithms to improve a table which meets hard constraints، and is provided by a searching algorithm for CSPs. Our results show that the proposed method is very suitable for real data with a large size and complex constraints.

In this paper، the notion of the digital divide has been described، and a few analyzing methods of digital divide have been reviewed. Analyzing methods of digital divide are called indices which have different indicators and different formulas for calculation. Since data collection for an indicator may be difficult، calculating an index is an essential problem. We collected and calculated some indicators in provinces of Iran. But they were insufficient to calculate a standard index. These indicators terribly show the deep digital divide between the provinces. To show more accurately the social inequalities in the adoption of ICT between provinces in Iran، we used the well-known K-means clustering algorithm on the indicators of the provinces. The clustering results appropriately showed the unique status of Tehran among provinces because Tehran always falls in a different cluster alone. It means that the information technology does not fairly spread through the provinces in Irān.

In this paper the development of quantum neural networks (QNN)، and some of presented models and physical implementation are reviewed. How of making use of double-slit experiment for implementing QNN and methods of designing as well as examples of two-layer hybrid networks in QNN constructed from quantum neurons and classical neurons are represented. Some application models of the networks (QNN) is compared with classical models and capabilities of QNN together with the quantum mechanical concepts in solving difficult problems، which is hard to solve by classical models، is noted.

Electronic devices in space applications may be Radiation Tolerant or Commercial off-the Shelf (COTS). Due to cost and unavailability in many applications، the latter is usually used. In applications such as spatial ones، the COTS equipment lacks reliability against threats like heavy ion radiation; therefore، some alternatives should be considered to make the equipment resistant against the probable threats. In this paper to checking an Intra-Inter block control flow، a software redundancy method in instruction level is delivered. In comparison with other methods، the proposed method has less memory overhead، better performance and more fault coverage.