Minimum attribute reduction algorithm based on quick extraction and multi-strategy social spider optimization

Intelligent optimization algorithm combined with rough set theory to solve minimum attribute reduction (MAR) is time consuming due to repeated evaluations of the same position. The algorithm also finds in poor solution quality because individuals are not fully explored in space. This study proposed an algorithm based on quick extraction and multi-strategy social spider optimization (QSSOAR). First, a similarity constraint strategy was called to constrain the initial state of the population. In the iterative process, an adaptive opposition-based learning (AOBL) was used to enlarge the search space. To obtain a reduction with fewer attributes, the dynamic redundancy detection (DRD) strategy was applied to remove redundant attributes in the reduction result. Furthermore, the quick extraction strategy was introduced to avoid multiple repeated computations in this paper. By combining an array with key-value pairs, the corresponding value can be obtained by simple comparison. The proposed algorithm and four representative algorithms were compared on nine UCI datasets. The results show that the proposed algorithm performs well in reduction ability, running time, and convergence speed. Meanwhile, the results confirm the superiority of the algorithm in solving MAR.

Fast and Secure Packet Scheduling with Packet Level Redundancy Elimination in Dynamic Mobile Networks

Mobile networks are fast and flexible for effective communication, where it only needs a dynamic topology and unstructured network construction. Due to this flexible nature, the resource utilization and energy consumption is high when comparing to the other static networks. In this scenario, packet redundancy is major problem on mobile networks, which increases network traffic and energy. Currently redundancy elimination is performed using redundancy elimination (RE) solution, which affects the end-to-end privacy. To eliminate redundant packet transfers and provide fast and secure packet transfer in the dynamic mobile network, a new middle box framework is proposed in this paper. This framework is named as Sift Hub, which is a centre point to verify the packet redundancy and security violations. This includes four algorithms to perform packet redundancy elimination, packet scheduling and verification. The algorithms are One pass signature generation algorithm (OSGA) for packet security, Predictive encryption technique using enhanced hidden vector encryption algorithm for redundancy detection on encrypted traffic, Packet level data filtering algorithm (PLDF) for eliminating un-authenticated and redundant packets and RAPS-Redundancy aware packet scheduling algorithm for fast data scheduling over dynamic mobile networks. The Shift_Hub supports inter and intra packet level redundancy elimination and secure packet scheduling without affecting the end-to-end privacy. The Shift_Hub is implemented and the performance is evaluated on dynamic mobile network scenario created on NS2 simulator. The results shows, the proposed work gives better performance in finding and eliminating redundant packet in secure manner than the existing works

A Predictive Acknowledgement Model for Enhancing Traffic Redundancy Elimination in Cloud

Due to the huge increase in the utilization of cloud storage in recent days, it leads to a massive growth in data traffic from application based servers to applications like smart phones, which not only influences batteries and computational capacities but as well swamp down multi-hopping strategies during data transmission. To resolve this crisis, traffic redundancy elimination (TRE) is an effectual solution, where the chunks to be transmitted will be directly fetched out from the receivers’ cache. Moreover, prevailing solutions cannot be directly applied or it is not appropriate for smart phones owing to its energy overhead ad high computation that is imposed on the applications. In order to overcome this problem, in this investigation, a novel a Predictive Acknowledgement for Eliminating Traffic (PACKET) is proposed which comprises of three significant elements. Initially, every application possess a clone in cloud that are responsible for calculating intensive tasks like detecting redundancy and parsing traffic. Secondly, consider that every cloud user has some specific applications like Facebook to be used in regular day to day life, every clones of cloud has to selectively determine the applications that are most frequently utilized and also reduce the high redundancy ratio. Thirdly, some cloud users always possess certain common applications; the proposed PACKET clusters those clones to co-operatively perform redundancy detection so as to diminish cache resource consumption in cloud. The simulation is carried out in MATLAB environment; the traces of applications are collected from online available data and are utilized for simulation purpose. Experimental outcomes demonstrates that PACKET can attain much higher hit ratio, reduced E2E delay, increased E2E throughput, energy efficiency and effectual bandwidth utilization in contrast to existing approaches. The proposed PACKET shows better and efficient trade-off than prevailing techniques.