Development of Hyperset Paging mechanism CURRENT STATUS: POSTED

Algorithms for page replacement play important roles in virtual memory management, especially in paging operating systems. Page substitution happens when the required page is not retained in the memory (file fault) or the free accessible file is not adequate to fulfill the requirement. It is either there are none or the amount of free sites is fewer than the required total. Two regular page replacements were hybridized as Hybrid Page replacements (HRA) in this analysis (LRU and OptimalOPT algorithms recently used). For its service, HRA is based on the principle of OPT and LRU. The HRA was determined by comparing the page failures caused with the default algorithms First InFirst Out FIFO, LRU and Optimal. Tests showed the amount of frames through to 4 and beyond the HRA outperformed FIFO, OPT and LRU.


I. Introduction
In the operating system, the paging replacement mechanism is crucial. Page replacement algorithms allow the decision from which pages the memory will be replaced if the required frame is not accessible. But to solve this problem, some algorithm was created. The algorithms differ from each other in the way in which the page removal process is been handled (Genta, 2015). All page replacement algorithms are internally similar to the following: insertion, detection and searching for a page. However, most of these algorithms depends on specialized data structure for effective performance (Kavaraid, 2013). Most of the algorithms handle huge amount of data, the locality of reference is considered as a shared attribute between programs. Hence, the majority of applications does not access all data at the same time. Instead of this, they reference only small part of data at different point in time (Shen et al, 2004). The locality of reference can be adopted in two different approaches: spatial locality of reference and temporal locality of references. The spatial locality is of the view that nearby memory location will be referenced in the near future while temporal locality depends on the time the page spends in the frame (O'neilet al., 1993).
The localities of reference can be increased by careful selection of data structure used in the algorithm. The data structure will reduce the page fault rate as well as the number of pages in 3 working set. For example, a stack has high locality because the replacement algorithm of the stack algorithm maintains a single page list in the LRU order similar to the original LRU algorithm. However, any dirty pages in the list will have a cold flag which may or may not be set as per the situational operation. The identification of Cold or Hot of a dirty page is taken care of by a cold detection algorithm. A dirty page may be cold, if that is the least referenced page in the list and its cold flag is set. A dirty page may be hot if that page is referenced again and the cold flag on it is cleared. Results from this work was compared with FIFO and found 18% better off than FIFO.Mark Page Discard Randomly Caching Algorithm (MPDRC) was proposed by Sumitet al., (2015), the work was developed in windows XP using C# .Net and the unique identification number was assigned to unique URL's to log of proxy server. These numbers were taken as a reference string that suits input to the innovative algorithms. The focus of the research was exploration of web proxy caching algorithm best for proxy server. With the help of log in details of proxy server, a real trace web reference was obtained. For the simulation, each URL was given a numeric identity which enables the numeric reference string to be acquired. During simulation the author compared MPDRC with FIFO, LRU and LFU replacement algorithms with the proposed innovative page replacement algorithm (MPDRC). It was discovered that MPDRC algorithm has 8.34% better than FIFO, 16.63% better than LFU and 10.6% better than LRU algorithm. Overall average this work innovative page replacement examined 11.85% better than existing algorithm. After comprehensive simulation experiments is summarized that for proxy caching the MPDRC hit ratio performance better than others existing algorithms.A study by Paajanen (2000), worked on Least Recently Used (LRU). The algorithm takes into considered the imbalance of the cost of read and writes memory when replacing pages. The algorithm defines that, the string been used in the list was replaced in the needed frame. The LRU algorithm has better performance when compared with FIFO.

Iii. Method
In this study, three standard algorithms were considered; FIFO, LRU and OPT i. ii.
Optimal: In operating system the algorithm swaps out the page whose next reference will take longer time for future reference. For example, a page that is not going to be used for the next 6 seconds will be swapped out over a page that is going to be used within the next 0.4 seconds(Khulbe, et al., 2014). iii.

Least Recently Use (LRU):
This based on the probability theories in a logic idea.
This idea consists of pages which have been heavily used in the last few instructions, which will probably be used again in the next few pages that are not use for a long time may say in this state again.
The following are some of the lapses observed with the existing algorithms i. The optimal algorithm fetches the string that will be used in the future, whereas the operating system cannot predict the future occurrence. Therefore, the optimal is not a good choice ii. FIFO algorithms fetch and replace the string that came into the frame first. This operation is based on first in first out (queue). As a result of this operation, more page fault is generated. Hence, FIFO is not a good choice.
iii. LRU, this algorithm is similar to FIFO just that it fetches in the page that has been used the least, this is better than FIFO but still have high page fault than OPTIMAL.

Hybrid Page Replacement Modeling
The HRA comprises of two standard page replacement algorithms namely, LRU and Optimal. These where hybridized to give a better performance in terms of number of page faults generated when compared to the individual paging algorithm. As shown in fig.1, the proposed HRA algorithm is designed for the purpose of less page fault in the page replacement algorithm.Knowledge of LRU and OPT algorithm is crucial in the usage of HRA algorithm. The victim page is been swapped in/out of the physical to the backing store (virtual memory) using HRA technique. The frame is shared between OPT and LRU i.e. the first frame is given to OPT, and it takes the first reference string, processed it 6 with the frame and swap out the particular frame that will not be used for the longest period of time and LRU will pick the second reference string, process it with the frames and swap out the particular frame that has been used least recently among all the frames on the pages. Java SE was used to implement the algorithm and the units that were tested includes: the first dialog box for the input frame, the second for the length of the strings and the third for the reference strings itself.

Iv. Results And Discussion
In this study, the results achieved are the outcome of simulation of the designed system. This result is the evaluation performance metric of the proposed HRA system in terms of NUNBER OF PAGE FAULTS generated. Fig. 2 fig. 8 shows thatHRA have poor performance at the lower number of frame, but from 4 frames and above, it outperformed others in terms of lower number of page fault generated.

V. Conclusion And Future Works
Developed and tested an efficient page substitution algorithm called HRA, which used LRU and OPTIMAL algorithm concepts. Via multiple experiments, HRA models have been found to be equivalent to the current Operating System page substitution by default link. Extensive work will be performed to Graph of HRA with 20 reference strings.