Cooperative Caching (COOP Cache)

Cooperative Caching

Strategies in Mobile Ad Hoc Networks Using Remote Client Memory to Improve System Performance.

                          Muhammad Arqam

                                (22 FEB, 2021)

ABSTRACT:

           With the emerging demand & need for data fetching with the speed as high as possible, the cooperative cache concept came into existence. It gives the facility to share caches of other users. In this way, users can fetch data from other user’s cache despite taking data from the hard disk that usually takes a lot of time. In this paper, we will discuss the strategies of data fetching the COOP (CO OP) cache. How the Operating System manages it to enhance the data transfer performance.

This paper is showing the dual data cache management architecture that is known as a cooperative cache. This has been designed for the low power cache for the embedded system to increase their efficiency. There are two types of COOP cache, i.e. direct map temporal oriented cache (TOC) and a 4 way set-associative spatial oriented cache (SOC).

Keywords: low power cache, cooperative cache, block size, associativity, and an embedded processor.

INTRODUCTION:

           The cooperative cache is mainly developed for mobile (non-static) ad-hoc networks commonly known as MANET (Keywords-Mobile Ad-hoc networks). It seeks to improve the performance of the file network system by allowing the client to get data from another’s client cache when the operating system doesn’t find the data in the local cache. When OS fails to get data from any cache attached to the network, it will then fetch data from the disk. All caches attached with the network create a shared coordinated system of caches in which each cache coordinates with another cache for the requested data. COOP cache is applicable for both wired (LAN) and wireless networks. The cooperative cache system improves the performance of the cache system by virtue of the structural characteristics of the two caches, which are inherently designed to help each other.

There are a lot of technologies that tend to cooperative cache but the following two are the main technologies.

1.    Processor working performance is increasing much more than the disk performance. This urges to reduce the number of disk hits to improve the performance.

2. High-speed networks are emerging day by day. This tends to use the cooperative cache.

This cache is very useful under a heavy workload and ad-hoc networks. The cooperative cache is helpful in resolving the issue of delaying retrieving the data from the server. It addresses two basic caching issues in ad-hoc networks (MANETs):

1.    Cache Resolution: COOP cache discovers the data source which induces less communication cost by using forwarding nodes and historical data.

2.    Cache Management: COOP cache reduces duplication of the data among nearby nodes and tries to keep distinct data as much as possible. This improves overall performance.

The power consumption of on-chip caches becomes more important especially in embedded processors because on-chip caches usually consume a large portion of power consumption in the processors. This paper presents a cooperative cache system that consists of two caches with different configurations as a power-efficient cache structure.

The dual data cache presented by Gonzalez et al., and the split temporal/spatial cache system proposed by Milutinovic et al., are constructed with two separate caches, i.e., spatial and temporal caches.

PARTS:

The cooperative cache system consists of two caches.

1.     Direct-mapped temporal oriented cache (TOC).

2.   Four-way set-associative spatial oriented cache (SOC)

COOPERATIVE CACHING RESEARCH WORK:

This is a challenging task to manage and overcome the caching issues or to enhance them because of the challenging change in technology in the current era. There are different cache techniques like cache invalidation, cache consistency, and cache discovery, etc.

Cooperative cache has enhanced the performance of the systems to a great extent. This has been proved in the latest study. Issues related to the implementation of COOP cache in the networks still under consideration for the areas like cache locality ache cost, cache memory size, data consistency, identification and placement of the area to be placed, replacement, and writing policy that what and when it can store and for how long.

COOPERATIVE CACHING FUTURE:

The amount of research work under process shows that the COOP cache is gaining a very important role in all fields of networks like wireless sensor networks, wired networks, social wireless networks, wireless Ad-hoc networks, social multimedia wireless networks, and also in microchip processing to increase the performance of overall network data sharing with low cost and high speed.

The availability of the data where the client needs in a very short time in the network with a load is another important issue always thinking of adding clients.

ALGORITHMS:

           There are 4 different COOP algorithms. All these algorithms will be used when the operating system will not find the data in the local cache of the client. The algorithm will not affect the data storage reliability because the client will just read the data using cache. If the client will do a write operation, that operation will be applied on both caches as well as main storage. So, in this way, caches usually reduce the throughput time of reading. Following are the four types of cooperative cache:

1.    Direct Client Cooperation: This is the simplest caching approach in which an active client can use the cache of an inactive or an idle client freely. This means an active client can access the cache of a remote idle client to improve data accessibility. It is called direct cooperation because a client is directly targeted towards the idle client cache in case of cache overflow of that client.

2.    Greedy Forwarding Cooperation: Another simple COOP cache algorithm in which cache memory of all the clients in the respective system will be treated as a global resource that may be accessed to fulfill the need of any client. When the client requests specific data, the server finds that data to its local client first. If the server couldn’t find that specific block of data, it then starts searching the data in another’s client cache. It will start searching data from nearby clients and then give data from the main storage if no cache contains the requested data.

It is called greedy forwarding because each client forwards the request when data is not available.

3.    Centrally Coordinated Cooperation: It adds coordination to the greedy forwarding algorithm by portioning each client’s cache into the local section to which that client manages greedily. There is also a global section coordinated by the server as an extension of its central cache.

4.    N-Chance Forwarding Cooperation: This is the last and final algorithm of the COOP cache. It dynamically adjusts the fraction of each client’s cache managed cooperatively.

OTHER ALGORITHMS:

1.    Hash-Distributed Caching

2.    Weighted LRU

Cooperative Caching Applications:

1.    Mobile Environment:

         Giving free access to the locally connected devices to share the server load and access the data and ignore the application cost of downloading and sharing such mechanism is also known as COOP cache in wireless networks carrying the mobile in workplaces like in public parks, university campus, markets, malls, airport terminals, and other public spots get connected to Social Wireless Networks would be automatically formed by using the devices by ad hoc wireless connections. The following devices will be included in it: Google’s, Apple’s iPhone, Androids, AWS, and other electronic vendors, etc.

SHARED CHANNEL DIAGRAM OF COOPERATIVE CACHE:

2.      WIRELESS P2P NETWORK:

         The purpose of it to reduce the overhead of copying data between the user space and kernel space and also to allow the data pipeline to reduce the end-to-end delay.

         After applying on different MAC layers such as 802.11 based ad hoc network and multi interference and multi-channel based mesh network on the performance of the cooperative cache. Results show that these approaches outperform the symmetric approach are traditional 802.11 based ad hoc networks by removing most of the processing overhead. Data Access delay is reduced in the asymmetric approach compared to the symmetric approach in mesh network because of the data pipeline.

CONCLUSIONS & SUMMARY:

           Cooperative cache plays a very important role in enhancing the data fetching/reading speed. This paper evaluates the performance of the COOP cache. The main purpose of this cache is to maximize the cache hit ratio or to minimize the cache miss ratio or minimize the hits on the main storage of data.

Over the past few decades, the use of applications to enable business processes has evolved significantly. What was just once a nice-to-have is currently an ordinary fastener that exists at the core of commerce, education, and other operations across the world. As application usage has greater than before, the reply time has turn into more and more significant with good explanation. So waiting is considered time-wasting and money-wasting with production lost. 

 Affecting factors for reply time today include overcrowded WAN pipes, ineffective/informal protocols, and latency due to extended distances between data and user. All these factors are exacerbated as applications and data resources are consolidated, centralized, or outsourced. So in a lot of cases delays translate into irritated users.

Following are the few main points regarding the cooperative cache:

·      Cooperative Caching (CC) has a main role in improving the performance in retrieving of data in mobile networks for mobile clients.

·      Cooperative Caching can save communication cost, bandwidth, and energy.

·      To decrease the memory size of caching data the compression technique can also be used.

·      Instead of using remote memory data can be fetched and accessed from the local memory domain.

·      Can be used as a new technique for eliminating redundancy that will reduce the space requirements for inter-network storage. This can be performed on the entire system at multiple levels.

·      Instead of going to remote memory for a cache miss, the local cache can be checked.

ACKNOWLEDGEMENTS:

           We are very thankful to Professor “Dr. Aslam” for leading us in this review paper. We would like to work with him in the future as well because he has given a lot of information to us regarding our topic. We have gained a lot of info regarding cooperate cache.

REFERENCES/CITES:

·       Smart collaborative caching for information-centric IoT in fog computing. Sensors (2017)

https://www.mdpi.com/1424-8220/17/11/2512

·       Ma, T., Hao, Y., and Shen, W., Tian, Y., Al-Rodhaan, and M.: An improved web cache replacement algorithm based on weighting and cost”. IEEE Transl. Content Min. Permitted Acad. Res. 6(16), 27010–27017 (2018).

https://ieeexplore.ieee.org/abstract/document/8344541

·      Matthew M. Mutka and Miron Livny. The Available Capacity of a Privately Owned Workstation Environment. Performance Evaluation, 12(4):269–84, July 1991.

·      Erik Hagersten, Anders Landin, and Seif Haridi. DDM–A Cache-Only Memory Architecture. IEEE Computer, 25(9):45–54, September 1992.

·       Saleh, A.I.: “An adaptive cooperative caching strategy (ACCS) for mobile ad hoc networks. Knowledge-Based Syst. 120, 133–172 (2017)

https://www.sciencedirect.com/science/article/abs/pii/S0950705117300060

·       Umamaheswari, S., Radhamani, G.: Enhanced antsec framework with cluster based cooperative caching in mobile ad hoc networks. IEEE Int. J. Commun. Netw 17(1), 40–46 (2015)

https://ieeexplore.ieee.org/abstract/document/7059533

·       González-Ca?ete, F.J., Casilari, E.: A study of the performance of cooperative caching in static ad hoc networks. Int. J. Adv. Netw. Serv. 6(1), 68–79 (2013)

https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.360.7067&rep=rep1&type=pdf#page=78

·       Theresa-Joy, P., Polouse-Jacob, K.: Cache Replacement Strategies for Mobile Data Caching. Int. J. Ad Hoc Sensor Ubiquitous Comput. (2012).

·       https://www.researchgate.net/publication/258493943_A_Review_of_Cooperative_Caching_Strategies_in_Mobile_Ad_Hoc_Networks

·       https://pages.cs.wisc.edu/~remzi/Classes/739/Fall2017/Papers/coop-caching-94.pdf

·       Yin, L., Cao, G.: Supporting Cooperative Caching in Ad Hoc Networks. In: 2004 IEEE Infocomm (2004) - Google Scholar - https://ieeexplore.ieee.org/abstract/document/4445675 (2017).

·       Artail, H., Safa, H., Mershad, K., Abou-Atme, Z., Sulieman, N.: COACS: A Cooperative and Adaptive Caching System for MANETs. IEEE Transactions on Mobile Computing 7(8) (August 2008a) - Google Scholar - https://ieeexplore.ieee.org/abstract/document/4445675

·       "Cooperative Caching in Ad Hoc Networks." 4th International Conference on Mobile Data Management. Melbourne, Australia. MDM 2003, 10 Sep 2009. Pp.13-28.

·       Beal, Vangie. WeboPedia. Property of Quintet Enterprise. 2017. https://www.webopedia.eom/TERM/C/cache.html (accessed March 10 March, 2017).

·       Jichuan Chang, G.S.Sohi. "Co-Operative Caching for Chip Multiprocessor." Computer Architecture ,2006,ISCA 06,33rd International Symosium On . IEEE xplore, Feburary 2006. 38.

·        Blue Coat Systems Inc. (2013). BLUE COAT,A TECHNICAL REVIEW OF. Retrieved March Tuesday, 2017, from https://www.bluecoat.com: https://www.bluecoat.com/documents/download/9457735e-8603-40a1-a566- 11d3f62aae8a/c645fb5d-746c-4be3-a40c-ae1cbce54141

·       M.S. Smith, L. S. (2011, May 05). Bright hub. Retrieved 03 11, 2017, from https://www.brighthub. com: https://www.brighthub.com/computing/hardware/articles/32165.aspx

• M. Dahlin, R. Wang, T. E. Anderson, and D. A. Patterson. Cooperative caching: Using remote client memory to improve file system performance. OSDI94, 1994
• Michael D. Dahlin, Clifford J. Mather, Randolph Y. Wang, Thomas E. Anderson, David A.

Patterson. Quantitative Analysis of Cache Policies fir Scalable Network File Systems.

Proc. 1994 ACM SIGMETRICS Conference on Measurement and Modeling of Computer Systems. May 1994.