Caching Methodologies in ICN - Paper Review

Caching Methodologies in Content Centric Networking (CCN)

            Internet is an invention in the world of Technology which has change the world into a global village. Where we can communicate with a person who is very far away from us. We can share a lot of information as well as can access the information from the others’ computer by using Internet. We become addicted of Internet in such a way that we even didn’t accept a little communication gap or delay. Current IP-based Internet is facilitating us but it still have some problems that are not acceptable by users. One of them is that mobility. Whenever communication is started and user move from one location to another location, its network changed and it cause to change the IP Address which effect the communication in such a way that downloading may start from the beginning or may continue after the renewal of connection etc.

            On the other hand, there is a huge traffic on the server. Nowadays, people are using Social Sites and Streaming Sites excessively and billions of requests goes to the server which become very difficult to manage for the server.

            An upcoming Internet model known as Information Centric Network (ICN) is introduced. It will replace our current Internet architecture. In current architecture focus is on establishing a connection between two devices. In current internet architecture, a naming system in which URLs are resolved through DNS to indicate the device to communicate with to receive data. [1]

            CCN provides mobility and enhance the transfer rate by reducing the load on the server using a concept of Content Caching. In which content is cached on different routers from the server first time and served from cached location for the next time.

            Author has discussed the concept of content caching and different techniques to cache the content for better utilization of the network as well as cache storage (router).

            Caching the content is the main feature of ICN which enables nodes on the request path to store a copy of the actual content from the content server temporarily. It will increase the availability of content closer to the requester (user) by reducing the user’s request time and also optimize the transfer rate.

            There are two main component in the caching node (router) that are known as Index Table and Packet Store.

Index Table:   Index Table is used to store the location (address) of packets stored in the packet store. It helps to access and search content stored in the packet store. Index Table is stored inside the Static Random Access Memory (SRAM). Index Table has smaller than Packet Store in size.

Packet Store: Packet Store is used to store the actual content data. All the data (packets) are allocated a particular location (cell) inside the Packet Store to be stored. These cells are assigned unique ID (address) that are stored in the Index Table.  [2, 3]

            Author has assumed different caching methodologies. Each one has its own focus point and advantages. All methodologies are proposed for the best usage of cache storage in different way and by different factors.

i.  Mixed integer program (MIP): The is a powerful model which is used to enhance the assignment of content to the router depending on the content’s popularity, cache size and capability of network link. As sharing the un-redundant content can get more access rate.  To make the response time shorter, distance of the content from the user / client is taken into the consideration.[4],

ii. Novel Cache Aware Approach: This approach is used in intra-domain cache aware routing technique which choose a particular path to rout and cache the content based on the lowest transportation cost and capability of network caching. In this approach, dynamic programming approach is used to measure the lowest transportation cost path by keeping the content or cache location and popularity and priority of the content in the consideration. [5],  

iii. Partition Based Caching: Also known as Segment Based Caching. In this caching technique, the content’s information is divided into portions (partitions) and all the segments are stored in different caches on the network. In this cache technique, data can be transmitted parallelly for better utilization of network and maximize the throughput of network. This parallel transmission of data minimize the average content transfer time. [6]

iv. Age Based Cooperative Caching: The purpose of this technique is to minimize network delay and traffic on the server by enhancing the whole network cache hit ratio. This technique also works on collaboration mechanism with the help of which the most requested content in over the network is spread by utilizing the storage space in Packet Store. Each content in the cache of router have an age and on the base of the age, that content is removed from the cache of the router. Each cached content is assign an age. Age of content is decided on its need or interest of users. If a content is more popular and requested a lot, its age will be greater and this content will get higher priority in the cache. [7]

v. CATT Approach:  The main focus in this approach is on routing and caching. Two different techniques are used for Routing and Caching. Potential Based Routing (PBR) is kept under focus for Routing and CATT is used for selective caching. Whenever a request is generated for a particular content by the user, it can be cached in all cache storages on different locations first time. Whenever a particular router receives a request for that content it serves if it have that content in its cache store or forward that request to the neighbor router if content is not available on its cache store based on PBR. Different techniques can be used to find the potential of a particular neighbor router. [8]

vi. Probabilistic in-network Caching: This scheme is very useful in the network which have decentralized in nature where real time content is distributed. This technique is used for well managed caching. As the cache redundancy is an issue where everything is cached everywhere. So the major objective is to reduce the redundancy of the content on the cache locations on the path. It also reduces the numbers of hops to enhance the speed by minimizing the response time. [9]

vii. Ubiquitous in-network Caching: Ambiguity in a network is that content requested by a user is available on more than one locations and this is a confusion that where we should get the content from. The purpose of this methodology is to focus on centralize network where there is no redundancy of the content in the network to remove the ambiguity and to ensure the centrality of the content. [10]

viii. Congestion Aware Caching: The network performance in ICN can be measured by the metrics that is the number of hopes between client and cache location as well as by the hit ratio. Time difference between the request generation time for a content from content server and the time taken for downloading the content from the content server is very important. The network’s performance can be calculated through the capacity / bandwidth of network link and network congestion. In short, the main purpose of this technique is how we can utilize better congested network path by cache the content that is requested instead of focusing on hit ration. [11]

ix. Cache Capacity: Flash crowd a very big problem in the network. Flash crowd is an event in the network is a problem where a large number of requests received on the server for a particular content (news, blog, video etc.). This cause that server temporarily down due to capacity exhausting and bad effects on the capabilities of content server. The purpose of this technique is to detect the Flash Crowd in the network and give a mechanism to handle this kind of problem ensure the availability of the content. The content request is redirected to another node and decision for the node is taken on the based on the previous forwarded request for that content. [12]

            CCN / ICN is expected to be the most advance internet architecture which can be deploy in the future. As the purpose of CCN is to convert the current Internet model which is based on IP into name based content centric model to provide a better secure communication by enhancing the scalability and mobility.  Scientists and researchers are still working on the CCN and methodologies to overcome the problems in CCN to make it better and better.

References

1.      5g whitepaper: The flat distributed cloud (fdc) 5g architecture revolution

2.      Somaya Arianfar, Pekka Nikander, Jörg Ott, Packet-level caching for information-centric networking, in: ACM SIGCOMM, ReArch Workshop, 2010.

3.      Mehmet Sayal, Yuri Breitbart, Peter Scheuermann, Radek Vingralek, Selection algorithms for replicated web servers, ACM SIGMETRICS Perform. Eval. Rev. 26 (3) (1998) 44–50.

4.      Yusung Kim, Ikjun Yeom, Performance analysis of in-network caching for content-centric networking, Comput. Netw. 57 (13) (2013) 2465–2482.

5.      Vasilis Sourlas, Paris Flegkas, Leandros Tassiulas, A novel cache aware routing scheme for information-centric networks, Comput. Netw. 59 (2014) 44–61.

6.      Moinuddin K. Qureshi, Yale N. Patt, Utility-based cache partitioning: A lowoverhead, high-performance, runtime mechanism to partition shared caches, in: Microarchitecture, 2006. MICRO-39. 39th Annual IEEE/ACM International Symposium on, IEEE, 2006, pp. 423–432.

7.      Zhongxing Ming, Mingwei Xu, Dan Wang, Age-based cooperative caching in information-centric networking, in: Computer Communication and Networks (ICCCN), 2014 23rd International Conference on, IEEE, 2014, pp. 1–8.

8.      Suyong Eum, Kiyohide Nakauchi, Yozo Shoji, Nozomu Nishinaga, Masayuki Murata, Catt: Cache aware target identification for icn, IEEE Commun. Mag. 50 (12) (2012).

9.      Ioannis Psaras, WeiKoong Chai, George Pavlou, Probabilistic in-network caching for information-centric networks, in: Proceedings of the second edition of the ICN workshop on Information-centric networking, ACM, 2012, pp. 55–60.

10.  Wei Koong Chai, Diliang He, Ioannis Psaras, George Pavlou, Cache less for more in information-centric networks, in: International Conference on Research in Networking, Springer, 2012, pp. 27–40.

11.   Mikhail Badov, Anand Seetharam, Jim Kurose, Victor Firoiu, Soumendra Nanda, Congestion-aware caching and search in information-centric networks, in: Proceedings of the 1st ACM Conference on Information-Centric Networking, ACM, 2014, pp. 37–46.

12.  Dabin Kim, Sung-Won Lee, Young-Bae Ko, Jae-Hoon Kim, Cache capacityaware content centric networking under flash crowds, J. Netw. Comput. Appl. 50 (2015) 101–113.