A Cache Content Replacement Scheme for Information Centric ... - Core

Available online at www.sciencedirect.com

ScienceDirect Procedia Computer Science 89 (2016) 73 – 81

Twelfth International Multi-Conference on Information Processing-2016 (IMCIP-2016)

A Cache Content Replacement Scheme for Information Centric Network Kumari Nidhi Lal∗ and Anoj Kumar Motilal Nehru National Institute of Technology, Allahabad 211 004, India

Abstract Nowadays, Information centric networking has tremendous importance in accessing internet based applications. The increasing rate of Internet traffic has encouraged to adapt content centric architectures to better serve content provider needs and user demand of internet based applications. These architectures have built-in network Caches and these properties improves efficiency of content delivery with high speed and great efficiency. By using Information centric architectures, users need not download content from content server despite they can easily access data from nearby caches. User requested content is independent of the location of the data by use of caching approaches and does not rely on storage and transmission methodologies of that content. There has been many researches going on which is based on caching approaches in the Content centric network. Efficient caching is essential to reduce delay and to enhance performance of the network. Along with caching, a good cache replacement scheme is also necessary to decide which content item should reside in the cache and which content should be evicted. So, in this paper, we presented a Cache content replacement (CCR) scheme for information centric network with evaluation of popularity of the content. Performance of CCR scheme will be evaluated in provision of cache hit ratio and Cache load. © 2016 The Authors. Published by Elsevier B.V. © 2016 The Authors. by Elsevier B.V. This iscommittee an open access under the CC BY-NC-ND license Peer-review under Published responsibility of organizing of article the Twelfth International Multi-Conference on Information (http://creativecommons.org/licenses/by-nc-nd/4.0/). Processing-2016 (IMCIP-2016). Peer-review under responsibility of organizing committee of the Organizing Committee of IMCIP-2016

Keywords:

CN; QoS; CCN; PIT; FIB; CS; LRU.

1. Introduction About 100 years ago,1 a mathematical foundation was created by Erlang which was used to analyze communication network based on telecommunication. Again 50 years later since that time,2 a concentric information network comes into account in the area of network research field which was also called as Information centric network (ICN).27 Information centric network basically a model who focuses more on the content item rather the source of content item where the content is residing. Centric network means, information requested by the client can easily accessed from one of the intermediate routers rather than the end server. This network exhibit data centric communication instead of end to end communication between computers.28 The Routers in the ICN network carry two functionalities: first caching content data and the second is routing of that content data. ICN supports in cooperation with multicasting, as well as broadcasting. Today, ICN concept can applied and tested easily on26 Software defined network (SDN),25 Wireless sensor network (WSN), Network Clouds, Wireless mesh network (WMN), Ad hoc Network etc. Now a days, our world is completely technology dependent. We use the internet for our lot of requirements like education, ∗ Corresponding author. Tel.: +91 -8173976148.

E-mail address: [email protected]

1877-0509 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of organizing committee of the Organizing Committee of IMCIP-2016 doi:10.1016/j.procs.2016.06.011

74

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81

Fig. 1. Host-to-Host Network Architecture Design.

Fig. 2. A Caching Router in ICN.

entertainment, knowledge and all other general purpose needs. These internet access based applications includes Gmail, YouTube, Facebook, Google and other web applications.3 In this scenario of the real world, it is reported that global internet traffic is increasing rapidly day by day or hour by hour and 70% above people are using internet toady in most of countries in the world. Many social and media applications like HD quality videos, software, Twitter, Facebook are used by user in large growth rate. So to cope up with these large amounts of requirements, Information Centric network has been introduced with many advantages like it does not depend on hop to hop communication. It exhibits the characteristic of communication which lies in receiver based data access methodology.4, 5 The expected benefits of the proposed content centric network over traditional end to end internet network are high efficiency, better scalability with respect to information/bandwidth demand, support of high degree of mobility, high performance in terms of content delivery guarantee and better robustness/reliability in challenging Communication environment. Many innovative architectures are proposed in the field in terms of DONA6 , CCN7 , NetInf8, PURSUIT9 and PRISP10 . There are so many options for making caching in ICN like chunk level caching, packet level caching and object information level caching. Naming of Contents is done by location independent identifying by using two key approaches: Two phase approach and one phase approach. In one phase approach routing of content is done natively in the network, whereas in two phase approach, first mapping is done to the content id locators.13, 22 In node centric design, Sometimes called host to host communication network, same content data are transmitted repeatedly between the nodes and that costs wastage of bandwidth, high traffic with excessive delay. ICN comes into account to overcome these drawbacks of host to host communication network. Figure 1 shows the architecture of Host-to-Host network design. In Fig. 2 showing below, the content data requested by the client is cached by one of intermediate router in the path from client to server. So content item is instantaneously directed by the caching router to client. Each caching router in information centric network maintains three data structure named as: Content store (CS), Pending information table (PIT) and Forwarding information base (FIB). There are two types of information object packets are used for establishing the communication: interest packet, which is created by client for getting services or content data and another one is a data packet.11 Content store uses to cache content and servers as a local cache. The task of FIB in caching router is mapping of information provided by client for requesting data to the output interfaces. PIT basically keeps records and tracks of incoming interfaces from the requesting interest message were generated/pending. When interest message from requesting client arrives at the caching router as shown in Fig. 2, then caching router checks the extracted information about requests content in its local cache called content store.12 If the operation is hit i.e. content is matching then it sends that content to the requesting client by using incoming interfaces otherwise caching router further matches the longest prefix of information object in its FIB table to select appropriate path and data source

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81

Fig. 3. Three Data Structures of Router in ICN.

for forward interest message further. In this case, if a forwarding entry is found in FIB then caching router record incoming interface of interest message in its PIT table. Following Fig. 3 shows the whole process of a router in ICN. Here, orange color denotes interest packet and green color bar denotes data packet. 2. Literature Survey Cache memory is the fastest memory as compared to primary and secondary memory. It is used to store frequently referred pages to increase the throughput of the system and with minimum delay.20 In ICN, cache placement and cache replacement are different terms. Cache placement basically references ‘in what place the content should be placed?’ but Cache replacement defines how the event cache content eviction should take place to achieve a high hit ratio and minimum latency. There are many researchers have been going on strategies of cache replacement. Some of them are illustrated below: 2.1 Random replacement strategy In random replacement strategy when a content data are requested by requester node (client) and caching router find that content in its content store (CS) then that event is called hit event and the content data is immediately sent to the client by the caching router. When the content is not found in the content store (CS) of the caching router then the respective data request is sent to the server and in returning of that data from server, caching router selects one of the content in its CS randomly and replace that content with requests incoming content from the server. The selection criteria of content which has to be replace is done random. This strategy chooses any one content item in its content store for making replacement.14 In the analysis of this random replacement strategy, it can be seen that Requirement for memory is less in this strategy as well as replacement time is also very less because there is no criteria for selection. 2.2 Least recently used strategy Least recently used (LRU) is one of famous and mostly used cache replacement strategy in ICN.18, 19 In Least recently used strategy when a content data are requested by requester node (client) and caching router finds that content in its content store (CS) then that event is called hit event and the content data is immediately sent to the client by the caching router.17, 23 When the content is not found in the content store (CS) of the caching router then the respective data request is sent to the server and in returning of that data from server, caching router selects one of the content in its CS on the behalf of recency of usage. Most popular content items will be demanded more in the network so its usage will be more and recency is directly proportional to the usage. So the item which having less recency will be

75

76

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81

selected for replacement by a caching router in its CS.15 LRU replacement strategy gives a high hit ratio because the most popular content is accessed many times in the modern world scenario. For example, in release of the latest action movie trailer, the percentage of access and demand of that movie trailer will be very high and if this content is stored by caching router then it will lead to high hit ratio with as compare to random strategy. 2.3 First in first out strategy First in first out replacement strategy is very simple to understand and implements.21, 24 In this replacement strategy, strategy when a content data is requested by requester node (client) and caching router find that content in its content store (CS) then that event is called hit event and the content data is immediately sent to the client by the caching router. When the content is not found in the content store (CS) of the caching router then the respective data request is sent to the server and in returning of that data from server, caching router selects one of the content in its CS on the behalf of oldness of usage. In this scenario, oldness is directly proportional to the time at which the content data was stored in the cache storage. The more old data have high probability to be replaced with new arrived cached content. 2.4 Least frequently used strategy Least frequently used Strategy works with the maintenance of a counter for each content data. This counter for each content item tracks how many number of times that particular content item is requested or referred.15 In Least frequently used (LFU) strategy, when a content data is requested by requester node (client) and caching router find that content in its content store (CS) then that event is called hit event and the content data is immediately sent to the client by the caching router. When the content is not found in the content store (CS) of the caching router then the respective data request is sent to the server and in returning of that data from server, caching router selects one of the content in its CS on the behalf of less value of counter. This less or low value of counter indicates less number of times that particular content item is referred. The caching router will select a content item in its cache who has a low value of the counter and replaces it with newly arrived content data. 2.5 Most recently used strategy As the name indicates, most recently used strategy is just opposite to least recently used replacement strategy.16 In Most recently used (MRU) strategy, when a content data are requested by requester node (client) and caching router finds that content in its content store (CS) then that event is called hit event and the content data is immediately sent to the client by the caching router. When the content is not found in the content store (CS) of the caching router then the respective data request is sent to the server and in returning of that data from server, caching router selects one of the content in its CS on the behalf of high recency of usage. Researches show that the MRU replacement holds good results for scenarios which having accessed old content data in spite of new one. 2.6 Most frequently used strategy As the name indicates, most recently used strategy is just opposite to least frequently used replacement strategy.16 In Most frequently used (MFU) strategy, when a content data are requested by requester node (client) and caching router finds that content in its content store (CS) then that event is called hit event and the content data is immediately sent to the client by the caching router. When the content is not found in the content store (CS) of the caching router then the respective data request is sent to the server and in returning of that data from server, caching router selects one of the content in its CS on the behalf of the high value of counter. This high value of counter indicates a large number of times that particular content item is referred. 3. Proposed Work By using Information centric architectures, users need not download content from content server despite they can easily access data from nearby caches. User requested content is independent of the location of that data by use of caching approaches and does not rely on storage and transmission methodologies of that content. There has been many researches going to base on caching approaches in the Content centric network. Efficient caching is essential to reduce delay and to enhance performance of the network. Along with caching, a good cache replacement scheme is

77

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81 Table 1. List of Symbols. Ci P(Ci) L P(Ci) Es P(Ci) L P A(Ci . . . n) U P(Ci) R n Ci P(Ci) L P(Ci) Es P(Ci) L P A(Ci . . . n) U P(Ci) R U P(C j)

ith content Popularity of ith content Local popularity of ith content Expected Popularity of ith content Local popularity of all the contents in CS Universal Popularity of ith Content in CS Rank of the content Total number of content in CS ith content Popularity of ith content Local popularity of ith content Expected Popularity of ith content Local popularity of all the contents in CS Universal Popularity of ith Content Rank of the content Universal Popularity of jth content newly arrived

also necessary to decide which content item should reside in the cache and which content should be evicted. So, in this paper, we presented a Cache content replacement (CCR) scheme for an information centric network with evaluation of the popularity of the content. Performance of CCR scheme will be evaluated in provision of cache hit ratio and cache load. Our work proceeds as follows: • In this paper, evaluation and calculation of content popularity in universal basis (overall in the network). • Estimation of local content popularity with respect to the overall demand of that item in locally and universally. • Selection of content to be replaced in the cache of caching router in relation with estimation of instantaneous hit rate. • Performance analysis of the proposed CCR replacement scheme in provision of cache hit ratio and cache load in comparison with replacement policies mentioned in the literature survey like LRU, MFU, RANDOM, FIFO. 3.1 Universal popularity estimation of content in ICN Characteristics of the internet content can be understood simply by Zipf law. This law is used to assign a rank to each content universally based on the frequency of access request of that content. The rank of each content is calculated using popularity distribution α which ranges 0.6 to 1.2. For analysis of YouTube related content the value of α lies greater than 2 (α > 2). As the frequency is inversely proportional to the rank of the content data item means the rank of content will be higher if the access rate (frequency) of that content will be higher. Below Table 1 represents the list of symbols used to propose CCR replacement scheme. Popularity of content item c is defined as with relation of rank R of that content item. P(Ci )α

1 R∞

1/R ∞ U P(Ci ) =  N ∞ n=1 1/n

(1) (2)

This value of P(Ci ) is the popularity of i th content item c where R denotes the rank of a content item. This universal popularity listing is maintained in the cluster head (CH) of the ICN. 3.2 Local popularity estimation of content in caching router In ICN, each caching router estimates local popularity of content which is selected for replacement in addition of newly arrived content. The requirement of estimation of local popularity is made due to the knowledge of rank listing of content data in its local cache of caching router. To make efficient replacement decision calculation of the local

78

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81

popularity of content is essential. A caching router estimates Expected popularity of each content item in its caching router by using equation 2 as follows:   1 1 L Es P(Ci ) = √ Exp − 2 U P(Ci ) (3) 2σ σ 2π By using equation 3, caching router estimates an expected popularity of content item Ci with respect to its universal popularity. Further, Estimation of Local popularity of all the contents denoted by the L P A(Ci . . . n) which are residing in a content store of caching router will be carried on by using equation 4. L P A(Ci . . . n) =

n 

(L Es P(C(i ), U P(C(i ))

i=1

  n 1  1 2 Exp − 2 (U P(Ci )) − L Es P(i )) = √ 2σ (σ 2π)n i=1

(4)

After the estimation of the local popularity of all the contents in the content store, caching router evaluates the local popularity L P(Ci ) of each content. This value of L P(Ci ) is calculated by using equation 3 and 4. L P(Ci ) =

Es P(Ci ) L P A(Ci . . . n)

(5)

3.3 Content cache replacement strategy In Information centric network, when a content data is requested by requester node (client) and caching router find that content in its content store (CS) then that event is called hit event and the content data is immediately sent to the client by the caching router. When the content is not found in the content store (CS) of the caching router then the respective data request is sent to the server and in returning of that data from server, caching router selects one of the content in its CS on the behalf of lower instantaneous hit rate (Hi ). Instantaneous hit rate keeps track averages of all the hit rates in previous history of content item. Hi = δi + σ Hi−1

(6)

In the above equation 6, Hi represents the instantaneous hit rate of content i . the value of δi is 1 in the event of hit event and 0 in the event of a cache miss. Hi−1 denotes the previous history of hit rate of content i and value of σi lies between 0 and 1 (0.5 in most of the cases). The higher value of instantaneous hit rate indicates higher recent hit rate. After estimating the value of instantaneous hit rate, the content which has a lower instantaneous hit rate will be selected. The local popularity of that selected content will be calculated by equation 5 as follows:

√1 Exp − 1 2 U P(Ci ) 2σ σ 2π

L P(Ci ) = 1 n √1 Exp − 2σ 2 i=1 (U P(Ci ) − Es P(i ))2 n (σ 2π)

The local popularity defines the rank of content which has a low instantaneous hit rate in the cache of caching router among all the content items stored in the content store. Next step is for estimation of universal popularity of content item U P(C j ) by using equation 1 and 2 which is newly arrived for storing in the content store with the replacement of selected data. After the estimation of universal popularity of newly arrived data and local popularity of content data in a content store of caching router, Decision of replacement will be done as follows: • If the difference between universal popularity of newly arrived content U P(C j ) and local popularity of content L P(Ci ) which is selecting for eviction in a content store is greater than 0 then it indicates that the popularity of content item which is newly arrived is high as compared to data which is selected for replacement in CS. So, according to high popularity to achieve high performance the selected data from content store will be replaced by newly arrived data.

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81

• If the difference between universal popularity of newly arrived content U P(C j ) and local popularity of content L P(Ci ) which is selecting for eviction is less than 0 then it indicates the popularity of content item which is newly arrived is low as compare to data which is selected for replacement in CS. So, according to high popularity to achieve high performance the selected data from content store will not be replaced by newly arrived data and that data will be directly forward towards the requesting client. if {U P(C j ) − L P(Ci ) > 0} = replacement will be done if {U P(C j ) − L P(Ci ) < 0} = replacement will not be done In our implemented Content cache replacement (CCR) Strategy, the replacement is done in the relation of content popularity and low instantaneous hit rate. By following this strategy, instantaneous hit rate maintains recent content data hits so that if a content has a lower instantaneous hit rate then it indicates that the content is not popular and due to that it have less hits. 4. Performance Analysis In this paper, we presented a Cache content replacement (CCR) scheme for an information centric network with the evaluation of popularity of the content. Performance of CCR scheme will be evaluated in provision of cache hit ratio and cache load. A. Cache hit ratio Cache hit ratio will be evaluated by achieving the number of hits for overall request of accessing content from the requested client. If the request for content item made by the client is found in the content store of a caching rather than this phenomenon is called cache hit. The increasing number of cache hit leads to high performance of the information centric network because of less delay and content will be reached for the client before expected time. Cache hit ratio is directly proportional to the number of times content item found in the content store of caching router. C(hi ) =

number of times content found in C S Total no. of request in that caching router

Our implemented CCR replacement strategy makes an efficient decision to making replacement of content based on instantaneous hit rate and local popularity. From above Fig. 4 shows that simulation results of comparison for cache hit ratio between proposed CCR strategy with Random, FIFO, LRU and MRU replacement schemes. It can be clearly seen that cache hit ratio of CCR scheme is high as compared to other schemes. A. Cache load The term cache, load is indicated that how much request is processed by a server or a caching router at any particular time. Load is the total number of requests for accessing the content data/ items from client to the server. For better performance of the system, it is necessary that the server or caching router should not be overloaded with incoming requests for content access. So by taking this point into consideration, Proposed CCR replacement scheme replaces the content in such a manner that the load on caching router will be very less and eventually it will give high performance as compared to other replacement strategy in terms of cache load. From above Fig. 5, it can be cleary shown that by using proposed CCR caching strategy the cache load of caching router of ICN will be low as compare to Random, FIFO, LRU and MRU strategies.

79

80

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81

Fig. 4. Cache Hit Ratio of CCR Replacement Scheme with other Schemes.

Fig. 5. Comparison of Cache Load with CCR Scheme with other Schemes.

5. Conclusions Information centric network becomes a tremendous research area nowadays. Several research challenges are addressed in this research are of ICN like caching, security, load balancing, naming, data retrieval, scalability and so much else. Our main focus on caching replacement strategy and by inspire of this topic in networking field, we studied several researches regarding caching in ICN. Caching is just storage of content data with aiming of speedily served future request. In this paper, we presented an overview of various caching replacement approaches in ICN with several features and regarding issues.This paper gives a simple idea about ‘what to cache’ problem with relation of popularity. Our proposed CCR scheme will give very high performance in terms of cache hit ratio and cach load with comparison of LRU, MRU, FIFO and Random strategy. Still, there are so many open issues in caching in current ICN architecture like maintaining the consistency of content data if anything updating occurs, Synchronization with various caching routers with respect to the server, reducing redundancy of content data item in various caches and optimization of cache space to achieve high capacity to store content data item and so much else have to consider to make an effective and efficient cache mechanism in information centric network. This will lead to bright future of ICN in current internet access scenarios. References [1] Kurose Jim, Information-Centric Networking: The Evolution from Circuits to Packets to Content, Computer Networks, vol. 66, pp. 112–120, (2014). [2] Chanda, Abhishek, Cedric Westphal and Dipankar Raychaudhuri, Content Based Traffic Engineering in Software Defined Information Centric Networks, IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), 2013 IEEE, (2013). [3] https://www.surrey.ac.uk/sites/default/files/5G-Network-Architecture-Whitepaper-(Jan-2016).pdf [4] Davies, Elwyn, et al. Information-Centric Networking: Baseline Scenarios, (2015). [5] You, Wei, Bertrand Mathieu and Gael Simon, Exploiting End-Users Caching Capacities to Improve Content-Centric Networking Delivery, P2P, Eighth International Conference on Parallel, Grid, Cloud and Internet Computing (3PGCIC), 2013 IEEE, (2013). [6] T. Koponen, et al., A Data-Oriented (and beyond) Network Architecture, ACM SIGCOMM Comput. Commun. Rev., vol. 37, no. 4, pp. 181–192, October (2007). [7] Content Centric Networking Project. [Online]. Available: http://www.ccnx.org/ [8] NSF Named Data Networking Project. [Online]. Available: http://www.named-data.net/ [9] FP7 PURSUIT project. [Online]. Available: http://www.fp7-pursuit.eu/PursuitWeb/ [10] FP7 PSIRP Project. [Online]. Available: http://www.psirp.org/

Kumari Nidhi Lal and Anoj Kumar / Procedia Computer Science 89 (2016) 73 – 81

[11] http://www.cnsm-conf.org/2013/documents/CNSM2013-Keynot2-G-Pavlou.pdf [12] G. Xylomenos, C. N. Ververidis, V. A. Siris, N. Fotiou, C. Tsilopoulos, X. Vasilakos, K. V. Katsaros and G. C. Polyzos, A Survey of Information-Centric Networking Research, In Communications Surveys & Tutorials, IEEE, vol. 16, no. 2, pp. 1024–1049, Second Quarter (2014). [13] Arianfar, Somaya, Pekka Nikander and J¨org Ott, Packet-level Caching for Information-Centric Networking, ACM SIGCOMM, ReArch Workshop, (2010). [14] Gallo, Massimo, et al., Performance Evaluation of the Random Replacement Policy for Networks of Caches, ACM SIGMETRICS Performance Evaluation Review, ACM, vol. 40, no. 1, (2012). [15] Rossi, Dario and Giuseppe Rossini, Caching Performance of Content Centric Networks Under Multi-Path Routing (and more). Relat´oriot´ecnico, Telecom Paris Tech, (2011). [16] K. Katsaros, G. Xylomenos and G. C. Polyzos, MultiCache: An Overlay Architecture for Information-Centric Networking, Computer Networks, pp. 1–11, (2011). [17] J. Choi, J. Han, E. Cho, T. T. Kwon and Y. Choi, A Survey on Content-Oriented Networking for Efficient Content Delivery, IEEE Communications Magazine, pp. 121–127, (2011). [18] I. Psaras, R. G. Clegg, R. Landa, W. K. Chai and G. Pavlou, Modelling and Evaluation of CCN-Caching Trees, IFIP Networking, (2011). [19] E. J. Rosensweig and J. Kurose, Breadcrumbs: Efficient, Best-Effort Content Location in Cache Networks, IEEE INFOCOM, (2009). [20] Suksomboon, Kalika, et al. Popcache: Cache More or Less based on Content Popularity for Information-Centric Networking, IEEE 38th Conference on Local Computer Networks (LCN), 2013 IEEE, (2013). [21] Ho, Cheng-Yun, Cheng-Yuan Ho and Chien-Chao Tseng, A Case Study of Cache Performance in ICN—Various Combinations of Transmission Behavior and Cache Replacement Mechanism, 17th International Conference on Advanced Communication Technology (ICACT), 2015 IEEE, (2015). [22] N. Lal, A. P. Singh and S. Kumar, Modified Trial Division Algorithm using KNJ-Factorization Method to Factorize RSA Public Key Encryption, International Conference on Contemporary Computing and Informatics (IC3I), 2014, pp. 992–995, 27–29 November (2014). [23] Lal, Kumari Nidhi, and Ashutosh Kumar Singh, Modified Design of Microstrip Patch Antenna for WiMAX Communication System, IEEE Students’ Technology Symposium (TechSym), 2014 IEEE, (2014). [24] Lal, Nidhi, et al., A Heuristic EDF Uplink Scheduler for Real Time Application in WiMAX Communication. arXiv preprint arXiv:1501.04553 (2015). [25] Yonghui Shim and Young Han Kim, Data Aggregation with Multiple Sinks in Information-Centric Wireless Sensor Network, In International Conference on Information Networking (ICOIN), 2014, pp. 13–17, 10–12 February (2014). [26] Salsano, Stefano, et al., Information Centric Networking Over SDN and OpenFlow: Architectural Aspects and Experiments on the OFELIA Testbed, Computer Networks 57.16, pp. 3207–3221, (2013). [27] S. Hassan, Z. Aziz and K. Nisar, On the Cache Performance of the Information Centric Network, In International Conference on Computing, Electrical and Electronics Engineering (ICCEEE), 2013, pp. 477–481, 26–28 August 2013. [28] B. Azimdoost, C. Westphal and H. R. Sadjadpour, On the Throughput Capacity of Information-Centric Networks, In 25th International Teletraffic Congress (ITC), 2013, pp. 1–9, 10–12 September (2013).

81