Information-Centric Networks: History
View Latest Version
Please note this is an old version of this entry, which may differ significantly from the current revision.
Contributor:
John Baugh
,
Jinhua Guo

Information-Centric Networking (ICN) is a new paradigm of network architecture that focuses on content rather than hosts as first-class citizens of the network. As part of these architectures, in-network storage devices are essential to provide end users with close copies of popular content, to reduce latency and improve the overall experience for the user but also to reduce network congestion and load on the content producers.

  • Named Data Networking (NDN)
  • Information-Centric Networking (ICN)
  • cache pollution

1. Introduction

As the current host-based Internet architecture ages, it is becoming increasingly apparent that it is insufficient to handle the content-driven applications in use today. The original Internet architecture was built largely to accommodate end-to-end communication between mutually trustworthy individuals who wished to share data and expensive resources. Since then, the use of the Internet as a communications network has shifted toward its use as a content delivery apparatus.
Because of these limitations, as technology and user needs have changed, overlay patches providing security, privacy, mobility, and content delivery have been created to sit on top of the underlying network. These application overlays provide services that the original Internet architects would have never imagined: mobile communications, banking, shopping, multiplayer gaming, streaming media, and many other services. However, the application overlays are not able to benefit from any optimizations at the network layer because they are simply not provided.
As a result, many researchers have pushed for a complete redesign of the fundamental architecture of the Internet to focus on content as the primary citizen of the network instead of location. For many applications, users want specific content but do not care where it comes from as long as the content is genuine and transmitted reliably and quickly. One architectural paradigm that seeks to accomplish the goals of a future Internet redesign is Information-Centric Networking (ICN) [1][2][3].
Several projects within the large umbrella of ICN have sought to achieve these goals with various approaches and to varying degrees of success. Prominent ICN projects include Data-Oriented Network (DONA) [4], Network of Information (NetInf) [5][6], Content-Centric Networking (CCN) [7][8], and Named Data Networking (NDN) [9].
Many interesting research challenges exist to provide the desired features of ICN. Researchers discuss technology related to in-network caching of data, which is a key feature to lowering latency and overall network congestion, using technology at the network layer. While in-network storage devices in ICN seek to provide similar services to those that Content Distribution (or Delivery) Networks (CDNs) [10] do today, in-network storage devices do so at the network layer and take advantage of various optimizations therein. In-network caching is discussed in a variety of recent research studies as a primary goal in ICN and NDN [11][12][13].
Given the pervasiveness of in-network storage devices in ICN architectures, these devices are prime targets for malicious users. In terms of in-network caching of popular data, an attacker would seek to reduce or remove any benefit experienced by the presence of in-network storage devices, such as content routers. An attack well suited for this type of service degradation is the cache pollution attack [14]. In this type of attack, the malicious attempt is to push truly popular content objects out of the cache in favor of less popular content objects. Ultimately, the goal is to reduce the hit rate experienced by users, such that the content that is genuinely popular is rarely or never in the content store of the in-network storage devices.
These types of attacks are not new—they have been experienced for years in the current IP-based Internet, such as in CDNs and other situations involving proxy caching servers [15]. However, these attacks will certainly be even more of a threat in an architecture where in-network storage is part of the fundamental fabric of the network and not just an overlay.
Several proposed solutions for preventing or reducing the effects of cache pollution attacks in ICN exist. However, most are insufficient for various reasons. Some lack adequate security and can be easily overcome, such as CacheShield [16]. Others assume all abnormal behaviors follow uniform distributions [17], and still others require significant computational or memory overhead, such as ANFIS [18]. Most current techniques attempt to detect attacks, which is often difficult when attackers employ various smart attacks. Additionally, these techniques can yield false positives if a user is even a little more aggressive in their request frequencies, and they may punish them even though they were not being malicious.

2. Cache Pollution Attacks

Fundamentally, cache pollution attacks come in two different forms: locality disruption and false locality. The term locality refers to a phenomenon in which certain content objects from a given requesting region tend to be more popular in that region than other content objects. Thus, a specific arrangement of content objects based on their popularity is representative of a locale or region.
Therefore, locality disruption refers to an attempt by an attacker to simply churn the cache of a given locale in hopes of knocking some of the popular content objects out of the cache. Thus, the attacker will request many different content objects from across the universe of content (the content space). A false locality attack, on the other hand, is an arguably more aggressive attack in which an attacker wishes to push even the most popular content objects out of the cache in favor of unpopular items. Thus, the attacker wants the cache full of content objects that are not at all representative of the desires of users utilizing the cache. Hence, the locality expressed by the cache is false relative to the actual users’ desires. It is important to note that most security approaches to cache pollution attacks tend to seek to prevent one (or sometimes both) of these types of attacks [10].

3. Approaches to Secure against Cache Pollution Attacks

One of the best-known examples in the literature in terms of attempting to fight cache pollution is CacheShield, developed by Xie et al. [12]. This approach records the interest frequency of items that are not cached. As the recorded frequency t increases, the probability that a content object will be cached increases as well due to the shielding function. However, a major weakness of this approach is that an attacker need only determine the threshold of t values and request content objects at a frequency beyond the threshold of t. Then, their own content objects will be cached.
A cache pollution detection approach using randomness checks was proposed by Park et al. [19]. The randomness check approach utilizes the fact that attackers launching a locality disruption attack request contents in a nearly uniform manner across the content universe. The authors reason that it is less likely that attackers will be able to know what contents are popular and that this popularity changes with time. Therefore, they consider false locality attacks far less likely than locality disruption attacks, which do not require detailed knowledge of the popularity of contents. The approach performs matrix operations, during which popular contents are more likely to be removed from the matrices due to the Zipf-like distributions, whereas attacker requests will remain. This causes the system to rank content requests as malicious when they remain with a given rank value.
As an attempt to improve upon CacheShield [16], the authors of [17] assume a slightly more intelligent attack scenario. They enhance their threat model such that the attackers only focus on a small subset of the content space, namely the items at the tail end of the Zipf distribution. This approach uses a machine-learning-based approach. However, it still suffers from weaknesses, including the assumption that all normal behavior is Zipf-like and attackers are always uniform in their distribution of requests.
A neural network (with fuzzy systems) approach was proposed by Karami et al. in [18]. This approach uses techniques that are used in linguistics to determine the degree to which different inputs and outputs are related. Although this approach is reasonably effective at detecting cache pollution attacks (both false locality and locality disruption), the overhead, both in terms of memory and computational complexity, is significant. For these reasons, this approach is not considered to be very scalable [14].

This entry is adapted from the peer-reviewed paper 10.3390/network3040022

References

  1. Ahlgren, B.; Dannewitz, C.; Imbrenda, C.; Kutscher, D.; Ohlman, B. A Survey of Information-Centric Networking. IEEE Commun. Mag. 2012, 50, 26–36.
  2. Xylomenos, G.; Ververids, C.N.; Siris, V.A.; Fotiou, N.; Tsilopoulos, C.; Vasilakos, X.; Katsaros, K.V.; Polyzos, G.C. A Survey of Information-Centric Networking Research. Commun. Surv. Tutor. 2014, 16, 1024–1049.
  3. Somayeh Kianpisheh, T.T. A Survey on In-Network Computing: Programmable Data Plane and Technology Specific Applications. IEEE Commun. Surv. Tutor. 2023, 25, 701–761.
  4. Koponen, T.; Chawla, M.; Chun, B.-G.; Ermolinskiy, A.; Kim, K.H.; Shenker, S.; Stoica, I. A Data-Oriented (and Beyond) Network Architecture. In Proceedings of the SIGCOMM ’07, Kyoto, Japan, 27–31 August 2007.
  5. Ahlgren, B.; Marchisio, M.; D’Ambrosio, M.; Marsh, I.; Dannewitz, C.; Ohlman, B.; Pentikousis, K.; Strandberg, O.; Remarz, R.; Vercellone, V. Design Considerations for a Network of Information. In Proceedings of the ACM ReArch ’08, Madrid, Spain, 9–12 December 2008.
  6. D’Ambrosio, M.; Dannewitz, C.; Karl, H. MDHT: A hierarchical name resolution service for information-centric networks. In Proceedings of the ACM SIGCOMM, Toronto, ON, Canada, 19 August 2011.
  7. Content Centric Networking Project. Available online: www.ccnx.org (accessed on 23 July 2023).
  8. Jacobson, V.; Smetters, D.; Thornton, J.; Plass, M.; Briggs, N.; Braynard, R. Networking Named Content. In Proceedings of the ACM CoNEXT ’09, Rome, Italy, 1–4 December 2009.
  9. Zhang, L.; Afanasyev, A.; Burke, J.; Jacobson, V.; Claffy, K.C.; Crowley, P.; Papadopoulos, C.; Wang, L.; Zhang, B. Named Data Networking. ACM SIGCOMM Comput. Commun. Rev. 2014, 44, 66–73.
  10. Nygren, E.; Sitaraman, R.; Sun, J. The Akamai Network: A Platform for High-Performance Internet Applications. ACM SIGOPS Oper. Syst. Rev. 2010, 44, 2–19.
  11. Zha, Y.; Cui, P.; Hu, Y.; Xue, L.; Lan, J.; Wang, Y. An NDN Cache-Optimization Strategy Based on Dynamic Popularity and Replacement Value. Electronics 2022, 11, 3014.
  12. Alubady, R.; Salman, M.; Mohamed, A.S. A review of modern caching strategies in named data network: Overview, classification, and research directions. Telecommun. Syst. 2023, 1–46.
  13. Liu, Z.; Jin, X.; Li, Y.; Zhang, L. NDN-Based Coded Caching Strategy for Satelite Networks. Electronics 2023, 12, 3756.
  14. Tourani, R.; Misra, S.; Mick, T.; Panway, G. Security, Privacy, and Access Control in Information-Centric Networking: A Survey. IEEE Commun. Surv. Tutor. 2018, 20, 566–600.
  15. Gao, Y.; Deng, L.; Kuzmanovic, A.; Chen, Y. Internet Cache Pollution Attacks and Countermeasures. In Proceedings of the 2006 IEEE International Conference on Network Protocols, Santa Barbara, CA, USA, 12–15 November 2006.
  16. Xie, M.; Widjaja, I.; Wang, H. Enhancing Cache Robustness for Content-Centric Networking. In Proceedings of the IEEE INFOCOM, Orlando, FL, USA, 25–30 March 2012.
  17. Conti, M.; Gasti, P.; Teoli, M. A lightweight mechanism for detection of cache pollution attacks in Named Data Networking. Comput. Netw. 2013, 57, 3178–3191.
  18. Karami, A.; Guerrero-Zapata, M. An ANFIS-based cache replacement method for mitigating cache pollution attacks in Named Data Networking. Elsevier J. Comput. Netw. 2015, 80, 51–65.
  19. Park, H.; Widjaja, I.; Lee, H. Detection of Cache Pollution Attacks Using Randomness Checks. In Proceedings of the ICC 2012—Communication and Information Systems Security Symposium, Ottawa, ON, Canada, 10–15 June 2012.
More
© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.
This entry is offline, you can click here to edit this entry!
Video Production Service
Feedback