enhancing replacement policy of content-centric networking
to support reaction toward natural disaster
raaid alubady*1, samah abdalhadi2, wesam abduladheem kamil3
1 collage of informationtechnology, university of babylon, babylon, iraq 2 collage of science for women, university of babylon, babylon, iraq 3 computer science department, university of thi-qar, thi-qar, iraq, *corresponding author e-mail: raaid.n.alubady@uobabylon.edu.iq
abstract
replacement policy in content-centric networking (ccn) is a necessary and current, function as an important part in interest packet caching. pending interest table (pit) is the main and core cache tables in ccn and plays a signi?cant role for recording the information of interest packets that are forwarded but are still waiting for matching with incoming data packets. however, pit management is more fundamental with regard to ccn operations for better memory ef?ciency. the pit size determination of the forwarding system is a difficult problem in pit management. due to the limited pit sizing, pit replacement is utilized to remove the current entry from pit and constructing a new space for the incoming entry to it. in a disaster area, this problem is due to the massive interest packet that generating by survivors from the disaster and rescuers. the pit over?ow could be subjected due to use of long interest lifetimes that would simultaneously increase the number of entries in the pit. thus particularly when there is no flexible replacement policy, hence affecting pit performance. therefore, the ultimate aim of this paper is to develop the replacement policy that can deal with this problem. the proposed policy is a pit management based on ccn pit replacement policy for managing the pit during a natural disaster, which can lead to mitigating pit over?owing. the results showed the overall scenarios, the proposed policy better pit memory usage as well as decreasing the interest droping, delay time, interest lifetime and interest retransmission. a positive signi?cance in?uence in this work would be to presents a formulate a rule as a function which can decrease the delay and thus be leading to increasing pit utilization, which will be very much useful for survivors, emergency rescue teams as well as emergency operation centers.
keywords: content-centric networking, future internet architecture, pending interest table, natural disaster, network simulation.
reduce the redundant traffic that passes through the ccn networks and it also reduces an average number of access hops. communication immediately after a natural disaster situation is an important component of response and recovery, in that it connects the citizen, survivors, emergency operation centers, rescue teams, and support systems [11]. accessible and reliable communication and network systems also are key to a community’s resilience [12][13]. most probably during the natural disasters, most of the infrastructures are damaged. in this case, without proper networks, communications cannot be established for the area, leading to longer delays in emergency operations that also causing huge losses in human and infrastructures [14][15].
therefore, the aim behind this paper is that the management of pit (i.e., content replacement) has not got much attention until now since it is a new novel information structure, which does not appear in any other icn architecture. in addition, pit management is challenging because it requires updatings for each package, and the requires that is stored in the pit for a long time require more memory [16]. this is the reason for focusing on the pit in this paper where pit plays a significant part in the performance of ccn routers. moreover, the current policies that are used with pit have not considered the popularity of request, entry lifetime and request-hop number as factors for determining the pit entry which must be replaced.
thus, leading to believe that these factors are required and may affect the on pit utilization in disaster area situation. the remainder of the paper has been organized into six sections. section 2 gives a brief description of ccn, pit, replacement
1. introduction
content-centric networking (ccn) [1][2][3] is an efficient and simple communication model that design based on information- centric networking paradigm for future of the internet, which completely redesign and develop the internet by replacing ip with content chunks as universal components of transport. although there are many studies that have focused on naming, caching, security, and scalability in order to make ccn getting perfect, nevertheless, the management field especially in pending interest table (pit) is still one of the essential concerns of high-speed forwarding. thus, the management of pit is one of the most significant design details that have not been studied in the ccn context to a significant extent. moreover, pit management is more fundamental with regard to the ccn operation for better memory ef?ciency [4][5]. although it is feasible with the current technology, it leads to suffering from many issues such as pit lookup [6], pit interest lifetime [7], pit scalability [8] and pit overflow [5].
according to [9], the management techniques of pit classified into four categories: pit implementation, pit placement, pit replacement, and adaptive interest packet lifetime. for cache performance, a replacement policy has a signi?cant in?uence on it since they are playing an effective and pivotal role in memory management [10]. in respect of increasing the effectiveness of pit cached, the popular requests are stored at the pit, which may
copyright © 2018 authors. this is an open access article distributed under the creative commons attribution license, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.international journal of engineering & technology
813
policies and disaster areaon existing works. propose policy describes in section 3. section 4 determines the simulation setup, performance metrics as well. section 5 presents the results and discussions, and we conclude the paper in section 6.
2. problem basic idea and related works
in this section, explaining the main components as a background, which is related to our proposed model.
2.1. content-centric networking
of course, the internet architecture for the future will not only bring new content and media. it should be aimed to retain some of the current semantics, but with added functionalities of how data would be handled[17]. also, attention requires to be paid to the privacy policies, presentation, services, communications, and infrastructure, which are essential building constitute of the internet architecture for the future [18]. therefore, the new generation of the internet namely information-centric network (icn) [1][19][20] introduced a radical change in internet communications. the various icn initiatives are focused on designing a new architecture for the future of the internet that will replace the current host-centric model. therewith, icn has gained the interest of the research communities as a new model for networks that can better meet the needs of users in a networked world [21]. a number of designs have emerged in the last few years, including data-oriented network architecture (dona), content-based network/combined broadcast and content-based, content-centric networking (ccn), network of information, and others [21][9].
ccn has attracted much emphasis in the research area recently, with different exploration activities focusing on the rising research with the point of moving from the present internet engineering which is manufactured and intended for a host to host correspondences model. in addition, ccn is an ef?cient and simple communication model driven by subscribers who broadcast requests (i.e., interest packet) to ask for a content by name regardless of the ip addresses of the nodes that supply the content. interests’ packets are forwarded by intermediate nodes (i.e., ccn router) upstream to publishers that are any node stores or owes the requested content. publishers simply respond to the data packets request, which goes through the way back to the subscribers [22][23].
2.2. pending interest table
on the context of ccn router design, pending interest table (pit) represented a core component for forwarding the interest packet. it is one of the three cache tables newly inculcated into the ccn router designs [4]. each pit record comprises ?ve ?elds [24] as illustrated in figure 1, including content name (a name related with the entry), incoming faces list, outgoing faces list, entry expire time and forwarding strategy.
trip time (rtt) [25]. hence, the current memory technologies are not capable to handle the pit implementation. on the other hand, the pit needs to store a number of interest packet in the order of 107[26]. therefore, huge memories are needed with their known limitations, especially when dealing with their access time. moreover, pit in the ccn router is very dynamic. thus, for all arriving interest packets and matching data packets, hence a special process (i.e., lookup process) must be happening in the pit which must be completed faster. this requires quick memories unfortunately available only for small storage size. hence, the pit table may be overflowed. because pit receives and removes them exponentially [27].
2.3. pit replacement policies
replacement policy in the pit is one of the important factors that determine the effectiveness of the cache (e.g., pit). it has become more significant as technology trends emerge towards highly correlate cache practices [28]. the state-of-the-art procedures, therefore, utilize many of the caches replacement policies, indicating that there is no common replacement that stands out as the best [9]. hence, this section explores some common pit replacement policies in greater perspective. it is therefore paramount to investigate the performance of different pit in relations to replacement policies for a contemporary workload in different pit con?gurations. this will determine how some existing policies relate to pit.
in addition, pit replacement policy has a different influence on instruction and entries into pit [29]. in ccn architecture, there are three replacement policies exist, which are least recently used, random as well as persistent, and it assumes persistent as the default replacement policy in ccn [24][30]. also, a new one is designed and implemented namely highest lifetime least request [29] for this purpose.
2.4. natural disasters communication
over recent years, a number of natural disasters have occurred in the world, causing damage to human, property and almost everything in the area [31]. these disasters of nature like heavy snowfall, floods, and heavy rain (see figure 2) have always struck at unpredictable times in different places in the world, leading to the increasing losses in human and infrastructure.
fig. 2: example of natural disaster
abdul hannan at. el., [32] proposed an ndn-iot-based dms in an sc scheme in ?re disaster environment, which is enabling push support on the ?xed sequence number. for that, researchers modified legacy ndn publishes and subscribes functionality with the addition of a threshold limit and ?re sensor monitoring modules as well as modified ndn cache tables (particularly fib and pit) in disaster scenario for better ef?ciency.
reference [33] proposed a framework for flooding disaster scenario in cities and shown the ability of their mechanism (i.e., smart threshold interest lifetime) in ndn to decrease the packet delay and increase network utilization. hence, it will be very
fig. 1: pit entry fields
because of the data packet and interests packet should exactly match, pit cannot utilize aggregation. indeed, pit requires a large memory for storing the pending entries. consider a case where the average interest packets arriving rate is 125 million packets/second and each interest packet needs 80ms for its round value
parameters
simulator topology
forward strategy
ns3-ndnsim rocketfuel-mapped
flooding
input requires:
1. interest packet lifetime
2. incoming interest packet main process:
3. while the pit is full do
4. calculate the frequency for each entry that has stored into pit based on:
5. if ( ==1)
6. calculate the average number of hop count. based
?
on: ?
7. calculate the entry that has a maximum lifetime based on:
?
8. calculate the replacement entry based on:
9. end_if
10.
11. else
12. calculate the evict entry since the first option is false based on:
?
13. calculate the replacement entry based on:
14. end_else
15. // (updating pit)
16. end_while
17. end
international journal of engineering & technology
814
helpful for the rescue teams, survivors, and the emergency operation center.
enabling communication in a delay tolerant network using information-centric networking is highlighted in reference [34]. in this work, researchers focus on pull support-based disaster situations where data mules movement is random and unpredictable.
baldini at. el., in [35] also designed a fast emergency deployment mobile communication nodes, which are used in many communication technologies to provide multiple communications services in disaster management situations. also, this work supported multi-media content sharing as well as supported live video streaming along with the services of traditional communication. another research paper [36] developed a flexible network architecture, which supplies a common networking platform for a performance in emergency s situations and for heterogeneous multiple operator s networks.
3. highest lifetime lest request-hop
one of the effective mechanisms used to manage ccn routers memory (i.e., cs and pit) is the replacement policies. replacement policies of the pit are useful in expelling entry from the pit and constructing new location for incoming interest packets. due to the limited pit size, thus, it cannot store all incoming requests. therefore, by using the replacement policy, pit will have the ability to monitor incoming flow entry to it and to manage the entries inside it when it is overflowing. highest lifetime least request [29][37] is a kind of memory management policy, which is utilized to manage and monitor pit entries within the ccn routers when it is full. this is usable when the pit is overflowing and then there is a new incoming entry that requires to be added to the pit. hllr is replacing an entry in whic