2.4. Infrastructure Sharing A key idea in ICN is that the network should secure information objects per se, not the communications channel that they are delivered over. This means that hosts attached to an information- centric network can share resources on an unprecedented scale, especially when compared to what is possible in an IP network. All devices with network access and storage capacity can contribute their resources thereby increasing the value of an information-centric
network, although compensation schemes motivating users to contribute resources remain a research challenge primarily from a business perspective. For example, Jacobson et al. [CBIS] argue that in ICN the "where and how" of obtaining information are new degrees of freedom. They illustrate this with a scenario involving a photo-sharing application that takes advantage of whichever access network connectivity is available at the moment (WLAN, Bluetooth, and even SMS) without requiring a centralized infrastructure to synchronize between numerous devices. It is important to highlight that since the focus of communication changes, keep-alives in this scenario are simply unnecessary, as devices participating in the testbed network contribute resources in order to maintain user content consistency, not link state information as is the case in the host-centric paradigm. This means that the notion of "infrastructure" may be completely different in the future. Muscariello et al. [SHARE], for instance, presented early work on an analytical framework that attempts to capture the storage/bandwidth tradeoffs that ICN enables and can be used as the foundation for a network planning tool. In addition, Chai et al. [CL4M] explore the benefits of ubiquitous caching throughout an information-centric network and argue that "caching less can actually achieve more." These papers also sit alongside a variety of other studies that look at various scenarios such as caching HTTP-like traffic [CCNCT] and BitTorrent-like traffic [BTCACHE]. We observe that much more work is needed in order to understand how to make optimal use of all resources available in an information-centric network. In real-world deployments, policy and commercial considerations are also likely to affect the use of particular resources, and more work is expected in this direction as well. In conclusion, scenarios in this category would cover the communication-computation-storage tradeoffs that an ICN deployment must consider. This would exercise features relating to network planning, perhaps capitalizing on user-provided resources, as well as operational and economical aspects of ICN, and contrast them with other approaches. An obvious baseline to compare against in this regard is existing federations of IP-based Content Distribution Networks (CDNs), such as the ones discussed in the IETF Content Delivery Networks Interconnection Working Group.
2.5. Content Dissemination Content dissemination has attracted more attention than other aspects of ICN. Scenarios in this category abound in the literature, including stored and streaming A/V distribution, file distribution, mirroring and bulk transfers, versioned content services (cf. Subversion-type revision control), as well as traffic aggregation. Decentralized content dissemination with on-the-fly aggregation of information sources was envisaged in [N-Scen], where information objects can be dynamically assembled based on hierarchically structured subcomponents. For example, a video stream could be associated with different audio streams and subtitle sets, which can all be obtained from different sources. Using the topology depicted in Figure 1 as an example, an application at C1 may end up obtaining, say, the video content from I1, but the user-selected subtitles from Px. Semantics and content negotiation, on behalf of the user, were also considered, e.g., for the case of popular tunes that may be available in different encoding formats. Effectively, this scenario has the information consumer issuing independent requests for content based on information identifiers, and stitching the pieces together irrespective of "where" or "how" they were obtained. A case in point for content dissemination are vehicular ad hoc networks (VANETs), as an ICN approach may address their needs for information dissemination between vehicles better than today's solutions, as discussed in the following section. The critical part of information dissemination in a VANET scenario revolves around "where" and "when". For instance, one may be interested in traffic conditions 2 km ahead while having no interest in similar information about the area around the path origin. VANET scenarios may provide fertile ground for showcasing the ICN advantage with respect to content dissemination especially when compared with current host- centric approaches. That said, information integrity and filtering are challenges that must be addressed. As mentioned above, content dissemination scenarios in VANETs have a particular affinity to the mobility scenarios discussed in Section 2.3. Content dissemination scenarios, in general, have a large overlap with those described in the previous sections and are explored in several papers, such as [DONA], [PSI], [PSIMob], [NetInf], [CCN], [CBIS], and [CCR], just to name a few. In addition, Chai et al. [CURLING] present a hop-by-hop hierarchical content resolution approach that employs receiver-driven multicast over multiple domains, advocating another content dissemination approach. Yet, largely, work in this area did not address the issue of access authorization in detail. Often, the distributed content is mostly assumed to be freely accessible by any consumer. Distribution of
paid-for or otherwise restricted content on a public ICN network requires more attention in the future. Fotiou et al. [ACDICN] consider a scheme to this effect, but it still requires access to an authorization server to verify the user's status after the (encrypted) content has been obtained. This may effectively negate the advantage of obtaining the content from any node, especially in a disruption-prone or mobile network. In summary, scenarios in this category aim to exercise primarily scalability and the cost and performance attributes of content dissemination. Particularly, they should highlight the ability of an ICN to scale to billions of objects, while not exceeding the cost of existing content dissemination solutions (i.e., CDNs) and, ideally, increasing performance. These should be shown in a holistic manner, improving content dissemination for both information consumers and publishers of all sizes. We expect that in particular for content dissemination, in both extreme as well as typical scenarios, can be specified by drawing data from current CDN deployments. 2.6. Vehicular Networking Users "on wheels" are interested in road safety, traffic efficiency, and infotainment applications that can be supported through vehicle- to-vehicle (V2V) and vehicle-to-infrastructure (V2I) wireless communications. These applications exhibit unique features in terms of traffic generation patterns, delivery requirements, and spatial and temporal scope, which pose great challenges to traditional networking solutions. VANETs, by their nature, are characterized by challenges such as fast-changing topology, intermittent connectivity, and high node mobility, but also by the opportunity to combine information from different sources as each vehicle does not care about "who" delivers the named data objects. ICN is an attractive candidate solution for vehicular networking, as it has several advantages. First, ICN fits well to the nature of typical vehicular applications that are geography- and time-dependent (e.g., road traveler information, accident warning, point-of-interest advertisements) and usually target vehicles in a given area, regardless of their identity or IP address. These applications are likely to benefit from in-network and decentralized data caching and replication mechanisms. Second, content caching is particularly beneficial for intermittent on-the-road connectivity and can speed up data retrieval through content replication in several nodes. Caching can usually be implemented at relatively low cost in vehicles, as the energy demands of the ICN device are likely to be a negligible fraction of the total vehicle energy consumption, thus allowing for sophisticated processing, continuous communication, and adequate storage in the vehicle. Finally, ICN natively supports asynchronous
data exchange between end-nodes. By using (and redistributing) cached named information objects, a mobile node can serve as a link between disconnected areas. In short, ICN can enable communication even under intermittent network connectivity, which is typical of vehicular environments with sparse roadside infrastructure and fast- moving nodes. The advantages of ICN in vehicular networks were preliminarily discussed in [EWC] and [DMND], and additionally investigated in [DNV2V], [RTIND], [CCNHV], [CCDIVN], [CCNVANET], and [CRoWN]. For example, Bai and Krishnamachari [EWC] take advantage of the localized and dynamic nature of a VANET to explore how a road congestion notification application can be implemented. Wang et al. [DMND] consider data collection where Road-Side Units (RSUs) collect information from vehicles by broadcasting NDN-like Interest packets. The proposed architecture is evaluated using simulation in a grid topology and is compared against a host-centric alternative based on Mobile IP. See Figure 3 for an indicative example of an urban VANET topology. Their results indicate high efficiency for ICN even at high speeds. That said, this work is a preliminary exploration of ICN in vehicular environments, so various issues remain for evaluation. They include system scalability to large numbers of vehicles and the impact of vehicles that forward Interest packets or relay data to other vehicles. + - - _- - -_- - - -_- - _- - - + | /_\ /_\ /_\ /_\ | | o o o o o o o o | | +-------+ +-------+ _ | | | | | |/_\ | | _ | | | |o o | | /_\| | | | | | o o+--_----+\===/+--_----+ | | /_\ |RSU| /_\ | | o o /===\ o o | | +-------+ +-------+ _ | | | | | |/_\ | | _ | | | |o o | |/_\ | | | | | |o o +_-----_+ +_-----_+ | | /_\ /_\ /_\ /_\ | +_ _ o_o_ _o_o_ _ _o_o_ _o_o_ _ + Figure 3. Urban Grid VANET Topology As mentioned in the previous section, due to the short communication duration between a vehicle and the RSU, and the typically short time of sustained connectivity between vehicles, VANETs may be a good
showcase for the ICN advantages with respect to content dissemination. Wang et al. [DNV2V], for instance, analyze the advantages of hierarchical naming for vehicular traffic information dissemination. Arnould et al. [CCNHV] apply ICN principles to safety information dissemination between vehicles with multiple radio interfaces. In [CCDIVN], TalebiFard and Leung use network coding techniques to improve content dissemination over multiple ICN paths. Amadeo et al. [CCNVANET] [CRoWN] propose an application-independent ICN framework for content retrieval and distribution where the role of provider can be played equivalently by both vehicles and RSUs. ICN forwarding is extended through path-state information carried in Interest and Data packets, stored in a new data structure kept by vehicular nodes, and exploited also to cope with node mobility. Typical scenarios for testing content distribution in VANETs may be highways with vehicles moving in straight lines, with or without RSUs along the road, as shown in Figure 4. With an NDN approach in mind, for example, RSUs may send Interest packets to collect data from vehicles [DMND], or vehicles may send Interest packets to collect data from other peers [RTIND] or from RSUs [CCNVANET]. Figure 2 applies to content dissemination in VANET scenarios as well, where C0 represents a vehicle that can obtain named information objects via multiple wireless peers and/or RSUs (I2 and I3 in the figure). Grid topologies such as the one illustrated in Figure 3 should be considered in urban scenarios with RSUs at the crossroads or co-located with traffic lights as in [CRoWN]. \___/ \___/ |RSU| |RSU| ================================ _ _ _ _ /_\ /_\ /_\ /_\ _ _ o_o_ _o_o_ _o_o_ _o_o_ _ _ _ _ _ _ _ /_\ /_\ /_\ /_\ o o o o o o o o ================================ Figure 4. Highway VANET Topology To summarize, VANET scenarios aim to exercise ICN deployment from various perspectives, including scalability, caching, transport, and mobility issues. There is a need for further investigation in (i) challenging scenarios (e.g., disconnected segments); (ii) scenarios involving both consumer and provider mobility; (iii) smart caching techniques that take into consideration node mobility patterns, spatial and temporal relevance, content popularity, and social relationships between users/vehicles; (iv) identification of new
applications (beyond data dissemination and traffic monitoring) that could benefit from the adoption of an ICN paradigm in vehicular networks (e.g., mobile cloud, social networking). 2.7. Delay- and Disruption-Tolerance Delay- and Disruption-Tolerant Networking (DTN) originated as a means to extend the Internet to interplanetary communications [DTN]. However, it was subsequently found to be an appropriate architecture for many terrestrial situations as well. Typically, this was where delays were greater than protocols such as TCP could handle, and where disruptions to communications were the norm rather than occasional annoyances, e.g., where an end-to-end path does not necessarily exist when communication is initiated. DTN has now been applied to many situations, including opportunistic content sharing, handling infrastructural issues during emergency situations (e.g., earthquakes) and providing connectivity to remote rural areas without existing Internet provision and little or no communications or power infrastructure. The DTN architecture [RFC4838] is based on a "store, carry, and forward" paradigm that has been applied extensively to situations where data is carried between network nodes by a "data mule", which carries bundles of data stored in some convenient storage medium (e.g., a USB memory stick). With the advent of sensor and peer-to- peer (P2P) networks between mobile nodes, DTN is becoming a more commonplace type of networking than originally envisioned. Since ICN also does not rely on the familiar end-to-end communications paradigm, there are clear synergies [DTNICN]. It could therefore be argued that many of the key principles embodied within DTN also exist in ICN, as we explain next. First, both approaches rely on in-network storage. In the case of DTN, bundles are stored temporarily on devices on a hop-by-hop basis. In the case of ICN, information objects are also cached on devices in a similar fashion. As such, both paradigms must provision storage within the network. Second, both approaches espouse late binding of names to locations due to the potentially large interval between request and response generation. In the case of DTN, it is often impossible to predict the exact location (in a disconnected topology) where a node will be found. Similarly, in the case of ICN, it is also often impossible to predict where an information object might be found. As such, the binding of a request/bundle to a destination (or routing locator) must be performed as late as possible.
Finally, both approaches treat data as a long-lived component that can exist in the network for extended periods of time. In the case of DTN, bundles are carried by nodes until appropriate next hops are discovered. In the case of ICN, information objects are typically cached until storage is exhausted. As such, both paradigms require a direct shift in the way applications interact with the network. Through these similarities, it becomes possible to identify many DTN principles that are already in existence within ICN architectures. For example, ICN nodes will often retain information objects locally, making them accessible later on, much as DTN bundles are handled. Consequently, these synergies suggest strong potential for marrying the two technologies. This could include, for instance, building new integrated Information-Centric Delay Tolerant Network (ICDTN) protocols or, alternatively, building ICN schemes over existing DTN protocols (and vice versa). The above similarities suggest that integration of the two principles would be feasible. Beyond this, there are also a number of identifiable direct benefits. Through caching and replication, ICN offers strong information resilience, whilst, through store-and- forward, DTN offers strong connectivity resilience. As such, both architectures could benefit greatly from each other. Initial steps have already been taken in the DTN community to integrate ICN principles, e.g., the Bundle Protocol Query Block [BPQ] has been proposed for the DTN Bundle Protocol [RFC5050]. Similarly, initial steps have also been taken in the ICN community, such as [SLINKY]. In fact, the Scalable and Adaptive Internet Solutions (SAIL) project has developed a prototype implementation of NetInf running over the DTN Bundle Protocol. Of course, in many circumstances, information-centricity is not appropriate for use in delay- and disruption-tolerant environments. This is particularly the case when information is not the key communications atom transmitted. Further, situations where a single sink is always used for receiving information may not warrant the identification and routing of independent information objects. However, there are a number of key scenarios where clear benefits could be gained by introducing information-centric principles into DTNs, two of which we describe later in this section. For the purpose of evaluating the use of ICNs in a DTN setting, two key scenarios are identified in this document. (Note the rest of this section uses the term "ICDTN".) These are both prominent use cases that are currently active in both the ICN and DTN communities. The first is opportunistic content sharing, whilst the second is the use of ad hoc networks during disaster recovery (e.g., earthquakes). We discuss both types of scenarios in the context of a simulation-
based evaluation: due to the scale and mobility of DTN-like setups, this is the primary method of evaluation used. Within the DTN community, the majority of simulations are performed using the Opportunistic Network Environment (ONE) simulator [ONE], which is referred to in this document. Before exploring the two scenarios, the key shared components of their simulation are discussed. This is separated into the two primary inputs that are required: the environment and the workload. In both types of scenarios the environment can be abstractly modeled by a time series of active connections between device pairs. Unlike other scenarios in this document, an ICDTN scenario therefore does not depend on (relatively) static topologies but, rather, a set of time-varying disconnected topologies. In opportunistic networks, these topologies are actually products of the mobility of users. For example, if two users walk past each other, an opportunistic link can be created. There are two methods used to generate these mobility patterns and, in turn, the time series of topologies. The first is synthetic, whereby a (mathematical) model of user behavior is created in an agent-based fashion, e.g., random waypoint, Gauss-Markov. The second is trace-driven, whereby the mobility of real users is recorded and used. In both cases, the output is a sequence of time- stamped "contacts", i.e., periods of time in which two devices can communicate. An important factor missing from typical mobility traces, however, is the capacity of these contacts: how much data can be transferred? In both approaches to modeling mobility, links are usually configured as Bluetooth or Wi-Fi (ONE easily allows this, although lower-layer considerations are ignored, e.g., interference). This is motivated by the predominance of these technologies on mobile phones. The workload in an ICDTN is modeled much like the workload within the other scenarios. It involves object creation/placement and object retrieval. Object creation/placement can either be done statically at the beginning of the simulations or, alternatively, dynamically based on a model of user behavior. In both cases, the latter is focused on, as it models far better the characteristics of the scenarios. Once the environment and workload have been configured, the next step is to decide the key metrics for the study. Unlike traditional networking, the QoS expectation is typically far lower in an ICDTN, thereby moving away from metrics such as throughput. At a high level, it is of clear interest to evaluate different ICN approaches with respect to both their delay- and disruption-tolerance (i.e., how effective is the approach when used in an environment subject to significant delay and/or disruption) and to their active support for operations in a DTN environment.
The two most prominent metrics considered in a host-centric DTN are delivery probability and delivery delay. The former relates to the probability by which a sent message will be received within a certain delay bound, whilst the latter captures the average length of time it takes for nodes to receive the message. These metrics are similarly important in an ICDTN, although they are slightly different due to the request-response nature of ICN. Therefore, the two most prominent evaluative metrics are satisfaction probability and satisfaction delay. The former refers to the probability by which an information request (e.g., Interest) will be satisfied (i.e., how often a Data response will be received). Satisfaction delay refers to the length of time it takes an information request to be satisfied. Note that the key difference between the host-centric and information-centric metrics is the need for a round-trip rather than a one-way communication. Beyond this, depending on the focus of the work, other elements that may be investigated include name resolution, routing, and forwarding in disconnected parts of the network; support for unidirectional links; number of round trips needed to complete a data transfer; long-term content availability (or resilience); efficiency in the face of disruption; and so on. It is also important to weigh these performance metrics against the necessary overheads. In the case of an ICDTN, this is generally measured by the number of message replicas required to access content. Note that routing in a DTN is often replication based, which leads to many copies of the same message. 2.7.1. Opportunistic Content Sharing The first key baseline scenario in this context is opportunistic content sharing. This occurs when mobile nodes create opportunistic links between each other to share content of interest. For example, people riding on an underground train can pass news items between their mobile phones. Equally, content generated on the phones (e.g., tweets [TWIMIGHT]) could be stored for later forwarding (or even forwarded amongst interested passengers on the train). Such scenarios, clearly, must be based around either the altruistic or incentivized interaction amongst users. The latter is a particularly active area of research. These networks are often termed "pocket- switched networks", as they are independently formed between the user devices. Here, the evaluative scenario of ICDTN microblogging is proposed. As previously discussed, the construction of such an evaluative scenario requires a formalization of its environment and workload. Fortunately, there exist a number of datasets that offer exactly this information required for microblogging.
In terms of the environment (i.e., mobility patterns), the Haggle project produced contact traces based on conference attendees using Bluetooth. These traces are best targeted at application scenarios in which a small group of (50-100) people are in a relatively confined space. In contrast, larger-scale traces are also available, most notably MIT's Reality Mining project. These are better suited for cases where longer-term movement patterns are of interest. The second input, workload, relates to the creation and consumption of microblogs (e.g., tweets). This can be effectively captured because subscriptions conveniently formalize who consumes what. For bespoke purposes, specific data can be directly collected from Twitter for trace-driven simulations. Several Twitter datasets are already available to the community containing a variety of data, ranging from Tweets to follower graphs. See <http://www.tweetarchivist.com> and <http://socialcomputing.asu.edu/datasets/Twitter>. These datasets can therefore be used to extract information production, placement, and consumption. 2.7.2. Emergency Support and Disaster Recovery The second key baseline scenario in this context relates to the use of ICDTNs in emergency scenarios. In these situations, it is typical for infrastructure to be damaged or destroyed, leading to the collapse of traditional forms of communications (e.g., cellular telephone networks). This has been seen in the recent North Indian flooding, as well as the 2011 Tohoku earthquake and tsunami. Power problems often exacerbate the issue, with communication failures lasting for days. Therefore, in order to address this, DTNs have been used due to their high levels of resilience and independence from fixed infrastructure. The most prominent use of DTNs in disaster areas would be the dissemination of information, e.g., warnings and evacuation maps. Unlike the previous scenario, it can be assumed that certain users (e.g., emergency responders) are highly altruistic. However, it is likely many other users (e.g., endangered civilians) might become far more conservative in how they use their devices for battery-conserving purposes. Here, we focus on the dissemination of standard broadcast information that should be received by all parties; generally, this is something led by emergency responders. For the environmental setup, there are no commonly used mobility traces for disaster zones, unlike in the previous scenario. This is clearly due to the difficultly (near impossibility) of acquiring them in a real setting. That said, various synthetic models are available. The Post-Disaster Mobility Model [MODEL1] models civilians and emergency responders after a disaster has occurred,
with people attempting to reach evacuation points (this has also been implemented in the ONE simulator). Aschenbruck et al. [MODEL2] focus on emergency responders, featuring the removal of nodes from the disaster zone, as well as things like obstacles (e.g., collapsed buildings). Cabrero et al. [MODEL3] also look at emergency responders but focus on patterns associated with common procedures. For example, command and control centers are typically set up with emergency responders periodically returning. Clearly, the mobility of emergency responders is particularly important in this setting because they usually are the ones who will "carry" information into the disaster zone. It is recommended that one of these emergency- specific models be used during any evaluations, due to the inaccuracy of alternate models used for "normal" behavior. The workload input in this evaluative scenario is far simpler than for the previous scenario. In emergency cases, the dissemination of individual pieces of information to all parties is the norm. This is often embodied using things like the Common Alert Protocol (CAP), which is an XML standard for describing warning message. It is currently used by various systems, including the Integrated Public Alert & Warning System and Google Crisis Response. As such, small objects (e.g., 512 KB to 2 MB) are usually generated containing text and images; note that the ONE simulator offers utilities to easily generate these. These messages are also always generated by central authorities, therefore making the placement problem easier (they would be centrally generated and given to emergency responders to disseminate as they pass through the disaster zone). The key variable is therefore the generation rate, which is synonymous with the rate that microblogs are written in the previous scenario. This will largely be based on the type of disaster occurring; however, hourly updates would be an appropriate configuration. Higher rates can also be tested, based on the rate at which situations change (landslides, for example, can exhibit highly dynamic properties). To summarize, this section has highlighted the applicability of ICN principles to existing DTN scenarios. Two evaluative setups have been described in detail, namely, mobile opportunistic content sharing (microblogging) and emergency information dissemination. 2.8. Internet of Things Advances in electronics miniaturization combined with low-power wireless access technologies (e.g., ZigBee, Near Field Communication (NFC), Bluetooth, and others) have enabled the coupling of interconnected digital services with everyday objects. As devices with sensors and actuators connect into the network, they become "smart objects" and form the foundation for the so-called Internet of
Things (IoT). IoT is expected to increase significantly the amount of content carried by the network due to machine-to-machine (M2M) communication as well as novel user-interaction possibilities. Yet, the full potential of IoT does not lie in simple remote access to smart object data. Instead, it is the intersection of Internet services with the physical world that will bring about the most dramatic changes. Burke [IoTEx], for instance, makes a very good case for creating everyday experiences using interconnected things through participatory sensing applications. In this case, inherent ICN capabilities for data discovery, caching, and trusted communication are leveraged to obtain sensor information and enable content exchange between mobile users, repositories, and applications. Kutscher and Farrell [IWMT] discuss the benefits that ICN can provide in these environments in terms of naming, caching, and optimized transport. The Named Information URI scheme (ni) [RFC6920], for instance, could be used for globally unique smart object identification, although an actual implementation report is not currently available. Access to information generated by smart objects can be of varied nature and often vital for the correct operation of large systems. As such, supporting timestamping, security, scalability, and flexibility need to be taken into account. Ghodsi et al. [NCOA] examine hierarchical and self-certifying naming schemes and point out that ensuring reliable and secure content naming and retrieval may pose stringent requirements (e.g., the necessity for employing PKI), which can be too demanding for low- powered nodes, such as sensors. That said, earlier work by Heidemann et al. [nWSN] shows that, for dense sensor network deployments, disassociating sensor naming from network topology and using named content at the lowest level of communication in combination with in- network processing of sensor data is feasible in practice and can be more efficient than employing a host-centric binding between node locator and the content existing therein. Burke et al. [NDNl] describe the implementation of a building automation system for lighting control where the security, naming, and device discovery NDN mechanisms are leveraged to provide configuration, installation, and management of residential and industrial lighting control systems. The goal is an inherently resilient system, where even smartphones can be used for control. Naming reflects fixtures with evolved identification and node- reaching capabilities, thus simplifying bootstrapping, discovery, and user interaction with nodes. The authors report that this ICN-based system requires less maintenance and troubleshooting than typical IP-based alternatives.
Biswas et al. [CIBUS] visualize ICN as a contextualized information- centric bus (CIBUS) over which diverse sets of service producers and consumers coexist with different requirements. ICN is leveraged to unify different platforms to serve consumer-producer interaction in both infrastructure and ad hoc settings. Ravindran et al. [Homenet] show the application of this idea in the context of a home network, where consumers (residents) require policy-driven interactions with diverse services such as climate control, surveillance systems, and entertainment systems. Name-based protocols are developed to enable zero-configuration node and service discovery, contextual service publishing and subscription, policy-based routing and forwarding with name-based firewall, and hoc device-to-device communication. IoT exposes ICN concepts to a stringent set of requirements that are exacerbated by the quantity of nodes, as well as by the type and volume of information that must be handled. A way to address this is proposed in [IoTScope], which tackles the problem of mapping named information to an object, diverting from the currently typical centralized discovery of services and leveraging the intrinsic ICN scalability capabilities for naming. It extends the base [PURSUIT] design with hierarchically based scopes, facilitating lookup, access, and modifications of only the part of the object information that the user is interested in. Another important aspect is how to efficiently address resolution and location of the information objects, particularly when large numbers of nodes are connected, as in IoT deployments. In [ICN-DHT], Katsaros et al. propose a Distributed Hash Table (DHT) that is compared with the Data-Oriented Network Architecture described in [DONA]. Their results show how topological routing information has a positive impact on resolution, at the expense of memory and processing overhead. The use of ICN mechanisms in IoT scenarios faces the most dynamic and heterogeneous type of challenges, when taking into consideration the requirements and objectives of such integration. The disparity in technologies (not only in access technologies, but also in terms of end-node diversity such as sensors, actuators, and their characteristics) as well as in the information that is generated and consumed in such scenarios, will undoubtedly bring about many of the considerations presented in the previous sections. For instance, IoT shares similarities with the constraints and requirements applicable to vehicular networking. Here, a central problem is the deployment of mechanisms that can use opportunistic connectivity in unreliable networking environments (similar to the vehicular networking and DTN scenarios). However, one important concern in IoT scenarios, also motivated by this strongly heterogeneous environment, is how content dissemination will be affected by the different semantics of the disparate
information and content being shared. In fact, this is already a difficult problem that goes beyond the scope of ICN [SEMANT]. With the ability of the network nodes to cache forwarded information to improve future requests, a challenge arises regarding whether the ICN fabric should be involved in any kind of procedure (e.g., tagging) that facilitates the relationship or the interpretation of the different sources of information. Another issue lies with the need for having energy-efficiency mechanisms related to the networking capabilities of IoT infrastructures. Often, the devices in IoT deployments have limited battery capabilities, and thus need low power consumption schemes working at multiple levels. In principle, energy efficiency gains should be observed from the inherent in-network caching capability. However, this might not be the most usual case in IoT scenarios, where the information (particularly from sensors or controlling actuators) is more akin to real-time traffic, thus reducing the scale of potential savings due to ubiquitous in-network caching. ICN approaches, therefore, should be evaluated with respect to their capacity to handle the content produced and consumed by extremely large numbers of diverse devices. IoT scenarios aim to exercise ICN deployment from different aspects, including ICN node design requirements, efficient naming, transport, and caching of time- restricted data. Scalability is particularly important in this regard as the successful deployment of IoT principles could increase both device and content numbers dramatically beyond all current expectations. 2.9. Smart City The rapid increase in urbanization sets the stage for the most compelling and challenging environments for networking. By 2050 the global population will reach nine billion people, 75% of which will dwell in urban areas. In order to cope with this influx, many cities around the world have started their transformation toward the "smart city" vision. Smart cities will be based on the following innovation axes: smart mobility, smart environment, smart people, smart living, and smart governance. In development terms, the core goal of a smart city is to become a business-competitive and attractive environment, while serving citizen well-being [CPG]. In a smart city, ICT plays a leading role and acts as the glue bringing together all actors, services, resources (and their interrelationships) that the urban environment is willing to host and provide [MVM]. ICN appears particularly suitable for these scenarios. Domains of interest include intelligent transportation systems, energy networks, health care, A/V communications, peer-to-
peer and collaborative platforms for citizens, social inclusion, active participation in public life, e-government, safety and security, and sensor networks. Clearly, this scenario has close ties to the vision of IoT, discussed in the previous section, as well as to vehicular networking. Nevertheless, the road to build a real information-centric digital ecosystem will be long, and more coordinated effort is required to drive innovation in this domain. We argue that smart-city needs and ICN technologies can trigger a virtuous innovation cycle toward future ICT platforms. Recent concrete ICN-based contributions have been formulated for home energy management [iHEMS], geo-localized services [ACC], smart-city services [IB], and traffic information dissemination in vehicular scenarios [RTIND]. Some of the proposed ICN-based solutions are implemented in real testbeds, while others are evaluated through simulation. Zhang et al. [iHEMS] propose a secure publish-subscribe architecture for handling the communication requirements of Home Energy Management Systems (HEMS). The objective is to safely and effectively collect measurement and status information from household elements, aggregate and analyze the data, and ultimately enable intelligent control decisions for actuation. They consider a simple experimental testbed for their proof-of-concept evaluation, exploiting open source code for the ICN implementation, and emulating some node functionality in order to facilitate system operation. A different scenario is considered in [ACC], where DHTs are employed for distributed, scalable, and geographically aware service lookup in a smart city. Also in this case, the ICN application is validated by considering a small-scale testbed: a small number of nodes are emulated with simple embedded PCs or specific hardware boards (e.g., for some sensor nodes); other nodes (which connect the principal actors of the tests) are emulated with workstations. The proposal in [IB] draws from a smart-city scenario (mainly oriented towards waste collection management) comprising sensors and moving vehicles, as well as a cloud-computing system that supports data retrieval and storage operations. The main aspects of this proposal are analyzed via simulation using open source code that is publicly available. Some software applications are designed on real systems (e.g., PCs and smartphones). With respect to evaluating ICN approaches in smart-city scenarios, it is necessary to consider generic metrics useful to track and monitor progress on services results and also for comparing localities between themselves and learn from the best [ISODIS]. In particular, it is possible to select a specific set of Key Performance Indicators (KPIs) for a given project in order to evaluate its success. These
KPIs may reflect the city's environmental and social goals, as well as its economic objectives, and they can be calculated at the global, regional, national, and local levels. Therefore, it is not possible to define a unique set of interesting metrics, but in the context of smart cities, the KPIs should be characterized with respect to the developed set of services offered by using the ICN paradigm. To sum up, smart-city scenarios aim to exercise several ICN aspects in an urban environment. In particular, they can be useful to (i) analyze the capacity of using ICN for managing extremely large data sets; (ii) study ICN performance in terms of scalability in distributed services; (iii) verify the feasibility of ICN in a very complex application like vehicular communication systems; and (iv) examine the possible drawbacks related to privacy and security issues in complex networked environments.