홈으로ArticlesAll Issue
ArticlesCost Modeling for Analyzing Network Performance of IoT Protocols in Blockchain-Based IoT
  • Minkyung Kim1, Kangseok Kim2,3 and Jai-Hoon Kim3,*

Human-centric Computing and Information Sciences volume 11, Article number: 07 (2021)
Cite this article 5 Accesses
https://doi.org/10.22967/HCIS.2021.11.007

Abstract

The Internet of Things (IoT) is a system of interrelated smart devices that share sensing data over networks. Heterogeneous data are collected, shared, and analyzed through the interconnection of massive physical devices. The analytical interaction between a series of IoT devices enables smarter decisions regarding the development of autonomous services. However, the scalability is essential to support the ever-growing number of connected computing devices. As the current IoT networks are centralized, the adoption of distributed peer-to-peer (P2P) networking could be a key solution to build successful decentralized blockchain-based IoT platforms. Therefore, this study attempts to characterize IoT networks in order to compare network performance under a variety of network conditions. In this paper, decentralized P2P computing is compared with the most commonly used centralized protocols, such as publish/subscribe and request/reply, within constrained IoT environments. First, cost modeling of network protocols is proposed to analyze the basic network performance. Moreover, the paper discusses a comparative analysis with push, pull, and P2P-based IoT data transmission approaches. The main purpose of the experiment is to identify the key factors that influence various IoT networks by comparing latency and availability generated by the request message. The results of the performance comparison show that the decentralized P2P-based approach has more key advantages than centralized networking in various performance metrics.


Keywords

Internet of Things, Peer-to-Peer, Publish/Subscribe, Request/Reply, Cost Modeling, Blockchain, IoT Collaboration


Introduction

The Internet of Things (IoT) consists of massive smart devices that are connected to the Internet for the purpose of sensing the physical world and sharing data in the digital world [13]. Real-time computational interaction with multiple things connected over networks is required to control automatically remote shared devices and systems. However, as the number of connected IoT devices grows, the amount of data and semantics generated by these devices is also growing constantly. In addition, IoT environments require lightweight sensing and communication protocols for resource-constrained devices. In a topic-based publish/subscribe manner, the data collected from physical devices are directed to central servers in order to support low bandwidth and high latency. In request/reply communications, all of the data are broadcast to servers in order to provide reliable security and allow resource observation [3]. The current IoT networks based on the centralized cloud need to be changed to reduce operational costs for the purposes of validation, computation, and storage. IoT networks must be re-designed to adopt distributed peer-to-peer (P2P) computing for processing the huge amounts of data [4]. P2P network is a distributed protocol for IoT networking in which there is no central server. The trustless P2P messaging protocols may provide a lightweight mechanism for interaction between devices on the IoT. A blockchain-based IoT is a key solution that allows untrusted nodes to send the data to distributed blockchain ledgers in P2P networking [5, 6]. The data can then be securely stored and shared in a blockchain framework to improve the security and availability of IoT data [7, 8].

Fig. 1. Overall architecture view of an IoT integrated collaboration and access system built on a blockchain-based distributed cloud.

The utilization of appropriate protocols for IoT collaboration according to the characteristics of various services and network conditions is crucial. Fig. 1 shows an overall view of the architecture of an IoT integrated collaboration and access system built on a blockchain-based distributed cloud. This system is a shared event collaboration model equipped with various IoT resources, in which a mashup framework integrates data collected from diverse network protocols, such as P2P, message queuing telemetry transport (MQTT) and constrained application protocol (CoAP), and the streaming server. It is composed of two major components: the IoT data management system and the mashup framework. The data management system operates so as to aggregate, process and analyze IoT data for extracting semantic data with an effective management scheme. The mashup framework shown in Fig. 1 is a basic structure for integrating collaboration data and activities, large-scale web contents, and sensing information obtainable from multiple sources. It also has the capability to maintain shared state consistency among connected things. Through the IoT mashup framework, heterogeneous data are integrated to generate semantic data and to create autonomous services. A key function of the IoT integrated collaboration and access system is to provide a generic solution for accessing and controlling the IoT, aggregating various IoT data, and maximizing the use of various collaborative capabilities by the collaborator. Another of its functions is to provide a structure for the development and deployment of data integration (mashup) applications that support the management of heterogeneous data. In addition, how to share and secure data effectively and efficiently has become one of the key concerns of the blockchain. Blockchain technologies are utilized to build a secure and reliable data mashup framework among multiple devices and applications. Based on the architecture of the IoT integrated collaboration system, the main contributions of the current research work are as follows:
This paper proposes cost analysis modeling IoT protocols to analyze the basic network performance of a decentralized P2P protocol and centralized publish/subscribe and request/reply protocols. Various performance metrics of IoT protocols, such as latency, usage, availability, and responsiveness, are analyzed.
Moreover, this paper introduces a comparative analysis of cost modeling based on push, pull, and P2P mechanisms. The comparative analysis of IoT protocols aims to identify key factors that influence network performance under various conditions.
The need for IoT-involved collaboration is presented in the following scenario. Due to the possibility of intermittent network disconnections when physical things or roaming users move from one place to another, they may be disconnected from collaboration for arbitrary periods of time. During disconnection, connected things may generate new objects on a shared application, or some objects on the application may be removed or transformed. Hence, disconnected things may have inconsistent data compared with those connected to the collaboration. Smart things can then be used to provide consistent data to disconnected things for synchronous collaboration. Intelligent customized services are created through interactions among sensing data, personal log data and activities, and large-scale web contents from multiple IoT sources over distributed networks.
The remainder of this paper is structured as follows: Section 2 reviews the related works; Section 3 presents the proposed cost analysis modeling on IoT network protocols; Section 4 presents the IoT data push, pull, and P2P transmission mechanisms and performance comparisons through a parameter analysis; and, finally, Section 5 presents the conclusion, future research perspectives, and additional improvements.


Related Work

This section presents an overview of various IoT technologies, protocols, and applications issues. The main elements, such as identification, sensing, interaction, computation, semantics, and services, are discussed to understand the functionality of the IoT [9–14]. Many previous studies have proposed ways of discovering, selecting, and composing the services provided by distributed physical objects. For example, an intelligent home-connected smart sensor for monitoring body conditions has been proposed to provide healthcare services needed in urgent situations [15]. In addition, Xu et al. [16] introduced an IoT-based system to collect, integrate, and interoperate IoT data to provide emergency medical services, while Park et al. [17] introduced a network architecture designed to effectively analyze data gathered from applications in a smart city environment. Furthermore, mashup technology is becoming a popular research technique for integrating web contents and providing novel integrated services to end users [18].
However, the issues of data consistency and availability caused by internet disconnection and malicious attacks may need to be considered. To guarantee data integrity and confidentiality, Kim et al. [19] proposed a video sharing scheme by storing hash values rather than actual video in the blockchain. Furthermore, as the number of connected computing devices grows, the operational costs of validation, computation, and storage may increase exponentially. Also, a novel lightweight secure message exchange protocol for low-power mobile edge devices has been proposed to achieve a higher degree of security while maintaining low computational costs [20]. To provide authentication and data integrity, Megouache et al. [21] introduces a security model used in a distributed, interoperable environment. In addition, Lee et al. [22] proposes the application of a distributed P2P IoT network architecture to the gateway of a traditional centralized architecture to prevent data forgery by utilizing blockchain technology. The use of blockchain technology in the IoT could be a scalable and secure solution for building a decentralized IoT. Therefore, this paper attempted to analyze cost modeling to characterize each network and to analyze the differences.
Next, the characteristics of network protocols that are commonly used in IoT environments are discussed. First, the P2P protocol enables decentralized collaboration by integrating heterogeneous networking services. P2P networking is a widely used distributed computing protocol for file sharing and highly parallel computing, among other things [23]. A peer can request data from other peers and simultaneously respond to requests from other peers. The distributed and decentralized aspects of the P2P protocol can help ensure high scalability and low server failure. One study proposed the integration of heterogeneous SNSs and making global relationships based on the integrated P2P system [24]. Moreover, a trustworthiness management method was proposed for P2P and social IoT [25].
IoT infrastructures require energy-efficient communication for sharing sensing data. Lightweight connectivity protocols, such as message queuing telemetry transport (MQTT) and Constrained Application Protocol (CoAP), are utilized to reduce battery power consumption, bandwidth usage, and communication latency. The MQTT protocol [26–29] is a topic-based publish/subscribe messaging protocol with a low message overhead and low battery power. When topic-based data are generated, the publisher node asynchronously publishes them to the broker. Then, the broker disseminates specific topic-based data to pre-registered subscribers. Furthermore, the request/reply protocol is a traditional client–server mechanism. When the client sends a query to devices, the response is sent back to the client via query processing. CoAP [30–36] is based on a request/reply interaction protocol between endpoints, allowing representational state transfer (REST) based communications based on HTTP methods, such as GET, POST, PUT, and DELETE. As a web application protocol, it supports tiny, low message overhead and low-power embedded sensors for resource-constrained and high-performance devices in the IoT.
In addition, various studies have compared the performance of lightweight protocols considering diverse network conditions. In a comparative study [37], MQTT was observed to have a better function than CoAP in persistent message transmission. By contrast, CoAP is appropriate for IoT environments requiring a low bandwidth and low resource usage. Moreover, research involving performance evaluations of MQTT and CoAP has been performed [389], the results of which show that CoAP is more reliable for message transmission when transferring a small volume of messages with a lower packet loss rate.


Cost Modeling on IoT Protocols

This study proposes cost modeling for analyzing the network performance of IoT protocols (such as P2P, publish/subscribe, and request/reply) which are used in the proposed blockchain-involved IoT collaboration environment. Based on previous analyses [39–43], this study extended the cost analysis models to analyze the network performance of various IoT protocols. A cost analysis of decentralized P2P protocols was also performed to compare centralized protocols, such as publish/subscribe and request/reply protocols. Moreover, the additional conditions of IoT environments were added for comparison. Table 1 shows the definition of the parameters. The following system parameters are governed by a Poisson process in the average inter-arrival time. The arrival time of the access request occurs as a Poisson process because the inter-arrival time of the request is an independent random variable with an exponential distribution with a pre-known average arrival rate.
Thus, the cost of the three different protocols (P2P, publish/subscribe, and request/reply) was analyzed in different situations in order to characterize each network and analyze the differences, as shown in Table 2. Fig. 2 shows diagrams of the three protocols to demonstrate the cost modeling clearly. Multiple network conditions were considered based on the basic aspects of the three protocols in IoT environments. Several models were considered as follows: (1) the conceptual total cost per unit time for each model; (2) the cost for each access by the client (or subscriber or client-side peer); (3) the average server’s service (processing) time per unit time interval considering the number of replicas (sever-side peers); (4) the waiting time in the server (based on queueing theory, M/D/1, and M/D/c); (5) the time delay between the client’s (or subscriber’s or client-side peer’s) intention and access; (6) the time delay between the occurrence of the event and the notification (or recognition) of the event to the client (or subscriber or client-side peer); (7) the hit ratio of the available data; (8) the availability of the server (or server-side peer); and (9) the transfer delay when considering server (or server-side peer) error.

Table 1. Definition of parameters

Parameter Definition
α The event generation rate of the provider (or publisher or server-side peer).
β The client’s (or subscriber’s or client-side peer’s) event access rate of uploaded (or published) events.
CP2P The P2P communication cost per generated event.
n The average number of clients (or subscribers or client-side peers) per generated event.
nreplica The average number of replicas (server-side peers).
Cpub The publish cost per event (publishers publish events to server).
Csub The subscribe cost per event (subscribers subscribe events from the server).
s(n) The sharing effect among n nodes (between 1/n and 1).
Crr The request and reply communication cost per generated event.
λ The failure event rate of the communication link (governed by a Poisson process with the average inter-arrival time of 1⁄λ).
μ The recovery (reconnection) event rate of the communication link (governed by a Poisson process with the average inter-arrival time of 1⁄μ).
tP2P The time delay for P2P communication per generated event.
tpub The time delay for publishing events to the server.
tsub The time delay for relaying events to the subscriber from the server.
trr The time delay for request and reply communication per generated event.

Table 2. Cost of selected models
Models Peer-to-Peer (P2P) Publish/Subscribe Request/Reply
(1) Conceptual total cost per unit time βnCP2P α(Cpub+ns(n)Csub) βnCrr
(2) Cost of each access by client (or subscriber or client-side peer) CP2P (α/β)(Cpub/n+Csub) Crr
(3) Average server’s service (processing) time per unit time interval considering the number of replicas (server-side peers) βn αns(n) βn
βn/nreplica
(4) Waiting time in the server (based on the queueing theory, M/D/1, and M/D/c) 1+(β/(2(1−β))) 1+(α/(2(1−α))) 1+(β/(2(1−β)))
≅ 1+((P(L(∞) ≥ nreplica)/(nreplica−β))
(5) Time delay between the client’s (or subscriber’s or client-side peer’s) intention and access tP2P 0 trr
tP2P/nreplica
(6) Time delay between event occurrence and notification (or recognition) of the event to the client (or subscriber or client-side peer) ((1/β)+3tP2P)/2 (tpub+tsub)/2 ((1/β)+trr)/2
((1/β)+((3tP2P)/nreplica))/2
(7) Hit ratio of available data β 1 β
β
(8) Availability of server (or server-side peer) μ/(λ+μ) μ/(λ+μ) μ/(λ+μ)
1−(λ/(λ+μ))nreplica
(9) Transfer delay as server (or server-side peer) error occurs (μ/(λ+μ))tP2P+(λ/(λ+μ))  ((1/μ)+tP2P)) (μ/(λ+μ))  (tpub+tsub) (μ/(λ+μ))trr+(λ/(λ+μ))
{1−(λ/(λ+μ))nreplica}tP2P+(λ/(λ+μ))nreplica)(1/nreplicaμ)+tP2P) +(λ/(λ+μ))((1/μ)+tpub+tsub) ((1/μ)+trr)
First, the conceptual total cost per unit time for each protocol was analyzed, as shown in (1) of Table 2. The conceptual total cost per unit time was formulated after receiving the client-side peer’s request. In P2P networks, this study assumes that CP2P is the P2P communication cost, and n is the average number of peers per data. When the request rate is β, the conceptual total cost per unit time in P2P networks is βnCP2P. A publish/subscribe system consists of publishers, a server (or broker), and subscribers. Cpub is the cost when events are published to a server by publishers, and Csub is the cost when subscribers subscribe events from a server. Moreover, n is the average number of subscribers and s(n) is the sharing effect among n nodes. Because a server disseminates events to subscribers, the sharing effect can be calculated ((1⁄n)≤s(n)≤1). When the publish rate is α, the cost of the publish/subscribe protocol is α(Cpub+ns(n)Csub). The process of the request/reply protocol is similar to the cost analysis method in the P2P protocol, as shown in the protocol diagram of Fig. 2. The cost for request/reply networks is assumed to be Crr. Thus, when the request rate is β and n is the number of clients, the total cost is βnCrr.
Fig. 2. Diagrams of the three protocols: P2P, publish/subscribe, and request/reply.

In addition, (2) is the cost of each access by a client (or subscriber or client-side peer). CP2P is the average cost to access the data of a server-side peer in P2P communication. In a publish/subscribe protocol, α/β is the average number of events that occur before each access. To calculate the subscriber’s access cost, the average cost to publish events to a server is divided among n subscribers and Csub is added for each subscriber. Then, the average cost for each access is (α/β)((Cpub/n)+Csub). Similar to P2P networking, the cost of request/reply networking to download the data is Crr.
The average server’s service (processing) time per unit time interval, considering the number of replicas (server-side peers), is formulated in (3). It is assumed that βn is the average processing time by applying the client’s request rate and the number of client-side peers. Moreover, the case of server redundancy was added in P2P networking. For example, BitTorrent is a common P2P protocol that is used to transfer large amounts of data from redundant server-side peers to a massive number of client-side peers over the Internet [44]. The number of redundant servers will reduce the average processing time, such that βn/nreplica. The network performance tends to improve as the number of server-side peers who have the same data in networks increases. In the publish/subscribe protocol, every generated event is transferred to the subscribers. Thus, the average processing time is calculated as αns(n) by applying the event generation rate and sharing the effects of the number of subscribers. The average processing time of the request/reply mechanism is equal to βn.
(4) is the waiting time in the server. Here, it is assumed that the calculation of the waiting time in a server redundancy case is based on the queueing theory, M/D/1, and M/D/c with the Poisson arrival process.
In (5), the time delay between a client’s (or subscriber’s or client-side peer’s) intention and access will be generated while a client recognizes the generated event and then accesses the data. In the P2P protocol, the time delay between intention and access is tP2P. In the case of server redundancy, the time delay between intention and access from multiple providers to a client is divided by nreplica. Therefore, increasing the number of redundant servers can reduce the time delay. However, publish/subscribe networking has no time delay for access after the subscriber’s intention, because the events have already been delivered to a server. Moreover, the time delay of request/reply networking for access after a client’s intention is trr.
In (6), the time delay between event occurrence and notification to a client (or subscriber or client-side peer) is formulated. The overall time delay of the P2P protocol between the client-side peer’s intention and access is tP2P, and the longest time delay is supposed as 3tP2P. The time delay is determined depending on the distance between a client-side peer’s request and the generated data. Moreover, it is assumed that 1/β is an additional time delay for accessing data by any digital device. These time delays will be generated on the client-side. The time delay between event occurrence and notification in a peer is ((1/β)+3tP2P)/2. A case of server redundancy is ((1/β)+(3tP2P/nreplica))/2. Moreover, the time delay of the publish/subscribe protocol between event occurrence and recognition is (tpub+tsub)/2. Similarly, ((1/β)+trr)/2 is supposed in the case of the request/reply protocol.
(7) is the hit ratio of available data. This means that the cost is basically required to use the generated data depending on the client’s request. It can be used, unless there are no generated data. The hit ratio of available data for the P2P and request/reply protocols is β. Because all of the data are published to a server in publish/subscribe networking, the generated data are available at any time. Therefore, the hit ratio of available data is 1.
Moreover, the availability of a server (or server-side peer) will be calculated for downloading data from a server (or server-side peer) in (8). Unless there are no generated data or there are no available data in any available server (or peers), the three networks can be available. Availability supposed by the rate of failure (λ) and the rate of recovery (μ) is μ/(λ+μ). In the server redundancy of P2P, the number of servers is added as 1−(λ/(λ+μ))nreplica.
The transfer delay caused by a server (or server-side peer) error is formulated in (9). If there is a server error, then the delay time for recovery is calculated. Moreover, the failure probability can be calculated with the recovery rate and delay time. In P2P networking, (μ/(λ+μ))tP2P is calculated as the delay time for downloading. In addition, the failure probability from the separated server is supposed by the average number of servers with the same uploaded event as (λ/(λ+μ))((1/μ)+tP2P)). Similarly, the transfer delay of the publish/subscribe protocol caused by a server error is calculated as (μ/(λ+μ))(tpub+tsub)+ (λ/(λ+μ))((1/μ)+tpub+tsub). Furthermore, the transfer delay caused by a server error in request/reply networking is (μ/(λ+μ))trr+(λ/(λ+μ))((1/μ)+trr). In the server redundancy of P2P, the number of servers are added as {1−(λ/(λ+μ))nreplica}tP2P+(λ/(λ+μ))nreplica)(1/nreplicaμ)+tP2P).


Performance Comparison by the Parameter Analysis

This section describes the comparative analysis of cost modeling for IoT protocols. This paper focused on IoT data push-and-pull transmission approaches with the IoT messaging protocols for IoT collaboration. During collaboration among users, the data generated from the IoT can be disseminated to users according to the three types of data transmission approaches, i.e., the P2P, push-based (publish/subscribe) and pull-based (request/reply) approaches. The communication channel (event message broker or messaging middleware based on publish/subscribe) enables one type of collaborative application to exchange events with other types of collaborative applications.
The shared event collaboration model for IoT collaboration with various IoT resources is shown in Fig. 3, where the mashup framework integrates the data that can be accessed by users and facilitates seamless communication between smart things and collaborative applications. The integration architecture of the IoT collaboration has already been described in previous works [45, 46]. The IoT mashup framework in Fig. 3 shows that the clients share the states that are maintained between the clients by transmitting changes through the event messaging middleware. In IoT-integrated collaboration, collaborators can monitor or detect the location and status of others in real time using various sensors. IoT-integrated collaboration services can then be improved through seamless communication among collaborators and prevent the state of inconsistency caused by the disconnected situations of roaming users. For the decentralized IoT, devices and applications, as IoT data providers, may need to directly communicate with one another. After transferring data or information into the blockchain-based distributed cloud, disseminating data from the cloud to applications may be wasteful and slower. Therefore, the P2P approach may allow better interaction among IoT applications.

Fig. 3. Shared event collaboration model in push, pull and peer-to-peer based data dissemination approaches.


Table 3. Values of parameters for analysis
Parameter Value
α 0.5
β 0.5
CP2P 0.5
n 1–1000
nreplica 1–1000
Cpub 1
Csub 1
s(n) 1
Crr 2
λ 0.1
μ 0.5
tP2P 0.75
tpub 1
tsub 1
trr 1
Fig. 4. Conceptual total cost per unit time.
Fig. 5. Cost of each access by client (or subscriber or client-side peer).

Table 3 shows the values of the system parameters by referencing our previous cost analysis [39–43]. The capability of data processing is shown in Figs. 4 and 5, which show the graphs generated by our analysis of the IoT protocols. The decentralized P2P-based approach has a key advantage in that it can immediately and directly be communicated to all shared peer objects. Moreover, Fig. 6 shows that the average server’s service (processing) time per unit time interval, depending on the number of replicas (server-side peers), can cost relatively less than other approaches according to the increase of redundant servers in P2P networking. The reason for this is that the duration time for transferring can be reduced by increasing the number of server-side peers. In P2P networking, the waiting time can also be reduced as the number of servers increases, as shown in Fig. 7. However, it has one obvious disadvantage in that it may be rather difficult to customize data transmission among all peer collaborators.
Fig. 6. Average server’s service (processing) time per unit time interval depending on the number of replicas (server-side peers).
Fig. 7. Waiting time in the server (based on the queueing theory, M/D/1, and M/D/c).

In a request/response paradigm (pull-based approach), data can be disseminated depending on the requests of users. It is cost-effective to use the request/reply message protocol in the IoT with constrained resources. However, as shown in the first graph of Fig. 8, recognizing and accessing the desired data cause more time delays in the pull-based data dissemination approach. Accessing data in the publish/subscribe paradigm (push-based approach) causes no time delay because a broker may already have the data to publish. Moreover, if each request is associated with the deadline in time, and of the collaborative application needs multiple requests as well, then the pull-based approach has a disadvantage when compared with the push-based approach because the application has to retrieve the data before the deadline of the respective request. The pull-based approach may result in a state of inconsistency. Thus, the processing time for critical data in multiple pull request cases necessitates a mechanism to ensure consistency among collaborators. In the push-based data dissemination approach, data collected from various sensors are disseminated to collaborators through the mashup framework and the communication channel (a broker for publish/subscribe). The latest version of the data is periodically updated in the database in the push-based data dissemination approach. In the second graph, the publish/subscribe protocol has a shorter communication delay between event occurrence and notification due to the concurrent dissemination of all the data to the users or devices being subscribed from a broker, when compared with other protocols. Moreover, as shown in the third graph of Fig. 8, the hit ratio of available data for the publish/subscribe protocol is higher than that of the other protocols due to the increase in data availability according to connection-oriented things. The fourth graph shows that the cost of availability is similar in those approaches when only the failure and recovery rate are considered. The cost of availability in the P2P approach involving redundant servers can rise according to the number of server-side peers with the same data. The fifth graph shows that the P2P protocol is more tolerable than other approaches in a transfer delay caused by a server error. Moreover, the P2P approach with redundant servers requires a shorter time delay when transferring data to a client-side peer because the probability of delay is considered by increasing the number of server-side peers. The publish/subscribe approach has a much longer transfer delay because it takes time to connect to a broker after the recovery of a server, as shown in the last graph of Fig. 8. Several tradeoffs exist among these approaches in the process of disseminating IoT data to collaborators in synchronous or asynchronous situations for IoT integrated collaboration. Clearly, the proposed comparative analysis has limitations in estimating network performance in blockchain-based IoT environments because it relies on simulations through our cost analysis modeling. However, this analysis can be used to formulate the impact of network performance when building an IoT blockchain platform.
Fig. 8. Comparative analysis of the models.


Conclusion

The IoT enables wireless and mobile technologies, data processing, and analytics to physical objects. IoT collaboration systems can monitor, share, and adjust the interactions between connected things over the Internet. With the ever growing number of devices deployed in the IoT, applying blockchain technologies to IoT devices makes it possible to solve the problems associated with scalability and reliability. The P2P features of blockchain technologies allow scalable and secure communications. This paper presents cost analysis modeling for IoT protocols and comparisons of network performance through an analysis of the parameters of the IoT-integrated collaboration and access system built on a blockchain-based distributed cloud. The comparison of the IoT networks aims to determine the key factors that influence network performance under various conditions. Moreover, a comparative analysis of cost modeling is introduced based on push, pull, and P2P mechanisms. The results show that the decentralized P2P-based approach outperforms centralized networking in a variety of performance metrics including latency and availability.
To build an IoT blockchain platform, the cost modeling proposed in this study could be used to formulate the impact of network performance. Future works will need to extend the research with specific IoT service environments in order to reflect the network performance characteristics through system implementation. For example, an efficient monitoring method based on the proposed cost analysis will be required to track and analyze networks. Moreover, it will be necessary to examine the current research with network applications and devices in real time so as to tolerate any failures, such as disconnected communication and inconsistency. In addition, it will be necessary to consider such issues as data security and personal privacy preservation with the proposed IoT collaboration system built on a blockchain-based distributed cloud.


Related Work

Acknowledgements


Author’s Contributions

All of the authors have contributed to the cost analysis modeling of IoT protocols and approved the final manuscript.


Funding

This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT) (No. NRF-2018R1D1A1B07040573 and No. NRF-2019R1F1A1059036).


Competing Interests

The authors declare that they have no competing interests.


References

[1] E. Borgia, “The Internet of Things vision: key features, applications and open issues,” Computer Communications, vol. 54, pp. 1-31, 2004.
[2] Z. Ma, M. Xiao, Y. Xiao, Z. Pang, H. V. Poor, and B. Vucetic, “High-reliability and low-latency wireless communication for Internet of Things: challenges, fundamentals, and enabling technologies,” IEEE Internet of Things Journal, vol. 6, no. 5, pp. 7946-7970, 2019.
[3] K. Gafurov and T. M. Chung, “Comprehensive survey on internet of things, architecture, security aspects, applications, related technologies, economic perspective, and future directions,” Journal of Information Processing Systems, vol. 15, no. 4, pp. 797-819, 2019.
[4] V. Pureswaran and P. Brody, “Device democracy: saving the future of the Internet of Things,” IBM Institute for Business Value, 2015 [Online]. Available: https://www.ibm.com/downloads/cas/Y5ONA8EV.
[5] S. Rahmadika, S. Noh, K. Lee, B. J. Kweka, and K. H. Rhee, “The dilemma of parameterizing propagation time in blockchain P2P network,” Journal of Information Processing Systems, vo. 16, no. 3, pp. 699-717, 2020.
[6] L. Tseng, L. Wong, S. Otoum, M. Aloqaily, and J. B. Othman, “Blockchain for managing heterogeneous internet of things: a perspective architecture,” IEEE Network, vol. 34, no. 1, pp. 16-23, 2020.
[7] L. Cui, S. Yang, Z. Chen, Y. Pan, Z. Ming, and M. Xu, “A decentralized and trusted edge computing platform for Internet of Things,” IEEE Internet of Things Journal, vol. 7, no. 5, pp. 3910-3922, 2019.
[8] O. Novo, “Blockchain meets IoT: an architecture for scalable access management in IoT,” IEEE Internet of Things Journal, vol. 5, no. 2, pp. 1184-1195, 2018.
[9] A. Al-Fuqaha, M. Guizani, M. Mohammadi, M. Aledhari, and M. Ayyash, “Internet of things: a survey on enabling technologies, protocols, and applications,” IEEE Communications Surveys & Tutorials, vol. 17, no. 4, pp. 2347-2376, 2015.
[10] A. Al-Fuqaha, A. Khreishah, M. Guizani, A. Rayes, and M. Mohammadi, “Toward better horizontal integration among IoT services,” IEEE Communications Magazine, vol. 53, no. 9, pp. 72-79, 2015.
[11] C. Sarkar, A. U. Nambi, R. V. Prasad, A. Rahim, R. Neisse, and G. Baldini, “DIAT: a scalable distributed architecture for IoT,” IEEE Internet of Things Journal, vol. 2, no. 3, pp. 230-239, 2014.
[12] S. S. Gill, S. Tuli, M. Xu, I. Singh, K. V. Singh, D. Lindsay, et al., “Transformative effects of IoT, blockchain and artificial Intelligence on cloud computing: evolution, vision, trends and open challenges,” Internet of Things, vol. 8, article no. 100118, 2019. https://doi.org/10.1016/j.iot.2019.100118
[13] J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, “Internet of Things (IoT): a vision, architectural elements, and future directions,” Future Generation Computer Systems, vol. 29, no. 7, pp. 1645-1660, 2013.
[14] J. H. Nord, A. Koohang, and J. Paliszkiewicz, “The Internet of Things: review and theoretical framework,” Expert Systems with Applications, vol. 133, pp. 97-108, 2019.
[15] G. Yang, L. Xie, M. Mantysalo, X. Zhou, Z. Pang, Z., L. D. Xu, S. Kao-Walter, Q. Chen, and L. R. Zheng, “A health-IoT platform based on the integration of intelligent packaging, unobtrusive bio-sensor, and intelligent medicine box,” IEEE Transactions on Industrial Informatics, vol. 10, no. 4, pp. 2180-2191, 2014.
[16] B. Xu, L. D. Xu, H. Cai, C. Xie, J. Hu, and F. Bu, “Ubiquitous data accessing method in IoT-based information system for emergency medical services,” IEEE Transactions on Industrial informatics, vol. 10, no. 2, pp. 1578-1586, 2014.
[17] J. H. Park, M. M. Salim, J. H. Jo, J. C. S. Sicato, S. Rathore, and J. H. Park, “CIoT-Net: a scalable cognitive IoT based smart city network architecture,” Human-centric Computing and Information Sciences, vol. 9, article no. 29, 2019. https://doi.org/10.1186/s13673-019-0190-9
[18] H. Han, Y. Xue, and K. Oyama, “Mashup technology: beyond open programming interfaces,” IEEE Annals of the History of Computing, vol. 46, no. 12, pp. 96-99, 2013.
[19] T. Kim, I. Y. Jung, and Y. C. Hu, “Automatic, location-privacy preserving dashcam video sharing using blockchain and deep learning,” Human-centric Computing and Information Sciences, vol. 10, article no. 36, 2020. https://doi.org/10.1186/s13673-020-00244-8
[20] F. Concone, G. L. Re, and M. Morana, “SMCP: a secure mobile crowdsensing protocol for fog-based applications,” Human-centric Computing and Information Sciences, vol. 10, article no. 28, 2020. https://doi.org/10.1186/s13673-020-00232-y
[21] L. Megouache, A. Zitouni, and M. Djoudi, “Ensuring user authentication and data integrity in multi-cloud environment,” Human-centric Computing and Information Sciences, vol. 10, article no. 15, 2020. https://doi.org/10.1186/s13673-020-00224-y
[22] Y. Lee, S. Rathore, J. H. Park, and J. H. Park, “A blockchain-based smart home gateway architecture for preventing data forgery,” Human-centric Computing and Information Sciences, vol. 10, article no. 9, 2020. https://doi.org/10.1186/s13673-020-0214-5
[23] D. Talia and P. Trunfio, “Toward a synergy between P2P and grids,” IEEE Internet Computing, vol. 7, no. 4, pp. 96-95, 2003.
[24] P. Lin, P. C. Chung, and Y. Fang, “P2P-ISN: a peer-to-peer architecture for heterogeneous social networks,” IEEE Network, vol. 28, no. 1, pp. 56-64, 2014.
[25] M. Nitti, R. Girau, and L. Atzori, “Trustworthiness management in the social Internet of Things,” IEEE Transactions on Knowledge and Data Engineering, vol. 26, no. 5, pp. 1253-1266, 2014.
[26] S. Wagle, “Semantic data extraction over MQTT for IoT centric wireless sensor networks,” in Proceedings of 2016 International Conference on Internet of Things and Applications (IOTA), Pune, India, 2016, pp. 227-232.
[27] J. E. Luzuriaga, J. C. Cano, C. Calafate, P. Manzoni, M. Perez, and P. Boronat, “Handling mobility in IoT applications using the MQTT protocol,” in Proceedings of 2015 Internet Technologies and Applications (ITA), Wrexham, UK, 2015, pp. 245-250.
[28] P. Fremantle, B. Aziz, J. Kopecky, and P. Scott, “Federated identity and access management for the internet of things,” in Proceedings of 2014 International Workshop on Secure Internet of Things, Wroclaw, Poland, 2014, pp. 10-17.
[29] A. P. Haripriya and K. Kulothungan, “Secure-MQTT: an efficient fuzzy logic-based approach to detect DoS attack in MQTT protocol for Internet of Things,” EURASIP Journal on Wireless Communications and Networking, vol. 2019, article no. 90, 2019. https://doi.org/10.1186/s13638-019-1402-8
[30] A. P. Castellani, M. Gheda, N. Bui, M. Rossi, and M. Zorzi, “Web services for the Internet of Things through CoAP and EXI,” in Proceedings of 2011 IEEE International Conference on Communications Workshops (ICC), Kyoto, Japan, 2011, pp. 1-6.
[31] T. A. Alghamdi, A. Lasebae, and M. Aiash, “Security analysis of the constrained application protocol in the Internet of Things,” in Proceedings of the 2nd International Conference on Future Generation Communication Technologies (FGCT), London, UK, 2013, pp. 163-168.
[32] M. Kovatsch, M. Lanter, and Z. Shelby, “Californium: scalable cloud services for the Internet of Things with CoAP,” in Proceedings of 2014 International Conference on the Internet of Things (IOT), Cambridge, MA, 2014, pp. 1-6.
[33] C. Bormann, A. P. Castellani, and Z. Shelby, “CoAP: an application protocol for billions of tiny internet nodes,” IEEE Internet Computing, vol. 16, no. 2, pp. 62-67, 2012.
[34] B. C. Villaverde, D. Pesch, R. D. P. Alberola, S. Fedor, and M. Boubekeur, “Constrained application protocol for low power embedded networks: a survey,” in Proceedings of 2012 6th international Conference on Innovative Mobile and Internet Services in Ubiquitous Computing, Palermo, Italy, 2012, pp. 702-707.
[35] L. Rodrigues, J. Guerreiro, and N. Correia, “RELOAD/CoAP architecture with resource aggregation/disaggregation service,” in Proceedings of 2016 IEEE 27th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC), Valencia, Spain, 2016, pp. 1-6.
[36] R. H. Randhawa, A. Hameed, and A. N. Mian, “Energy efficient cross-layer approach for object security of CoAP for IoT devices,” Ad Hoc Networks, vol. 92, article no. 101761, 2019. https://doi.org/10.1016/j.adhoc.2018.09.006
[37] N. De Caro, W. Colitti, K. Steenhaut, G. Mangino, and G. Reali, “Comparison of two lightweight protocols for smartphone-based sensing,” in Proceedings of 2013 IEEE 20th Symposium on Communications and Vehicular Technology in the Benelux (SCVT), Namur, Belgium, 2013, pp. 1-6.
[38] D. Thangavel, X. Ma, A. Valera, H. X. Tan, and C. K. Y. Tan, “Performance evaluation of MQTT and CoAP via a common middleware,” in Proceedings of 2014 IEEE 9th International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP), Singapore, 2014, pp. 1-6.
[39] S. Oh, S. L. Pallickara, S. Ko, J. H. Kim, and G. Fox, “Publish/subscribe systems on node and link error prone mobile environments,” in Computational Science – ICCS 2005. Heidelberg, Germany: Springer, 2005, pp. 576-584.
[40] S. Oh, S. L. Pallickara, S. Ko, J. H. Kim, and G. Fox, “Cost model and adaptive scheme for publish/subscribe systems on mobile grid environments,” in Computational Science – ICCS 2005. Heidelberg, Germany: Springer, 2005, pp. 275-278.
[41] S. Oh, J. H. Kim, and G. Fox, “Real-time performance analysis for publish/subscribe systems,” Future Generation Computer Systems, vol. 26, no. 3, pp. 318-323, 2010.
[42] M. Kim, J. H. Kim, and K. Kim, “comparative analysis of cost modeling on lightweight messaging protocols for IoT integrated ubiquitous collaboration system,” International Journal of Applied Engineering Research, vol. 12, no. 23, pp. 13314-13320, 2017.
[43] M. Kim, “Decentralized data management schemes for IoT blockchain,” Ph.D. dissertation, Department of Computer Engineering, Graduate School of Ajou University, Suwon, Korea, 2019.
[44] M. A. D’Alessandro Costa and M. Goncalves Rubinstein, “Performance analysis of a locality-aware BitTorrent protocol in enterprise networks,” Peer-to-Peer Networking and Applications, vol. 12, pp. 751-762, 2019.
[45] K. Kim and G. C. Fox, “Modeling, simulation, and practice of floor control for synchronous and ubiquitous collaboration,” Multimedia Tools and Applications, vol. 53, no. 1, pp. 213-236, 2011.
[46] K. Kim, W. Ha, O. Choi, H. Yeh, J. H. Kim, M. Hong, and T. Shon, “An interactive pervasive whiteboard based on MVC architecture for ubiquitous collaboration,” Multimedia Tools and Applications, vol. 74, no. 5, pp. 1557-1576, 2015.

About this article
Cite this article

Minkyung Kim1, Kangseok Kim2,3 and Jai-Hoon Kim3,*, Cost Modeling for Analyzing Network Performance of IoT Protocols in Blockchain-Based IoT, Article number: 11:07 (2021) Cite this article 5 Accesses

Download citation
  • Recived12 September 2020
  • Accepted30 December 2020
  • Published15 February 2021
Share this article

Anyone you share the following link with will be able to read this content:

Provided by the Springer Nature SharedIt content-sharing initiative

Keywords