An Efficient Algorithm for Media-based

Surveillance System (EAMSuS)

in IoT Smart City Framework

Vasileios A. Memos, Kostas E. Psannis, Byung‐Gyu Kim, and B.B Gupta

Abstract

Internet of Things (IoT) is the new technological revolution that aspires to connect all the everyday

physical objects to the Internet, making a huge global network of uniquely things which can share

information amongst each other and complete scheduled tasks, bringing significant benefits to users

and companies of a Smart City (SC). A Smart City represents a new future framework, which

integrates multiple information and communication technology (ICT) and Internet of Things (IoT)

solutions, so as to improve the quality life of its citizens. However, there are many security and privacy

issues which must be taken into account before the official launching of this new technological

concept. Many methods which focus on media security of wireless sensor networks have been

proposed and can be adopted in the new expandable network of IoT. In this paper, we describe the

upcoming IoT network architecture and its security challenges and analyze the most important

researches on media security and privacy in wireless sensor networks (WSNs). Subsequently, we

propose an Efficient Algorithm for Media-based Surveillance System (EAMSuS) in IoT network for

Smart City Framework, which merges two algorithms introduced by other researchers for WSN packet

routing and security, while it reclaims the new media compression standard, High Efficiency Video

Coding (HEVC). Experimental analysis shows the efficacy of our proposed scheme in terms of users’

privacy, media security, and sensor node memory requirements. This scheme can be successfully

integrated into the IoT network of the upcoming Smart City concept.

Keywords Encryption Algorithm, HEVC, IoT, Smart City, Video Surveillance Systems, WSN.

1 Introduction

Last decade, Internet of Things (IoT) came gradually into our daily routine, thanks to the evolution of

the wireless communication technologies, such as RFID, WiFi, 4G, IEEE 802.15.x etc [1]. IoT is a

computing concept that envisages a future where all the everyday physical objects (the ―things‖) will

be linked to the internet and be uniquely identifiable and ubiquitously connected amongst each other,

so as to communicate and exchange information and data. Internet specialists assume that the IoT

network will consist of about 50 billion objects by 2020 [2]. By using IoT, everything around us, such

as computers, mobile phones, TVs, cars, etc, will collect all the necessary information, and then they

will send them to the concerned devices, thanks to wireless technologies, which will automatically

accomplish the proper actions [3].

The main goal of IoT is to establish advanced connectivity of systems, devices and services that exceed

machine-to-machine (M2M) communications, supporting various protocols, domains and applications

[21]. The interconnection of all these "things" will contribute in process automation and will support

advanced applications such as Smart Grid and Smart Surveillance.

The term "surveillance" includes all the ways which are used to monitor the behavior, activities, or

other changing information for human influencing, managing, directing, or protecting purposes. Thus,

the term covers remote monitoring by using various electronic devices (such as CCTV cameras) or

interception of electronically transmitted stream (e.g. internet traffic, phone calls etc). The term ―smart

surveillance‖ denotes the ability of real-time video analyzing in relevant surveillance applications. The

main advantage of video surveillance systems use is the application of video compression technology,

which can multiplex effectively or store images from a large number of cameras onto mass store

devices (e.g. video tapes, discs) [22]. Surveillance systems convert video surveillance from a data

acquisition and analysis tool to information and intelligence acquisition systems [23]. By using real-

time video analysis, surveillance systems can respond to an activity in real time, gathering the

important information at much higher resolution [22].

Fig. 1. The Future Smart City Project.

It is fact that Wireless Sensor Networks (WSNs) tend to be integrated into the whole IoT network.

WSNs are ad hoc networks which consist of many small sensor nodes with limited computing

resources and one or more base stations. In most cases, sensor nodes have a processing unit with

limited computational capacity and power [14]. WSNs are developed in specific environments, such as

glaciers, forests, and mountains, so as to gather environmental quantitative parameters during long

periods. Temperature, moisture and light sensor readings are some of the factors which allow the

analysis of the environmental phenomena, such as the climate changes and their effects [13].

There are many benefits of integrating WSN into IoT. However, this integrated scheme carries several

security challenges [42], such as privacy issues [43], which must be considered in order to be regarded

as secure. Several schemes and algorithms have been proposed so as to improve the security level of

IoT and WSN infrastructures separately. In this paper we focus on the WSN security mechanisms so as

to make them applicable into the IoT, and thus to improve the general IoT security.

Our scheme makes use of High Efficiency Video Coding (HEVC) compression standard in the video

which will be captured from the sensors (e.g. a micro-cam) surround the world and be sent to the user’s

smart surveillance system, such as a tablet, a smartphone etc. at his location, such as his home [45], in

order to provide the suitable information about a fact. HEVC (or H.265) is the new video compression

format and is the successor of H.264. It was formalized on the 25th of November 2013 and published

as ISO/IEC 23008-2:2013. An open source HEVC decoder and encoder (x265) has been developed and

is widely adopted. There are many advantages by using HEVC standard instead of H.264/MPEG-AVC

standard or more outdated compression standards, because the fact that HEVC presents high quality

multimedia streaming, even on low-bandwidth networks, as it consumes about half bandwidth less than

its predecessor H.264 [15].

All the above technologies will be adapted to the new future challenge, called ―Smart City‖. A "Smart

City" (SC) is a future concept of an urban environment which integrates multiple information and

communication technology (ICT) and Internet of Things (IoT) services so as to manage a city's assets,

such as local departments, schools, hospitals, transportation systems, law enforcement and other

innovative community services [26]. Information and communications technology (ICT) is an extended

term for information technology (IT) which stresses the role of unified communications and the

integration of telecommunications, computers as well as necessary enterprise software, middleware,

storage, and audio-visual systems, which enable users to access, store, transmit, and manipulate

information [49].

There are many advantages of developing a smart city, as it enhances the overall quality of life of its

citizens [27]. A "Smart City" is a city well performing in 6 characteristics, built on the "smart"

combination of endowments and activities of self-decisive, independent and aware citizens. The

European Commission promoted ―the smart city‖ calls regarding energy efficiency, renewable energy

and green mobility for the large urban cities [28]. Citizens of a smart city will be surrounded by an

environment that rapidly evolves, leading them to demand community-centric experiences with better

quality of experience (QoE) [41]. Figure 1 shows the future smart city project which will interconnect

various innovative technologies so as to contribute to many areas, such as Economy, Business,

Environment, Agriculture, Mobility, Education, Governance, Retail, Communication, Buildings, etc,

with main goal: the improvement of citizens’ life quality [51]. There have been developed

corresponding security frameworks for each of these areas, such as for example the security framework

for business clouds [52].

In literature, various definitions of "Smart City" can be found, such as:

- ―A city connecting the physical infrastructure, the ICT infrastructure, the social infrastructure, and the

business infrastructure to leverage the collective intelligence of the city‖ [29].

- ―A city that invests in human and social capital and traditional and modern (ICT) communication

infrastructure in order to sustain the economic growth and a high quality of life, with a wise

management of natural resources, through participatory governance‖ [30].

- ―A city whose community has learned to learn, adapt and innovate. People need to be able to use the

technology in order to benefit from it‖ [31].

- ―A city that reflects a particular idea of local community, one where city governments, enterprises

and residents use ICTs to reinvent and reinforce the community’s role in the new service economy,

create jobs locally and improve the quality of community life‖ [32].

The rapid growth of ―Internet of Things‖ in combination with the quality of service and the privacy

concerns of its users reinforced our motivation for this research. Specifically, the main goal of the

paper is to contribute to the efficient real-time video transmission from the sensor nodes – with the

desirable gathered information from the environment – to citizens of the smart city and their devices

(tablets, smartphone etc), by applying improved encryption algorithm so as to ensure confidentiality

and security of the transmission. In addition, our proposed algorithm contributes to the reduction of the

memory consumption in the sensor nodes, while it also ensures a faster routing of the media sharing

among the users of the smart city. Therefore, our proposed scheme improves the current state-of-the-art

architecture.

The rest of this paper is organized as follows: Section 2 presents the related work that has been

conducted in IoT and WSNs. Section 3 describes the security architecture of IoT and the probable

integration of WSNs in IoT infrastructure. In Section 4 we present our proposed algorithm, while

Section 5 includes the experimental results of our scheme. Section 6 identifies conclusions and future

work.

2 Related Work

The future Internet, promoted as ―Internet of Things‖, will be a world-wide network which will consist

of a significant number of interconnected things. Many researches have been conducted on the IoT

model. Security, privacy and trust in IoT are analyzed by S. Sicari et al. in [12], while a comprehensive

critical analysis on the security concerns of IoT is conducted by M.U. Farooq [3] and it presents the

architecture of IoT with the necessary security requirements in each layer. In the same way, D. Kozlov

et al. [11] analyze the security and privacy threats in IoT architectures by considering particular threat

scenarios for several levels.

On the other hand, a wide range of applications in various domains, including health-care, assisted and

enhanced-living scenarios, industrial and production monitoring, control networks, and many other

fields are based on Wireless Sensor Networks (WSNs). This characteristic shows that it is expected

soon in the future that WSNs will be integrated into IoT, where sensor nodes join the Internet

dynamically, and use it to collaborate and accomplish their tasks. However, WSNs technology has to

face many security and privacy issues [44] in order to be safe to be integrated into IoT infrastructure.

For example, C. Alcaraz et al. [10] analyzed the advantages and disadvantages of a possible full

integration of WSN into the Internet and conclude that some applications should not connect their nodes directly to the Internet, but other applications can benefit from using TCP/IP directly. D. Christin

et al. evaluate different approaches to integrate WSNs into the Internet and outline a set of IoT

challenges [13].

In addition, many studies have been conducted in WSNs so as to improve their security level and

protection of their users’ privacy, while promise energy savings. Q. Chen et al. [5] designed a

lightweight high-level sensor network encryption algorithm (IAES) which is based on Advanced

Encryption Standard Algorithms, but dominate it due to the fact that it reduces sensor node energy

consumption, cost, and space requirements. Experimental results showed the efficacy of their proposed

algorithm in the sections of RAM and ROM storage space overhead, CPU clock cycles and network

latency. A lightweight sensor node authentication algorithm has been proposed by S. Tripathy [6] for

WSN confidentiality and authenticity. This proposed lightweight security algorithm (LISA) is not

based on traditional cryptographic intensive computations and is regarded as ideal for WSNs

infrastructure. It has the ability to achieve a very good security level in a reduced implementation

complexity. Data freshness and data integrity security services with minimum computation and communication overhead, while protects over data loss. I. J. de Dieu et al. [7] have developed an

energy-efficient secure path algorithm (ESPA) for WSNs which provides a better performance in

maximizing the network lifetime. Distance-based energy-aware routing and data privacy protection

techniques are used by this ESPA so as to provide a good security level both in authenticity and

integrity of the sensed data, while maximize the network lifetime.

J. Albath and S. Madria [8] have designed a practical algorithm for data security (PADS) in WSNs

which presents protection with negligible overhead while the throughput is similar to the same network

without security. This algorithm is based on the embedding of a one-time pad so as to make any

information gleaned from eavesdropping useless to an attacker. The use of the one-time pad provides

both indeed data receive from a specific sensor, and sufficient protection level against injected

messages. A. Ramos and R. H. Filho [9] have designed a high accuracy sensor data security level

estimation scheme (SDSE) for WSNs which analyze attack prevention and detection mechanisms

additionally. This security model solves the previous existing security estimation schemes which full

ignored detection mechanisms and analyzed the security of WSNs based exclusively on prevention mechanisms.

In addition, various studies based on smart city concept have been conducted by many authors, so as to

contribute to the improvement of the quality life of citizens and the overall quality of services (QoS)

used into it. QoS routing protocols and privacy in wireless sensor networks proposed in [46]. The

authors proposed a system which provides additional trustworthiness, less computation power,less

storage space, more reliability, and a high level of protection against privacy disclosure attacks. A QoS

routing scheme which improves the probability of success of a local route repair in wireless sensor

networks was proposed in [47] and thus, it optimizes the QoS management, such as in the case of node

disappearance. In addition, an unified routing for data dissemination in smart city networks proposed in

[48], which promises better delivery ratio and latency compared to other relevant algorithms.

The upcoming project of building a smart city by integrating to its communication network multiple

ICT and IoT technologies has been presented and analyzed in various studies. Recently, Sian Lun Lau

et al. [33] analyzed the technology oppurtunities of the capital of Malaysia to be converted to a smart

city, improving the quality life of its citizens. Moreover, a comprehensive comparative analysis

between various smart cities projects in Spain is presented in [34]. More specifically, in [35], the

authors conduct an IoT experimentation over the city of Santander in Spain (SmartSantander testbed),

while smart Cities of the future are analyzed in [36]. Technological challenges and socioeconomic opportunities for developing an IoT Smart City Framework are described and analyzed in [37]. An

information framework of creating a smart city through IoT technology is proposed in [38].

Furthermore, a comparison analysis of various smart city projects is presented in [39] and concludes to

useful valuable implications which they should be considered when launching smart city frameworks.

Finally, an advanced information centric platform for supporting the typical ICT services of a Smart

City is presented in [40].

As mentioned above, we additionally make use of HEVC standard in our proposed scheme, so as to

ensure better compression to the transmitted media files over the IoT network of the smart city. Many

studies have been conducted on HEVC compression standard and how it can be used in terms of

reliability and security in video transmission over the internet. Specifically, a new scheme based on

selective encryption for HEVC is proposed in [16] and ensures transparent and sufficient encryption

and protection against attacks. Moreover, this scheme allows fast encryption and decryption while

preserving the format and length of the video stream. Furthermore, an efficient SE system for CABAC

entropy coding of HEVC video standard is proposed in [17], which presents sufficient protection against cryptanalysis attacks, while making it proper for streaming on eterogeneous networks, due to

the fact that bit-rate remains the same and the system requirements are minimal. Finally, an improved

encryption algorithm for efficient transmission of HEVC media in [18] demonstrates more security and

effectiveness compared to previous algorithms used for H.264 standard, while shows better overall

performance.

3 WSN Integration into IoT SC Framework

The architecture of IoT [50] is based on an Electronic Product Code (EPC), which was introduced by

GS1 EPCglobal Architecture Framework. All the connected objects (―things‖) will be equipped with

RFID tags with a unique EPC. Thus, RFID tags will operate as electronic identification for their

physical and virtual connected objects of the Smart City (SC).

A wireless sensor network (WSN) is composed of large number of small sensor nodes with limited

resources and densely deployed in an environment. Whenever end users require information about any

event related to some object(s), they send a request to the sensor network via the base station, and then

the base station conveys the request to the whole network or to a specific region of the network. In

response to this request, sensor nodes send back required information to the base station [19].

The integration of WSNs to the IoT is possible in three basic approaches as referred in [20], varying

from the WSN integration grade into the Internet structure. The first approach consists of connecting

both independent WSN and the Internet through a single gateway (independent network). This scheme

has been adopted by most of the WSNs accessing the Internet, and presenting the highest abstraction

between networks [13]. The second approach includes a hybrid network which is composed of both

independent networks, where few dual sensor nodes can access the Internet [13]. The third approach

includes the current WLAN structure with a dense 802.15.4 access point network, where multiple

sensor nodes can link the Internet in one hop [13].

Despite the fact that the integration of WSN into IoT provides many benefits, this scheme has to face

several big security challenges, such as privacy issues, so as to improve the protection level of the

infrastructure which will be used in SC framework. It is obvious that the establishment and

implementation of such a big network of interconnected devices will have to face some new security

threats and privacy issues. As it is expected, there will be many security gaps initially, making the

network vulnerable to hackers who will try to intrude and cause malicious actions for their benefit.

Specifically, the easy accessibility of the ―things‖ makes the infrastructure vulnerable with security

exploits, which can presume upon hackers [4].

Therefore, IoT should cover the following security challenges: authentication, access control,

confidentiality, privacy, secure middleware, policy enforcement, mobile security, and trust. Proper

authentication mechanisms must be developed so as to prevent malicious users to gain access to the

data which will be transferred into the IoT network of the SC. These mechanisms should ensure data

confidentiality, data integrity, and data availability [3]. This principal is known as CIA Triad. In the

next section we present our proposed method as an enhanced solution to this challenge.

Fig. 2. The proposed algorithm in flowchart format.

Fig. 3. Multihop packet structure in our proposed scheme.

4 Proposed Scheme

The objective of this research is to ensure to the community a safer and more lightweight infrastructure

for Smart City framework, so as to optimize the life quality of its citizens. From this perspective, we

propose a new routing security scheme which will ensure a faster, lighter and more secure transmission

of media sharing among the citizens of the smart city, such as in the case of a cybernetics social cloud

[57]. Cloud storage may ensure the personal data of the citizens of the smart city to be safe [54].

Furthermore, for terror or natural disasters, our scheme can make use of an important solution which

was proposed in [55] and refers to big data system recovery in private clouds.

Our proposed algorithm for efficient media transmission over the IoT network of the Smart City

merges Identity, Route and Location (IRL) Privacy Algorithm [19] for routing at sensor and

intermediate node, and Practical Algorithm for Data Security (PADS) [13] for WSN, adapted properly

so as to be applicable to the new video compression standard, HEVC for media files transmission.

Figure 2 shows in flowchart format the steps of media packet routing used in our proposed scheme.

As it is shown in Figure 2 and according to [19], each node classifies its neighboring nodes into

forward neighboring nodes (F), right side backward neighboring nodes (Br), left side backward

neighboring nodes (Bl), and middle backward neighboring nodes (Bm). Additionally, M(tF), M(tBr),

M(tBl), and M(tBm) represent the set of trustworthy nodes for each direction respectively. This scheme

contributes to the trustworthy forward of each media packet to the proper destination, on time and via reliable nodes.

Our proposed model operates in two rounds, such as IRL/r-IRL: neighbor node state initialization

round, and routing round. When a wireless node is to forward a packet, it is called the routing round for

source or intermediate node of EAMSuS algorithm. When a source node is to forward the packet, it

must check if there are available trustworthy neighboring nodes in its forward direction setM(tF), before

it forwards the packet. If there are trustworthy nodes, it will randomly choose one node as a next hop

from the setM(tF) and forward the packet towards it. If there is no trustworthy node in its forward

direction, the source node must check if there is available trustworthy node in the right (M(tBr)) and left

(M(tBl)) backward sets. If there are available trustworthy nodes, the source node will randomly choose

one node as a next hop from these sets and forward the packet towards it. If there is not trustworthy

node in these sets, the source node will randomly choose one trustworthy node from the backward

middle set (M(tBm)) and forward the packet towards it. If there are no trustworthy nodes available in all

of the sets, the packet will be dropped.

In the case of an intermediate node which receives the packet, either from the source node or from

another en-route node, it must firstly check whether the packet is new or old. If it is new, the node must check if there is trustworthy node from the forward direction set (MF), excluding the prevhop node if it

belongs to forward set. If there are trustworthy nodes in the forward set, the node will randomly choose

any one trustworthy node as a next hop and forward the packet towards it. If there is no trustworthy

node available in the forward direction, it must check to which set the sender of the packet belongs to.

The detailed description of the operation of the routing process which we adopt in our proposed

algorithm is available in [19].

By security perspective, our scheme differs from IRL/r-IRL one, in the fact that we embed PADS

algorithm [8] in the IRL/r-IRL, by setting the resulting p(xi) as the payload of the packet form: ―Set

payload= p(xi)” instead of using the command: ―payload = [E(IDx||Rn, k+

bs),E(d, kx,bs)]”, used in

IRL/r-IRL. The one-time pad of PADS algorithm calculates a MAC over the static part of the packet

[8]. Multi-hop routing has to change the packet header properly so as to route the packet, and thus, only

the part of the packet that does not change is used. The calculated MAC is appended to the data and the

secret key shared between the sender and the receiver is used to create a time synced key. This ensures

that an attacker has to be time synced with the network in order to break the encryption [8]. This

feature introduces additional protection in terms of encryption and authentication, while it makes more

lightweight the algorithm, as it will be proved in the next section. In addition, the algorithm is modified so as to be adaptable to support HEVC media files transmitted over the IoT network of a smart city.

Note that the detection process of the transmitted media packet in each sensor node works in the same

way with the embedding process and is identical with the PADS for finding the embedded pad,

removing it and returning to the original sensor value.

5 Experimental Results

The sequences that we used in the experiments are classified into five classes based on their resolution

(class A, B1, B2, C, D), retrieved by JCT-VC main configuration common conditions [52]. Class A

sequences correspond to ultra-high definition (HD) with a resolution of 2560 x 1600. Class B1 and B2

sequences correspond to full high-definition sequences with a resolution of 1920 x 1080. Class C and

Class D sequences correspond to WVGA and WQVGA resolutions of 800 x 480 and 400 x 240,

respectively. For the experiments, Class A includes the Traffic, PeopleOnStreet, Nebuta, and

SteamLocomotive sequences; Class B1 includes the Kimono, ParkScene sequences; Class B2 includes

the Cactus, BQTerrace and BasketballDrive sequences; Class C includes the RaceHorses, BQMall,

PartyScene and BasketballDrill sequences; and Class D includes the RaceHorses, BQSquare,

BlowingBubbles and BasketballPass sequences. The sequence media packets with the corresponding

frame count, frame rate and duration, are summarized in Table 1.

Table 1. The sequence packets and their properties based on the JCT-VC main configuration common conditions.

Table 2. Memory requirement for each scheme.

Scheme Memory Calculation

PFR (16+1)M bits

PSR (16+16+1)M bits

CAS K(6+7M)+16M bits

SAS K(4M+2N)+16M bits

IRL/r-IRL M(26+32Δt)+k+

bs+kx,bs bits

EAMSuS M(26+32Δt)+24 bits

The original sizes of these test sequences are calculated with the maximum class bitrates [25], as they

have been presented in [18]. Note that previous subjective video performance comparisons [24] shows

that Class A’ sequences compressed in HEVC standard (4K UHD) present an average 64 % bitrate

reduction compared to H.264, Class B’ (1080p) 62 %, Class C’ (720p) 56 %, and Class D’ (480p) 52

%, respectively.

Class Sequence Packet

Frame

Count

Frame

Rate (fps)

Duration

(sec)

A

Traffic 150 30 5

PeopleOnStreet 150 30 5

Nebuta 300 60 5

SteamLocomotive 300 60 5

B1

Kimono 240 24 10

ParkScene 240 24 10

B2

Cactus 500 50 10

BQTerrace 600 60 10

BasketballDrive 500 50 10

C

RaceHorses 300 30 10

BQMall 600 60 10

PartyScene 500 50 10

BasketballDrill 500 50 10

D

RaceHorses 300 30 10

BQSquare 600 60 10

BlowingBubbles 500 50 10

BasketballPass 500 50 10

Let us assume that we have 4 different sensor nodes where each one receives the test sequences –

packets – of each Class respectively. Therefore, sensor 1 receives one by one the Traffic,

PeopleOnStreet, Nebuta, and SteamLocomotive packets; sensor 2 receives one by one Kimono and

ParkScene packets, sensor 3 receives one by one the Cactus, BQTerrace and BasketballDrive (B1)

packets, and the RaceHorses, BQMall, PartyScene and BasketballDrill (B2) packets; and sensor 4

receives one by one the RaceHorses, BQSquare, BlowingBubbles and BasketballPass packets.

Cycle prevention strategy requires some short term memory to store signature of the packet for short period time (δt) [19]. Thus, for short term memory consumption, we assume two scenarios:

1. A media packet transmission method which makes use of both the PADS and IRL algorithm.

2. Our proposed packet transmission method which is based on these two algorithms, modified

properly for less memory consumption costs.

For both scenarios we consider two cases: H.264 and HEVC compression standards of the test

sequences.

In our scheme, signature of the packet comprises of the following fields, as it is shown in Figure 3: 1)

Sequence number (2 bytes), 2) previous hop identity (2 bytes), 3) next hop identity (2 bytes), 4) type (1

byte), 5) reading (2 bytes), 6) parent address (2 bytes), 7) data length (variable size), 8) MAC (4 bytes),

9) CRC (2 bytes), 10) counter (2 bits), and 11) δt time (4 bytes). Fields 4, 5, 6 and 7 constitute the

payload. This multihop packet structure is an assumption based on the contribution of J. Albath and S.

Madria for wireless sensor networks [8] and our proposals for our scheme. In our scenario, data length

is variable and dependent by the compression standard which is used in the transmitted media file.

Therefore, in our proposed scheme, each packet sensor node requires 21.25 + data length bytes of

memory. The packet signature will be removed from the buffer after δt time. Note that the additional

overhead does not make sensor nodes overloaded [19].

Figures 4-7 depicts the total short term memory consumption of each sensor node for each scenario in

H.264 and HEVC as follows: Fig.4 shows the total required memory for Class A sequence packets for

both PADS+IRL and EAMSuS algorithm; Fig.5 shows the total required short term memory for Class

B (B1&B2) sequence packets for both PADS+IRL and EAMSuS algorithm; Fig.6 shows the total

required short term memory for Class C sequence packets for both PADS+IRL and EAMSuS

algorithm; and Fig.7 shows the total required short term memory for Class D sequence packets for both

PADS+IRL and EAMSuS algorithm. As it is clearly shown in these diagrams, our proposed encryption

routing scheme (EAMSuS) overcomes the scenario of PADS+IRL in short term memory consumption

for each case (H.264 and HEVC compression standard use). Short term memory consumption presents

a significant decrease with our proposed algorithm and overcomes in efficiency the PADS+IRL

scheme.

For memory consumption analysis, we calculate the total memory required for our EAMSuS scheme

based on the corresponding calculation of PFR, PSR, CAS, SAS, and IRL/r-IRL schemes, which were

presented in [19]. Table 2 shows the memory requirement of these privacy schemes, in which M

represents the neighborhood size, K represents pseudonym space, 4Δt represents size of time window, and N is the total number of nodes in the network. Because of the fact that our proposed scheme does

not make use of base station’s public key (k+

bs) and shared secret key (kx-bs), which are used in IRL/r-

IRL scheme, but only one-time pad of PADS algorithm, it consumes: M(26+32Δt)+24 bits memory at

each sensor node. The value 24 represents the 3 bytes required for PADS algorithm.

Figure 8 depicts the total memory required for each scheme per each sensor node for a total of 100

sensor nodes (N=100), 8 bytes pseudonym space (K=8), Δt=5 seconds size of time window, and

k+

bs=20 bytes, kx-bs= 8 bytes (the keys used in IRL/r-IRL), respect to the total number of its neighbor

nodes (M).

Figure 9 depicts the maximum short term memory savings (%) per each sensor node by using the

proposed EAMSuS for HEVC compression standard, instead of using PADS+IRL for H.264 and

HEVC respectively. As it is mentioned above, sensor node 1 contains the sequence packets of Class A;

sensor node 2 contains the sequence packets of Class B; sensor node 3 contains the sequence packets of

Class C; sensor node 4 contains the sequence packets of Class D. The optimum overcoming of our

proposed scheme versus PADS+IRL occurs in sensor node 1 for Class A sequence packets and

approaches the 47% short term memory savings.

Figure 10 depicts the total memory savings (%) per sensor nodes’ neighborhood size. As it is clearly

shown our scheme achieves up to 50% memory savings per each sensor node for a total of 100 sensor

nodes (N=100), Δt=5 seconds size of time window, and k+

bs=20 bytes, kx-bs= 8 bytes (the keys used in

IRL/r-IRL).

As it is shown in these experiments, EAMSuS model seems to be more lightweight compared to other relevant schemes, such as IRL/r-IRL, while ensures a high level of privacy protection of the users in the IoT. Therefore, this concept could be adopted in Smart City Framework for a better quality level of communication among the citizens, optimizing their overall life quality and enhancing their privacy protection.

Fig.4. Total short term memory consumption at sensor node 1.

0

50

100

150

200

250

1 2 3 4

Mem

ory

Req

uir

ed

(MB

)

Class A Sequence Packets

Memory Usage in Sensor Node 1

PADS+IRL (H.264)

PADS+IRL (HEVC)

EAMSuS (H.264)

EAMSuS (HEVC)

0

50

100

150

200

250

300

350

1 2 3 4 5

Me

mo

ry R

equ

ired

(MB

)

Class B Sequence Packets

Memory Usage in Sensor Node 2

PADS+IRL (H.264)

PADS+IRL (HEVC)

EAMSuS (H.264)

EAMSuS (HEVC)

Fig.5. Total short term memory consumption at sensor node 2.

Fig.6. Total short term memory consumption at sensor node 3.

Fig.7. Total short term memory consumption at sensor node 4.

0

20

40

60

80

100

120

140

160

1 2 3 4

Me

mo

ry R

eq

uir

ed

(MB

)

Class C Sequence Packets

Memory Usage in Sensor Node 3

PADS+IRL (H.264)

PADS+IRL (HEVC)

EAMSuS (H.264)

EAMSuS (HEVC)

0

20

40

60

80

100

120

140

160

1 2 3 4

Mem

ory

Re

qu

ire

d (M

B)

Class D Sequence Packets

Memory Usage in Sensor Node 4

PADS+IRL (H.264)

PADS+IRL (HEVC)

EAMSuS (H.264)

EAMSuS (HEVC)

Fig.8. Memory consumption analysis per each sensor node with respect to the number of its

neighbor nodes.

Fig.9. Maximum short term memory savings (%) per each sensor node by using EAMSuS

(HEVC) against PADS+IRL (H.264) and PADS+IRL (HEVC) respectively.

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

1 2 3 4 5 6 7 8 9 10

Me

mo

ry R

eq

uir

ed

(b

its)

Neighborhood Size (M)

Memory Consumption Analysis

PFR

PSR

CAS

SAS

IRL/r-IRL

EAMSuS

0,00%

5,00%

10,00%

15,00%

20,00%

25,00%

30,00%

35,00%

40,00%

45,00%

50,00%

1 2 3 4

Max

. Sh

ort

Ter

m M

emo

ry S

avin

gs (

%)

Sensor Node

EAMSuS (HEVC) vs PADS+IRL

PADS+IRL (H.264)

PADS+IRL (HEVC)

Fig.10. Memory savings (%) per each sensor node with respect to the number of its neighbor

nodes.

6 Conclusions and Future Work

A new security scheme for efficient and secure media packet routing in IoT network was presented. This scheme is based on two other proposed methods by other researchers for WSNs, which merges them properly for an enhanced security level for the upcoming IoT infrastructure of smart city concept. Experimental analysis demonstrates that our proposed EAMSuS achieves less memory consumption at the wireless sensor nodes of the IoT, compared to IRL/r-IRL scheme, while ensures a high level of security, by using PADS algorithm for privacy and authentication reasons. Thus, we regard this scheme to be adaptable to the upcoming future concept of IoT Smart City framework, which will offer better quality and privacy level of the communication of its citizens, contributing to the overall quality of their life.

The memory savings per each sensor node depend on the number of its neighbor nodes, the total number of sensor nodes in the area, the values of the keys used in IRL/r-IRL algorithm, and the size of

time window. In a test with a total of 100 sensor nodes (N=100), Δt=5 seconds size of time window,

and k+

bs=20 bytes, kx-bs= 8 bytes, our scheme achieved up to 50% memory savings per each sensor

node. Moreover, the maximum short term memory savings per each sensor node by using EAMSuS

instead of PADS+IRL approach the 47% in our experiments. Future work may include more comprehensive analysis of our proposed scheme by simulating and testing it with a sample of real physical objects of a smart city project connecting to the internet and a data processing system for automated scheduled tasks by analyzing the gathered data of the objects. Furthermore, energy consumption at the wireless nodes for various scenarios would be calculated and compared to other proposed algorithms which focus on energy savings, such as IAES, ESPA, LISA, etc. Finally, we will consider ways to improve ad hoc and mobile networks security [53], while we will ensure the location privacy in the smart cities by considering corresponding mechanisms for this goal [56].

0%

10%

20%

30%

40%

50%

60%

1 2 3 4 5 6 7 8 9 10

Me

mo

ry S

avin

gs (

%)

Neighborhood Size (M)

EAMSuS vs IRL/r-IRL

References

[1] D. Miorandi, S. Sicari, F. De Pellegrini, I. Chlamtac, Survey internet of things: vision, applications and research challenges, Ad Hoc Netw. 10 (7), 1497–1516, 2012.

[2] Dave Evans (April 2011). "The Internet of Things: How the Next Evolution of the Internet Is Changing Everything" (PDF). Cisco. Retrieved 4 September 2015.

[3] M.U. Farooq, Muhammad Waseem, Anjum Khairi, Sadia Mazhar, "A Critical Analysis on the Security Concerns of

Internet of Things (IoT)", International Journal of Computer Applications (0975 8887), Volume 111 - No. 7, February 2015.

[4] Rodrigo Roman, Pablo Najera and Javier Lopez, ‖Securing the Internet of Things,‖ in IEEE Computer, Volume 44, Number 9, pp. 51-58, 2011.

[5] Qingzhang Chen, Zhongzhe Tang, Yidong Li, Yibo Niu, Jianhua Mo, "Research on Encryption Algorithm of Data Security for Wireless Sensor Network", Journal of Computational Information Systems 7:2, 369-376, 2011.

[6] Somanath Tripathy, "LISA: LIghtweight Security Algorithm for Wireless Sensor Networks", Distributed Computing and Internet Technology, Volume 4882 of the series Lecture Notes in Computer Science pp 129-134, 2007.

[7] Imanishimwe Jean de Dieu, Nyirabahizi Assouma, ManiraguhaMuhamad,Wang Jin, and Sungyoung Lee, "Energy-Efficient

Secure Path Algorithm for Wireless Sensor Networks", International Journal of Distributed Sensor Networks, Volume 2012, 2012.

[8] Julia Albath and Sanjay Madria, "Practical Algorithm for Data Security (PADS) in Wireless Sensor Networks", Proceedings of the 6th ACM international workshop on Data engineering for wireless and mobile access, Pages 9-16, 2007.

[9] Alex Ramos and Raimir Holanda Filho, "Sensor Data Security Level Estimation Scheme for Wireless Sensor Networks", Sensors Journal, 15, 2104-2136, 2015.

[10] Cristina Alcaraz, Pablo Najera, Javier Lopez, Rodrigo Roman, "Wireless Sensor Networks and the Internet of Things: Do We Need a Complete Integration?", 1st International Workshop on the Security of the Internet of Things, 2010.

[11] Denis Kozlov, Jari Veijalainen, Yasir Ali, "Security and Privacy Threats in IoT Architectures", Proceedings of the 7th International Conference on Body Area Networks Pages 256-262, September 2012.

[12] Sabrina Sicari, Alessandra Rizzardi, Luigi Alfredo Grieco, Alberto Coen-Porisini, "Security, privacy and trust in Internet of Things: The road ahead", Computer Networks, January 2015.

[13] Delphine Christin, Andreas Reinhardt, Parag S. Mogre, Ralf Steinmetz, "Wireless Sensor Networks and the Internet of

Things: Selected Challenges", Proceedings of the 8th GI/ITG KuVS Fachgespräch "Drahtlose Sensornetze", pp. 31--34, August 2009.

[14] M.Tharani, N. Senthilkumar, "Integrating Wireless Sensor Networks into Internet Of Things For Security", International Journal of Innovative Research in Computer and Communication Engineering, Vol.2, Special Issue 1, March 2014.

[15] Ohm, J.R., Sullivan, G.J., Schwarz, H., Tan, T.K., Wiegand, T.: Comparison of the coding efficiency of video coding

standards—including high efficiency video coding (HEVC). IEEE Trans. Circuits Syst. Video Technol. 22(12), 1669–1684, 2012.

[16] Hofbauer H., Uhl A., Unterweger A.: Transparent encryption for HEVC Using bit -stream-based selective coefficient sign encryption, In: IEEE international conference on acoustics, Speech and Signal Processing (ICASSP), pp. 1986–1990, 2014.

[17] Shahid Z., Puech W.: Investigating the structure preserving encryption of high efficiency video coding (HEVC), In: Proceedings SPIE, real-time image and video processing, vol. 8656, 2013.

[18] Vasileios A. Memos and Kostas E. Psannis, ―Encryption Algorithm for efficient transmission of HEVC media‖, Journal of Real-Time Image Processing, May 2015.

[19] Riaz Ahmed Shaikh, Hassan Jameel, Brian J. d’Auriol, Heejo Lee, Sungyoung Lee, and Young-Jae Song, "Achieving Network Level Privacy in Wireless Sensor Networks", Vol. 10, pp. 1447-1472, Sensors, Feb. 2010.

[20] R. Roman and J. Lopez, ―Integrating Wireless Sensor Networks and the Internet: a Security Analysis,‖ Internet Research: Electronic Networking Applications and Policy, vol. 19, no. 2, 2009.

[21] J. Holler, V. Tsiatsis, C. Mulligan, S. Karnouskos, S. Avesand, D. Boyle, "From Machine-to-Machine to the Internet of Things: Introduction to a New Age of Intelligence", Elsevier, 2014.

[22] Bhargava M., Chen Ch., Ryoo M. S., and Aggarwa J. K., "Detection of object abandonment using temporal logic", Springer Berlin, pp. 271-281, January 2009.

[23] Khot Harish S, Gote Swati R, Khatal Sonali B, Pandarge Sangmesh, Smart Video Surveillance, International Journal of Emerging Engineering Research and Technology, Volume 3, Issue 1, pp. 109-112, January 2015.

[24] Tan T.K., Mrak M., Baroncini V., Ramzan N.: Report on HEVC compression performance verification testing. Joint Collab. Team Video Coding (JCT-VC), 2014.

[25] Bossen F., ―Common HM test conditions and software reference configurations‖, document JCTVC-L1100 of JCT-VC, Geneva, CH, 2013.

[26] Sam Musa, "Smart Cities – A Roadmap for Development", January 2016.

[27] Paolo Neirotti, Alberto De Marco, Anna Corinna Cagliano, Giulio Mangano, and Francesco Scorrano, "Current trends in Smart City initiatives: some stylised facts", Cities, Vol 38, pp. 25-36, December 2013.

[28] European Smart Cities, Centre of Regional Science Vienna University of Technology, 2012.

[29] C. Harrison, B. Eckman, R. Hamilton, P. Hartswick, J. Kalagnanam, J. Paraszczak, and P. Williams, "Foundations for smarter cities", IBM Journal of Research and Development, Vol. 54, I. 4, pp. 1-16, 2010.

[30] A. Caragliu, C. del Bo, and P. Nijkamp, "Smart cities in Europe", Proc. of Central European Conf. in Regional Science, CERS, 2009.

[31] A. Coe, G. Paquet, and J. Roy, "E-governance and smart communities: a social learning challenge", Journal on Social Science Computer Review, Vol. 19, I. 1, pp. 80-93, 2001.

[32] A.-V. Anttiroiko, P. Valkama, and S. Bailey, "Smart cities in the new service economy: building platforms for smart services", Journal of AI and SOCIETY, Vol. 28, I. 1, pp. 1-12, 2013.

[33] "Greater Kuala Lumpur as a smart city: A case study on technology opportunities", 8th International Conference on Knowledge and Smart Technology (KST), February 2016.

[34] Stijn Bossuyt, "Smart cities projects in Spain: Comparative analysis between various cities", 2013.

[35] Luis Sanchez, Luis Munoz, Jose Antonio Galache, Pablo Sotres, Juan R. Santana, Veronica Gutierrez, Rajiv Ramdhany,

Alex Gluhak, Srdjan Krco, Evangelos Theodoridis, Dennis Pfisterer, "SmartSantander: IoT experimentation over a smart city testbed", Computer Networks, Vol. 61 pp. 217–238, 2014.

[36] M. Batty, K. Axhausen, G. Fosca, A. Pozdnoukhov, A. Bazzani, M. Wachowicz, G. Ouzounis, and Y. Portugali, "Smart cities of the future", The European Physical Journal, November 2012.

[37] Evangelos Theodoridis , Georgios Mylonas, and Ioannis Chatzigiannakis, "Developing an IoT Smart City Framework", Conference Paper, July 2013.

[38] Jiong Jina, Jayavardhana Gubbi, Slaven Marusic, and Marimuthu Palaniswami, "An information framework of creating a smart city through Internet of Things", IEEE Internet of Things Journal, Vol. 1, I. 2, pp. 112-121, April 2014.

[39] Yejin Yoo, Kabsung Kim, and Jihye Han, "Comparison Analysis of Smart City Projects - Implications for U-City", Business, Economics, Social Science & Humanities (BESSH), Vol. 76, I. 3, pp. 41-50, February 2016.

[40] G. Piro, I. Cianci, L. A. Grieco, G. Boggia, and P. Camarda, "Information centric services in Smart Cities", Journal of Systems and Software, January 2013.

[41] Y. Kryftis, G. Mastorakis, C. Mavromoustakis, J. Mongay Batalla, E. Pallis and G. Kormentzas, ―Efficient Entertainment Services Provision over a Novel Network Architecture‖. IEEE Wireless Communications, 2016.

[42] Kriti Bhushan, B. B. Gupta, "Security Challenges in Cloud Computing: State-of-art," International Journal of Big Data Intelligence (IJBDI), Inderscience, 2015.

[43] Gou Zhaolong, Shingo Yamaguchi, B. B. Gupta, ―Analysis of Various Security Issues and Challenges in Cloud Computing

Environment: A Survey,‖ Handbook of Research on Modern Cryptographic Solutions for Computer and Cyber Security, IGI-Global's Advances in Information Security, Privacy, and Ethics (AISPE) series, USA, 2016.

[44] S. Veluru, Y. Rahulamathavan, B. B. Gupta, M. Rajarajan, ―Privacy Preserving Text Analytics: Research Challenges and

Strategies in Name Analysis,‖ Book on Securing Cloud-Based Databases with Biometric Applications, IGI-Global's Advances in Information Security, Privacy, and Ethics (AISPE) series, 2014.

[45] J. Mongay Batalla, G. Mastorakis, C. Mavromoustakis, ―On cohabitating networking technologies with common wireless access for Home Automation Systems purposes‖ To be published in IEEE Wireless Communications Magazine, 2016.

[46] S. Phani Praveen M.Tech and Cse B. Sukumar Babu, "Qos Routing Protocols and Privacy in Wireless Sensor Networks", Global Journal of Computer Science and Technology, Volume 12, Issue 4, Version 1.0, February 2012.

[47] M. Belghachi and M. Feham, "Qos Routing Scheme and Route Repair in WSN", nternational Journal of Advanced ComputerScience and Applications, Vol. 3, No. 12, 2012.

[48] Viet-Duc Le, H. Scholten, and P. Havinga, "Unified routing for data dissemination in smart city networks", 3rd International Conference on the Internet of Things (IOT), October 2012.

[49] Murray James, "Cloud network architecture and ICT - Modern Network Architecture", ITKnowledgeExchange, TechTarget, Retrieved 2016-12-20.

[50] Jayavardhana Gubbi, Rajkumar Buyya, Slaven Marusic, and Marimuthu Palaniswami, "Internet of Things (IoT): A vision, architectural elements, and future directions", Future Generation Computer Systems, Vol. 29, pp. 1645–1660, 2013.

[51] M. V. Moreno, F. Terroso-Saenz, A. Gonzalez, M. Valdes-Vela, A. F. Skarmeta, M. A. Zamora-Izquierdo, and V. Chang,

"Applicability of Big Data Techniques to Smart Cities Deployments", IEEE Transactions on Industrial Informatics, Volume: PP, Issue: 99, September 2016.

[52] Victor Chang, Yen-Hung Kuo, and Muthu Ramachandran, "Cloud computing adoption framework: A security framework for business clouds", Future Generation Computer Systems, Vol. 57, pp. 24-41, April 2016.

[53] P. Vijayakumar, Victor Chang, L. Jegatha Deboraha, Balamurugan Balusamy, and P.G. Shynu, "Computationally efficient

privacy preserving anonymous mutual and batch authentication schemes for vehicular ad hoc networks", Future Generation Computer Systems, pp. 1-12, April 2017.

[54] Victor Chang, and Gary Wills, "A model to compare cloud and non-cloud storage of Big Data", Future Generation Computer Systems, Vol. 57, pp. 56-76, April 2016.

[55] Victor Chang, "Towards a Big Data system disaster recovery in a Private Cloud", Ad Hoc Networks, Vol. 35, pp. 65-82, December 2015.

[56] Gang Sun, Victor Chang, Muthu Ramachandran, Zhili Sun, Gangmin Li, Hongfang Yu, and Dan Liao, "Efficient location

privacy algorithm for Internet of Things (IoT) services and applications", Journal of Network and Computer Applications, October 2016.

[57] Victor Chang, "A cybernetics Social Cloud", Journal of Systems and Software, Vol. 124, pp. 195-211, February 2017.

nexthop

← ∅

node is the

source nodeprev

hop ← ∅

M(tF ) ≠ ∅

nexthop

(k) = Rand(M(tF));

nexthop

(k) = Rand(M(tBr

) ∪ M(tBl

))

M(tBr ) ∪ M(tBl) ≠ ∅

M(tBm) ≠ ∅

nexthop(k) = Rand(M(tBm))

Signature of new packet already exists in buffer

Mtemp

= {Mtemp

} + LasttimePrevhop

Mtemp

= {Mtemp

} + LasttimeNexthop

Set counter = timesReceivedBefore + 1

Remove signature from buffer

counter=3

Mtemp

= { Mtemp

} + prevhop

(M(tF) − { M(tF)∩Mtemp

} ) ≠ ∅

nexthop(k) = Rand(M(tF) − {M(tF)∩Mtemp

} )

Mtemp1

= M(tBl) ∪ M(tBm)

Mtemp1

≠ ∅

nexthop (k) = Rand(Mtemp1

)

M(tBr) ≠ ∅

nexthop(k) = Rand(M(tBr ) − {M(tBr )∩Mtemp

})

Yes

NoYes

NoYes

Yes No

No

Yes

No

packet camefrom Br

packet camefrom Bl

Yes

Yes

No

No

Yes

No

Mtemp2

= M(tBr ) ∪M(tBm)

Mtemp3

= M(tBr) ∪ M(tBl)

Mtemp2 ≠ ∅

nexthop

(k) = Rand(Mtemp2

− {Mtemp2

∩Mtemp

})

M(tBl

) ≠ ∅

nexthop

(k) = Rand(M(tBl) − { M(tBl)∩Mtemp

})

YesNo

NoYes

No

Yes

Mtemp3

≠ ∅

nexthop

(k) = Rand( Mtemp3

− { Mtemp3

∩Mtemp

} )

M(tBm) ≠ ∅

nexthop

(k) = Rand(M(tBm) − { M(tBm)∩Mtemp

} )

Yes

YesNo

Mtemp

= ∅

Yes

No

Yes

No

No

Drop packet

Exit

Set prevhop = myid

calculate MAC over p(xi) (excluding routing information) and save in p(x

i)

ktime = kij modified and time synced

β = ktime mod (b(MAC) − α + 1)

α = ktime mod b(MAC) + 1

temp_pad = substring of MAC starting at α for β bits

pad=Append temp_pad to pad until pad is the same length as the HEVC media file

xi = x

i XOR pad

Replace xi with xi in p(xi)

Set payload= p(xi)

Form pkt p = {prevhop

; nexthop

; seqID; payload}

Create Signature and save in buffer

Forward packet to nexthop

Set timer Δt = D/dnexthop

* pt

Δt = true Signature remains in buffer

Signature removed from buffer

Yes

No