The Study on Supporting Big Data Framework in Wireless

Surveillance Networks

Huan Du

and Haiyan Chen

The Third Research Institute of the Ministry of Public Security, Shanghai, China

East China University of Political science, Shanghai, China

Keywords: Big Data, Wireless Surveillance Networks, Data Sensing.

Abstract: Nowadays, Wireless Sensor Networks (WSNs) are based on techniques more and more oriented towards

image, video and sound processing, hence the recent need of Wireless Multimedia Sensor Networks

(WMSNs). One of the important challenges for real-time surveillance system is end-to end delay QoS for

packet deliveries. Providing end-to-end QoS is difficult due to two reasons. As wireless sensor nodes may

require multichip transmissions to reach the sink and some of the wireless transmissions may be not

successful. Multimedia data are characterized by their large volume, and have strict requirements in terms of

quality of service (QoS) such as bandwidth, delay, packet loss, delay jitter, etc. In this paper, we are

interested in routing protocols based on clusters that aim to reduce congestion in order to have reliable data

transmission and a reduced loss rate. This is achieved by balancing the traffic load, which results into a

balanced energy consumption within the network.

1 INTRODUCTION

Nowadays, Wireless Sensor Networks (WSNs)

(Takaishi, 2014) are based on techniques more and

more oriented towards image, video and sound

processing, hence the recent need of Wireless

Multimedia Sensor Networks (WMSNs) (Nadkarni,

2014). Examples of applications for WMSNs

include the monitoring of elderly people, the

monitoring of fields in precision agriculture, intruder

detection through video cameras, etc.

With recent developments in low cost hardware

such as CMOS cameras, microphones and PIR

sensors, the wireless nodes can be equipped with

these modules and have contributed to the

development of Surveillance Wireless Multimedia

Sensor Networks (SWSNs). One of the important

challenge for real-time surveillance system is end-to

end delay QoS for packet deliveries. Providing end-

to-end QoS is difficult due to two reasons. As

wireless sensor nodes may require multihop

transmissions to reach the sink and some of the

wireless transmissions may be not successful (Hou,

2015). Due to significant growth in data volume,

The corresponding author: Haiyan Chen

WSNs require protocols to support big data which

are characterized by 3Vs. These sensor nodes gather

a large volume and wide variety of the sensed data.

Due to sensor nodes constraints, many protocols are

designed and proposed to conquest these constrains

which include energy, self-management, wireless

networking, decentralized management, design

constraints and security (Dargie, 2010; Luo, 2011;

Xu, 2015). Conventional layered approaches

enhances the performance of these individual layers

but the cross-layer approaches have proved to

achieve better optimization results than their layered

counterparts (Mendes, 2011).

Multimedia data are characterized by their large

volume, and have strict requirements in terms of

quality of service (QoS) such as bandwidth, delay,

packet loss, delay jitter, etc. In this paper, we are

interested in routing protocols based on clusters that

aim to reduce congestion in order to have reliable

data transmission and a reduced loss rate. This is

achieved by balancing the traffic load, which results

into a balanced energy consumption within the

network. Our ESCC balance the clusters based on

neighbor, it make a comparison between number of

members in clusters. The current CH search between

its neighbor the CH that have more members, if the

number of members of the CH found is more than

134

Chen H. and Du H.

The Study on Supporting Big Data Framework in Wireless Surveillance Networks.

DOI: 10.5220/0006020501340137

In Proceedings of the Information Science and Management Engineering III (ISME 2015), pages 134-137

ISBN: 978-989-758-163-2

the members of the current CH plus 2, A FORCE-

JOIN is sent to the CH that have the more members.

We propose two cross layer protocol for

surveillance wireless sensor networks. The first is

suitable for high ratio of packet deliver at the sink.

The proposed load-balancing algorithm enhances the

ratio of packet delivery. But in the second, the

performance of the proposed end-to-end delay QoS

protocol is superior for providing end-to-end QoS.

The packet delay is important in many surveillance

systems. In this system, there is a limited admissible

delay if the packet delay is higher than that, this

packet is no longer useful and after that it is

discarded.

2 THE METRIC FOR WIRELESS

SURVEILLANCE NETWORKS

CAA (Congestion Avoidance and Alleviation

routing protocol) is designed to avoid congestion in

nodes. It detects congestion and sets the rate of

packets arriving at the nodes equal to the rate of

packet service. CAA uses a clustering method based

on residual energy to elect its cluster-heads, and a

CDMA (Code Division Multiple Access) method for

the inter-cluster communication, and a TDMA

method for the communication between members

and CH. In the transmission phase, a centralized

method is established by the sink for routing. The

sink broadcasts a flooding message with hop number

0 in the network, and the CHs rebroadcast this

message with their distance to the sink. The

congestion detected using a congestion degree. The

cluster-heads forward their data over the route that

has the minimum degree of congestion. An

additional technique to avoid congestion is the use

of a local buffer to store data during the congestion

period.

CRAP (Cluster based congestion control with

Rate Adjustment based on Priority) monitors

proactively the congestion and adjusts the traffic rate

when a cluster has a high priority data flow to be

transmitted. The adjustment rate is done by

exchanging the estimated rate of traffic between

clusters. This reduces the number of broadcast

packets and the energy loss. Three kinds of nodes

are present: the CH which schedules the

transmissions, the gateway node which interconnects

adjacent clusters, and the node members. The

member nodes send data, traffic rate and other

information to their own cluster-heads. This

collected data is transmitted over routes using ZRP

(Zone Routing Protocol). CRAP calculates the rate

of traffic in the clusters in order to alleviate the CH

congestion. The congestion degree is analyzed by

the CH. If this degree exceeds a given threshold, the

CH broadcast this value to all neighboring CHs to

adjust their congestion rate.

In summary, these protocols aim to reduce

congestion by reducing the traffic rate. This solution

causes a problem for the multimedia data. Indeed,

the video data transmission tolerates only a small

margin of packet loss. The MPEG compression

codec provides three type of frames: I-frames, B-

frames and P-frames. The most important frames are

I-frames and P-frames. Losing one of these frames

degrades significantly the video quality. In WMSNs,

CHs receives process and aggregate the multimedia

data from all their members. Thus, their load and

energy consumption is related with the number of

members they have. In the following, we propose a

new metric called MCUR, an optimal election and

our ESCC.

Our protocol, called ESCC (Equal Size Clusters to

reduce Congestion), aims to reduce the MCUR in a

distributed manner. It adds a new period to the

election phase called balancing, that balances the

size of clusters using a Force-Join message. Each

member keeps track of neighbors CH (by receiving

ADV message) and on their members (through

overhead Join messages). We introduce in the

election step a modified method to elected CH and

to join CH from that in LEACH to reduce the

number of CH neighbors and to eliminate the

isolated node. In the election part, the nodes do not

send their ADV messages together but it send it in

randomly delay, so if one neighbor receive more

than th ADV before sending its own ADV it decide

to become a member. In the join part, the nodes that

do not receive any ADV message, they called

isolated nodes, decide to become CH or member of

new CH (isolated node become CH in join step) in

order to maintain connectivity over network.

3 THE QOS OF THE PROPOSED

FRAMEWORK

Many approaches are presented to overcome the

limitations of wireless multimedia sensor networks

in different applications. According to network

subjects, their efficiency comparison is most

important. The assignment of the WMSN must be

The Study on Supporting Big Data Framework in Wireless Surveillance Networks

135

The Study on Supporting Big Data Framework in Wireless Surveillance Networks

135

consider to compare proposed protocol, for instance,

throughput maximization is not as important for

event-based application as it for monitoring

application. One of the important challenges for

real-time surveillance system is end-to-end delay

QoS for packet deliveries. Providing end-to-end QoS

is difficult due to two reasons. As wireless sensor

nodes may require multihop transmissions to reach

the sink and the some of the wireless transmissions

may be not successful. Load balancing is used to

increase network lifetime by reducing energy

consumption in wireless sensor networks. Some

papers considered the overview of load balancing

algorithm. Using clustering for load balancing can

also decrease energy consumption and increase

network lifetime and scalability. The cluster head

needs to send aggregated data to the base station. It

consumes more energy for data aggregation than the

member nodes.

In my network, the nodes are randomly deployed

and gather information by its sensor and periodically

send it to a sink node. To send node data to sink,

each node sends data to a neighboring node which is

nearer to the sink. Other neighbors go to sleep until

transmission ends. We propose a cross-layer design

to increase network lifetime. This protocol divides

the network to several levels. Each level shows the

distance between the node and the sink. Each node

acquires its level in the initial phase. After the initial

phase, the node knows the number of hops to the

sink node. For instance, a node neighboring the sink,

knows that there is not any hop between itself and

the sink node. Thus the node sets its level to ‘one’. A

node, neighboring the ‘level-one’ node and is not a

neighbor of the sink node, knows that it is one hop

away from the sink node and then sets its level to

‘two’ and the rest of the nodes set their levels

likewise.

When the surveillance system needs to provide

end-to-end delay QoS, We propose new algorithm

providing end-to end delay QoS. This protocol uses

a route table and only a first initial phase for sending

data to sink. A node operates the same as the

aforementioned first initial phase with the exception

that it saves the mac address of each get-level

massage receiving from the lower level nodes. Each

nodes estimate Minimum Delay Time (MDT) that it

can deliver a packet to sink and knows the MDT of

its lower level neighbor. It finds the minimum MDT

of its lower level neighbor and selects this lower

level neighbor for next hop to send a packet. For

providing the end-to-end delay QoS, a delay field is

inserts to data packet and when the delay field value

become bigger than end-to-end delay QoS, this

packet is discarded. For end-to-end delay QoS, when

a node generates a data and send it form the

application layer to the cross layer, the cross layer

inserts a delay field to the data packet. This field is

henceforth called delay field throughout the present

paper. Each node adds a delay time, which a packet

stays in the queue until it is sent, to delay field.

When a node receives a data packet, it checks the

sum of the delay field and Minimum Delay Time

(MDT), which the node can deliver a packet to the

sink and the detailed descriptions are explained in

the following paragraph. If the result is bigger than

the admissible end-to-end delay, the packet is

discarded otherwise the receiver node push the

packet to its queue. The RTS/CTS handshaking is

used to data transmission. This transmission is the

same as the previous section but in this case, the

sender node select the next hop from its route table

therefore there is no longer receiver contention to

send CTS packet. The next hop is the node that its

MDT is minimum value in the route table. Each

node sends the packet that its delay field is

maximum value in the queue.

4 SIMULATIONS

We evaluate our proposed cross-layer protocols

using MIXIM package and Omnet++ simulator.

Each sensor node periodically samples the data and

sends it to sink node. The simulation result for 10

random network topologies with 100 sensor nodes in

400*400 m2 area is presented with the sink position

(250,250). When sensor node samples data with low

rate, the energy consumption is high. In low

sampling frequency, when data rate increases, the

average energy consumption decreases severely.

After the node sends its data, it stays awake and

listens to the channel until the next transmission

starts. The node wastes energy in this time. When

data rate increases, the time the node stays awake

and wastes its energy in a way that energy efficiency

of the node increases. In order to solve this problem,

many papers use duty cycle mechanism. This

mechanism increases the delay and is designed for

low rate data but our aim is to support the big data.

In high data rate, the energy efficiency of the node

slowly increases because the node has to be more

competitive for getting channel. The second initial

phase has a significant impact on average energy

consumption. It tries to balance energy consumption

through the network therefore the local congestion

decrease and energy efficiency increase. When the

ISME 2015 - Information Science and Management Engineering III

136

ISME 2015 - International Conference on Information System and Management Engineering

136

data rate is bigger than 50 packet per minute for

each node, the probability of a collision increases

and the energy efficiency decreases.

5 CONCLUSIONS

Nowadays, Wireless Sensor Networks (WSNs) are

based on techniques more and more oriented

towards image, video and sound processing, hence

the recent need of Wireless Multimedia Sensor

Networks (WMSNs). One of the important challenge

for real-time surveillance system is end-to end delay

QoS for packet deliveries. Providing end-to-end QoS

is difficult due to two reasons. As wireless sensor

nodes may require multichip transmissions to reach

the sink and some of the wireless transmissions may

be not successful. Multimedia data are characterized

by their large volume, and have strict requirements

in terms of quality of service (QoS) such as

bandwidth, delay, packet loss, delay jitter, etc. In

this paper, we are interested in routing protocols

based on clusters that aim to reduce congestion in

order to have reliable data transmission and a

reduced loss rate. This is achieved by balancing the

traffic load, which results into a balanced energy

consumption within the network.

ACKNOWLEDGEMENTS

This work was supported in part by the National

Science and Technology Major Project under Grant

2013ZX01033002-003, in part by the National High

Technology Research and Development Program of

China (863 Program) under Grant 2013AA014601,

the project of Shanghai Municipal Commission of

Economy and Information under Grant 12GA-19.

REFERENCES

Takaishi D, Nishiyama H, Kato N, Miura R (2014)

Toward Energy Efficient Big Data Gathering in

Densely Distributed Sensor Networks. IEEE Trans

Emerg Top Comput 2:388–397. doi:10.1109/

TETC.2014.2318177

Nadkarni A, Vesset D (2014) Worldwide Big Data

Technology and Services 2014–2018 Forecast. IDC

Rep.250485:

Luo, X., Xu, Z., Yu, J., and Chen, X. 2011. Building

Association Link Network for Semantic Link on Web

Resources. IEEE transactions on automation science

and engineering, 8(3):482-494.

Xu, Z. et al. 2015. Knowle: a Semantic Link Network

based System for Organizing Large Scale Online

News Events. Future Generation Computer Systems,

43-44:40-50.

Hou I-H (2015) Packet Scheduling for Real-Time

Surveillance in Multihop Wireless Sensor Networks

With Lossy Channels. IEEE Trans Wirel Commun

14:1071–1079. doi: 10.1109/TWC.2014.2363678

Dargie WW, Poellabauer C (2010) Fundamentals of

wireless sensor networks: theory and practice. John

Wiley & Sons

Mendes LDP, Rodrigues JJPC (2011) A survey on cross-

layer solutions for wireless sensor networks. J Netw

Comput Appl 34:523–534. doi: 10.1016/

j.jnca.2010.11.009

The Study on Supporting Big Data Framework in Wireless Surveillance Networks

137

The Study on Supporting Big Data Framework in Wireless Surveillance Networks

137