A SPATIOTEMPORAL DATA DISSEMINATION PROTOCOL

FOR SLOWLY-VARYING MOBILE SINKS IN WIRELESS

SENSOR NETWORKS

Seungmin Oh, Jeongcheol Lee, Hosung Park, Yongbin Yim and Sang-Ha Kim

Department of Computer Engineering, Chungnam National University, Daejeon, Republic of Korea

Keywords: Mobile Sink, Real-time Service, Wireless Sensor Networks, Region-based.

Abstract: In wireless sensor networks (WSNs), previous real-time data dissemination schemes for stationary sinks

exploits the spatiotemporal approach, which utilizes the delivery speed based on the fixed distance between

a source and a static sink to fulfil the desired time deadline. This approach cannot be directly applied to a

mobile sink since the distance can be varied depending on its movement. That is, the delivery speed cannot

be determined without knowing the distance between a source and a mobile sink. It must be the hardest

problem for all intermediate sensors to know the current location update of the mobile sink and control their

transmission speeds. We focus on the real-time data dissemination for a slowly varying mobile sink

compared to the transmission speed, which may be the most common case. By the slowly varying constraint,

the movable area of a mobile sink can be determined by the two factors: its speed and the distance between

a source and its initial location. In this paper, we propose a real-time data dissemination scheme for a

mobile sink by utilizing the movable area concept. Simulation results show that the proposed scheme

provides a high success ratio in terms of the guarantee of real-time requirement.

1 INTRODUCTION

Many applications in WSNs, such as battlefield

surveillance and earthquake response systems, are

tailored to interact with fast changing events and

required to gather the event data in an application

desired time deadline (He, 2005) (Felemban, 2006).

The existing protocols for the real-time applications

in WSNs mainly exploit the spatiotemporal

approach which forwards data at the delivery speed.

The desired delivery speed is calculated by dividing

end-to-end distance by the time deadline. In the

protocols, each node on the data dissemination route

selects a node as its next-hop node which is nearer to

the sink and provides a better relay speed. The relay

speed means the advance in distance to each next

node dividing by the delay to forward a packet to the

each next node. End-to-end real-time data

dissemination is achieved by maintaining the desired

delivery speed from sources to the sink. However,

the protocols typically assume sinks are stationary.

Recently, sink mobility has been receiving

attentions from the researchers of WSNs (Li, 2009)

(Akyildiz, 2002). Since the mobile sinks are able to

move within the sensor field, source nodes could not

accurately know the location of the mobile sink

because of continuous movement of the sink. For

mobile sinks, the spatiotemporal approach cannot be

directly applied since the distance could be varied

depending on the movement of mobile sinks. That is,

the delivery speed could not be determined without

knowing the intermediate sensors to know the

current location update of the mobile sink and

control their transmission speed. Moreover, since the

time overhead for sink location update is occurred

frequently, it might not fulfill the time deadline.

Since the end-to-end distance may be longer, the

initially calculated delivery speed could not

guarantee the real-time data dissemination. Hence,

the real-time data dissemination needs to adapt a

new approach for the distance between the source

and the mobile sink. We focus on the real-time

protocol only for a slowly varying mobile sink

compared to the transmission speed, which may be

the most common case. By the slowly varying

constraint, the movable area of a mobile sink can be

determined by the two factors: its speed and the

distance between a source and its initial location.

In this paper, we propose a real-time data

dissemination protocol for slow varying mobile

387

Oh S., Lee J., Park H., Yim Y. and Kim S..

A SPATIOTEMPORAL DATA DISSEMINATION PROTOCOL FOR SLOWLY-VARYING MOBILE SINKS IN WIRELESS SENSOR NETWORKS.

DOI: 10.5220/0003907303870390

In Proceedings of the 2nd International Conference on Pervasive Embedded Computing and Communication Systems (PECCS-2012), pages 387-390

ISBN: 978-989-8565-00-6

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

sinks based on a virtual region for movable area of

the sink. The protocol utilizes the movable area

concept. The virtual region of movable area provides

the maximum distance of the slow varying mobile

sink. Through the virtual region, the source node

could determine the data delivery speed for the

spatiotemporal approach. Simulation results show

that the proposed protocol provides a high success

ratio in terms of the guarantee of real-time

requirement.

Figure 1: Real-time data dissemination for mobile sink

with modified spatiotemporal approach.

2 PROPOSED PROTOCOL

In this section, we propose and describe a Region-

based Real-time Data Dissemination (RRDD)

scheme to support sink mobility of real-time data

dissemination.

2.1 Assumptions

RRDD relies on several assumptions that are

explicitly and implicitly exploited in other studies

about real-time routing (He, 2005) (Felemban, 2006)

(Li, 2009) and many geographic unicast routing

schemes (Akyildiz, 2002) (Hamida, 2008) (Karp,

2000) as follows.

• Once a phenomenon appears, the sensor nodes

surrounding the phenomenon collectively gather

information and one of them becomes the source

to generate data of the phenomenon (Akyildiz,

2002) (Hamida, 2008). And the mobile sinks

require knowing where the sources locate.

• The source nodes that generate event data could

be provided the location of sink by one of the

sink location services (Li, 2000) (Park, 2009).

• For the geographic unicasting routing, which is

one of the stateless routing method, each sensor

node is aware of its own location after

deployment by receiving Global Positioning

System (GPS) signals (Karp, 2000) or using

some localization techniques (Bulusu, 2000).

 All the nodes and the sinks have a priori

knowledge about the size and the center point (x

) of the sensor field.

2.2 Spatiotemporal Approach

While in multi-hop wireless sensor networks, since

communication is physically bounded, the end-to-

end delay depends not only on single hop delay

(temporal), but also on the distance a packet travels

(spatial). The spatiotemporal approach for real-time

data disseminations utilizes a data delivery speed.

The approach calculates the data delivery speed by

dividing the end-to-end distance by the desired time

deadline. Each node on the dissemination route

selects a node as its next-hop node if it is nearer to

the sink and provides a better relay speed than the

desired delivery speed. The relay speed means the

advance in distance to each next node dividing by

the delay to forward a packet to the node. The end-

to-end real-time data dissemination is achieved by

maintaining the desired delivery speed.

2.3 Modified Spatiotemporal Data

Dissemination Approach

In the proposed protocol, the data dissemination

procedure consists of two steps: 1) forwarding the

data from the source to one of the vertex nodes

registered by the sink, and 2) forwarding from the

vertex node to the sink. The time deadline T

setdeadline

depends on applications in WSNs.

To reduce data detour, the vertex node, which

locates between the source and the sink, is selected

as a data relay node. The source node selects a

vertex node (x

, y

) as follows:

()( )0

s cp cp src

scp cpsrc

xx x x

yy y y

−

×− ≥

−

×− ≥

(1)

where (x

, y

) and (x

src

, y

src

) is the location of the sink

and the source node, respectively. We define T

and

as the time of forwarding data from the source to

the selected vertex node, and the time of forwarding

from the vertex node to the sink, respectively.

setdeadline

should be larger than the total sum of both

and T

. The distance of forwarding data is the

total of the distance from the source node to the

vertex node and the distance between the vertex

node and the sink node. However, since the mobile

sink moves continuously, we could not estimate the

distance between the vertex node and the sink. The

length of the longest straight line from a crossing

Sink

Source

Vertex node

_Setspeed Modified

etdeadline

PECCS 2012 - International Conference on Pervasive and Embedded Computing and Communication Systems

388

point within a coarse-grained grid is

. From the

nature of wireless sensor networks, since the time

for data propagation is proportional to the distance

for itself (Park, 2009), the following equation is

established. The source node could calculate the

modified delivery speed S

setspeed_modified

, which is the

desired delivery speed from itself to the vertex node,

with the calculated time T

as follows:

2(,)

setspeed_modified

SetDeadline

dsrccp

(2)

A vertex node, which receives data from the source,

acts as the relay node for the virtual region. The

node also stores the current location information of

the sink to relay the data from the source to the sink.

However, since the time passes from the registration

time, the sink may not be on the registered point.

Therefore, before the new periodical update of sink

location, data from the source should be sent to the

area the sink will be expected to locate. For this, we

are able to exploit flooding data into a whole cell to

disseminate them. But it wastes a lot of energy since

all nodes in the region have to participate in the

communication mode. The width of a cell of the grid

is V

x T

setdeadline

. During the T

setdeadline

, the sink could

move out of the registered cell of the grid but within

the cell and at most its adjacent cells. The vertex

node concurrently sends the data to 9 cell leaders. If

some of the adjacent grid cells are not belonged to

the virtual region, the sink prepare a new region

through the current location. Through this process,

though the sink moves out of the virtual region, the

data could be continuously disseminated to the sink.

In the cell of the grid, the data are propagated to the

sink by limited flooding. The total of the distance

from a vertex node to the leader of grid and the

distance of data flooding is equal to or less than

Consequently, the data delivery in the grid cell is

achieved.

3 PERFORMANCE EVALUATION

3.1 Simulation Environments and

Metric for Performance Evaluation

In this section, we compare the performances of

RRDD with that of other real-time data

dissemination protocol, SPEED (He, 2005), which is

the most popular real-time protocol in WSNs.

However, since SPEED does not consider mobility

support, we add footprint chaining method to

SPEED. The protocols are implemented in Qualnet

(Scalable, SITE). The simulation network space

consists of 2,500 sensors randomly deployed in

500m × 500m area. The radio range of sensor is

10m. A source node generates 30 byte-packets with

interval 0.5s. The simulation time is 50 seconds and

the sink has random way mobility model.

Transmitting and receiving power consumption rates

are 21mW and 15mW, respectively (Hill, 2002). All

results are the average values of 100 times of

simulation.

3.2 Simulation Results

Figure 2 shows the values of TDSR in the proposed

scheme and SPEED in accordance of the sink speed.

We vary the speed from 0m/s (stationary) to 35m/s.

The width of the virtual region is 100m and the end-

to-end distance between the source and the sink is

fixed at 200m. Since SPEED calculates the desired

delivery speed with the initial location of the sink,

the possibility of data dissemination on the time

deadline is decreased. And as the speed of the sink

moves faster and the distance due to foot-print

chaining becomes longer, the possibility is decreased

much more. However, in the proposed scheme, since

the source node calculates the delivery speed

considering the region to be expected for the sink to

locate and sends its data with the desired delivery

speed, the value of TDSR in the scheme is higher

than that in SPEED though the speed of the sink

increases.

Figure 2: Time deadmine success ratio according to sink

speed.

The values of TDSR of the RRDD and the

SPEED according to the end-to-end distance are

presented in fig. 3. The longer the end-to-end

distance becomes, the larger the hop count of data

path is. The speed of the mobile sink is fixed to

15m/s and the width of the virtual region is 100m. As

0 5 10 15 20 25 30 35

100

Time Deadline Success Ratio (TDSR) (%)

Sink S

eed

(

m/s

)

Proposed Scheme

SPEED + Footprint Chaining

A SPATIOTEMPORAL DATA DISSEMINATION PROTOCOL FOR SLOWLY-VARYING MOBILE SINKS IN

WIRELESS SENSOR NETWORKS

389

described above, since SPEED does not consider the

increase of the end-to-end distance, the value of

TDSR is decreased as the distance increases. But

since the proposed scheme considers both the

distance and the speed of sink, it has higher value of

TDSR than SPEED.

Figure 3: Time deadline success ratio according to end-to-

end distance.

Figure 4 indicates the TDSR of the proposed

scheme according to both the end-to-end distance

and the virtual region size. We also vary the region

size from 60m to 140m. The speed of the mobile

sink is fixed to 20m/s and the time deadline is 1.0s.

If the grid size is smaller, the frequent updates are

need. Since the mobile sink could move out of the

grid region during data dissemination due to small

grid size, the TDSR is decreased.

Figure 4: Time deadline success ratio according to both

sink speed and grid size.

4 CONCLUSIONS

In this paper, we propose a new real-time data

dissemination scheme for mobile sinks in wireless

sensor networks. The existing schemes for real-time

service to stationary sinks exploits the

spatiotemporal approach which forwards data at the

delivery speed calculated through the fixed distance

between a source and a static sink and the desired

time deadline. However, in case of mobile sinks,

since the distance between the source and the sink is

dynamically changed, it is difficult to adapt the

approach. So, our scheme considers the virtual

region to be expected for the mobile sink to locate

in, and calculates the desired delivery speed based

on the region. The proposed scheme exploits the

virtual region and grid based infrastructure.

Simulation results show that the proposed scheme

has better performance than the existing scheme in

terms of guaranteeing the real-time dissemination.

REFERENCES

T. He, J. A. Stankovic, T. F. Abdelzaher, and C. Lu, “A

Spatiotemporal communication protocol for wireless

sensor networks,” IEEE Trans. Parallel and Distrib.

Syst., Vol. 16, No. 10, pp. 995-1006, Oct. 2005.

E. Felemban, C. Lee, and E. Ekici, “MMSPEED:

Multipath Multi-SPEED Protocol for QoS Guarantee

of Reliability and Timeliness in Wireless Sensor

Networks,” IEEE Trans. Mobile Computing, Vol. 5,

No. 6, pp. 738-754, Jun. 2006.

Y. Li, C. S. Chen, Y.-Q. Song, Z. Wang, and Y. Sun,

“Enhancing Real-Time Delivery in Wireless Sensor

Networks With Two-Hop Information,” IEEE

Transactions on Industrial Informatics, Vol. 5, No. 2,

pp. 113-122, May 2009.

I. F. Akyildiz et al., “A survey on sensor networks,” IEEE

Communications, Vol. 40, pp. 102-114, Aug. 2002.

E. B. Hamida, and G. Chelius, “Strategies for Data

Dissemination to Mobile Sinks in Wireless Sensor

Networks,” IEEE Wireless Communications, vol.15,

no.6, pp.31-37, Dec. 2008

J. Li, J. Jannotti, D. S. J. D. Couto, D. R. Karger, and R.

Morris, “A scalable location service for geographic Ad

hoc routing,” in Proc. 6th Annu. ACM/IEEE Int. Conf.

Mobile Computing and Networking (MobiCom ’00),

2000, pp. 120–130.

H. Park, T. Kim, J. Lee, M.-S. Jin, and S.-H. Kim, “Sink

location via Inner Rectangular in Wireless Sensor

Networks,” in Proc. IEEE International Conferences

on Advanced Information Networking and

Applications (AINA), May 2009.

B. Karp and H.T. Kung, “GPSR: Greedy perimeter

stateless routing for wireless networks,” In Proc. of the

6th Annual Int'l Conf. on Mobile Computing and

Networking, ACM Press, 2000.

N. Bulusu, J. Heidemann, and D. Estrin, “GPS-less Low

Cost Outdoor Localization for Very Small Devices,”

IEEE Personal Communications Magazine, vol. 7, no.

5, pp. 28-34, Oct. 2000.

J. Hill and D. Culler, “Mica: a wireless platform for

deeply embedded networks,” IEEE Micro, Vol. 22,

Iss. 6, pp. 12-24, Nov./Dec. 2002.

Scalable Network Technologies, Qualnet, [online]

available: http://www.scalable-networks.com.

0 50 100 150 200 250 300 350 400

100

Time Deadline Success Ratio (TDSR) (%)

Distance from source to sink (m)

Proposed Scheme

SPEED + Footprint Chaining

0 50 100 150 200 250 300 350 400

100

Time Deadline Success Ratio (TDSR) (%)

Distance from source to sink (m)

Proposed Scheme (60m)

Proposed Scheme (80m)

Proposed Scheme (100m)

Proposed Scheme (120m)

Proposed Scheme (140m)

PECCS 2012 - International Conference on Pervasive and Embedded Computing and Communication Systems

390