Performance Improvement in Beacon-enabled LR-WPAN-based

Wireless Sensor Networks

Hong Min Bae

, Chan Min Park

, Shinil Suh

, Rana Asif Rehman

and Byung-Seo Kim

Dept. of Electronics and Computer Eng., Hongik University, 2369 Sjongro, Sejongsi, Korea, Republic of

Graduate School of Smart City Science Management, Hongik University, 2369 Sejongro, Sejongsi, Korea, Republic of

Dept. of Electronics and Computer Eng., Hongik University, 2369 Sjongro, Sejongsi, Korea, Republic of

Dept. of Computer and Information Communications Eng., Hongik University, 2369 Sejongro, Sejongsi, Korea, Republic of

Keywords: LR-WPAN, Interference, WLAN, Beacon.

Abstract: LR-WPANs have two types of networks: beacon–enabled and non-beacon-enabled networks. In beacon-ena-

bled LR-WPANs, the high reliability of Beacon frame transmission is required because all transmissions is

controlled by the in-formation in the Beacon frame. However, the process to handle the case for the beacon-

loss is not well-defined in the standard. In this paper, an enhanced protocol for the case when a Beacon frame

is lost is proposed to improve network performances. The protocol allows a device not receiving a Beacon

frame to keep transmit its pending frames only within the minimum period of CAP based on the previously

received Beacon frame while the standard prevents the device from sending any pending frame during a whole

superframe. By simulation and evaluations, the effectiveness of the proposed protocol on improving perfor-

mances is proven.

1 INTRODUCTION

IEEE802.15.4 standard (IEEE, 2011) specifies Low

Rate-Wireless Personal Area Networks (LR-

WPANs) for the low-cost devices’ communications

with a short range, low data rate, and low power con-

sumption. Applications using such IEEE802.15.4

standards-based LR-WPANs have been increasing in

broad areas including situation awareness, medical

services, public safety, home entertainment system,

smart home automation systems, ubiquitous building

systems, traffic information systems, public safety

systems, and so on. IEEE802.15.4 standard defines

two types of LR-WPANs: beacon–enabled and non-

beacon-enabled networks. Any transmissions of any

device in the beacon-enabled LR-WPANs are con-

trolled by the information in the Beacon frames trans-

mitted by the central PAN coordinator. Therefore, the

high reliability of Beacon frame transmission is es-

sential for beacon-enabled LR-WPANs. Furthermore,

the importance of beacon-enabled LR-WPANs also

increases as multimedia traffics are served over

WSNs to meet QoS.

However, Beacon frames are not successfully de-

livered to member devices because of collisions,

interferences from other heterogeneous communica-

tion devices, and erroneous channel. The collision

problems between Beacon frames have been studied

in (Kim et al., 2008), (Koubaa et al., 2007), (Nam and

Hwang, 2014), (IEEE, 2012). Beacon frame colli-

sions occur when a device is in the transmission range

of two PAN coordinators. In this case, the device may

receive two Beacon frames from both coordinators at

the same time and as a consequence the Beacon

frames are in collision. To resolve this problem, re-

searches in (Kim et al., 2008), (Koubaa et al., 2007),

(Nam and Hwang, 2014), (IEEE, 2012) propose a few

methods. Most of the methods are to schedule or to

distribute the transmissions of Beacon frames of mul-

tiple piconets, so that the collision is prevented. Par-

ticularly, IEEE802.15.4e standard (IEEE, 2012) de-

fines ‘Beacon Scheduling” method to prevent from

beacon collision. On the other hand, the loss of Bea-

con frame due to interference occurs because many

communication networks like LR-WPANs, Wireless

Local Area Networks (WLANs), and even micro-

waves uses same frequency bands of 2.4GHz which

is called Industrial Scientific Medical (ISM) band

(Lau et al., 2009). As a consequence, LR-WPANs ex-

perience severe interferences from other devices. The

Bae, H., Park, C., Suh, S., Rehman, R. and Kim, B-S.

Performance Improvement in Beacon-enabled LR-WPAN-based Wireless Sensor Networks.

DOI: 10.5220/0005632400890094

In Proceedings of the 5th International Confererence on Sensor Networks (SENSORNETS 2016), pages 89-94

ISBN: 978-989-758-169-4

performance degradations of LR-WPAN due to inter-

ferences are reported by many experiments and stud-

ies as shown in (Howitt and Gutierrez, 2003), (Sikora

and Groza, 2005), (Yoon et al., 2006), (Yuan et al.,

2007), (Shin et al., 2007), (Park and Kim, 2014). As

the electric power grid systems recently utilize LR-

WPANs and WLANs, the interference issues in the

power grid system is reported as shown in (Stanci-

ulescu et al., 2012). Especially, as the number of de-

ployed WLANs rapidly increases, the impacts on LR-

WPANs of interferences from WLANs are actively

researched in (Sikora and Groza, 2005), (Yoon et al.,

2006), (Yuan et al., 2007), (Chen et al. 2015) and it

is shown that LR-WPANs coexisting with WLANs

experience 10~100 % degradations on the perfor-

mances depending on the distances between LR-

WPANs and WLANs, locations, the channels used by

LR-WPANs, and the traffic loads of WLANs. There

are many studies to avoid the interference. To resolve

the problem, the most of methods switch the operat-

ing channels to non-interference channel. Some other

methods allow piconets using interference channel to

borrow some part of superframe of piconets using

non-interference channel.

While all aforementioned studies propose meth-

ods to avoid a beacon loss, no aforementioned studies

mentions the process itself for the case that a device

fails to receive a Beacon frame. Even though many

solutions have been proposed, Beacon frame can still

be lost because of the channel characteristics like

noise, fading, Doppler effects, and so on. Based on

IEEE802.15.4 standard, devices failed to receive a

Beacon frames have to hold their pending transmis-

sions during a superframe, so that it cause perfor-

mance degradations. Therefore, we need to a better

way to improve the network performances when the

Beacon frame is lost.

Figure 1: Superframe Structure.

In this paper, enhanced protocol is proposed to

overcome the performance degradation when Beacon

frames are not successfully transmitted in beacon-en-

abled LR-WPANs. The proposed protocol allows de-

vices to transmit its pending frames during a Conten-

tion Access Period (CAP).

In Section 2, IEEE802.15.4 standard-based LR-

WPANs and the process when the Beacon frame is

lost are described. In Section 3, the proposed protocol

is described and In Section 4, the performances of the

proposed protocol is evaluated through extensive

simulations. Finally conclusions are made in the last

section.

2 PRELIMINARIES

2.1 IEEE802.15.4 Standard

Beacon-enabled LR-WPANs defined in

IEEE802.15.4 standard operates within a certain time

period, called ‘Superframe’. The superframe is re-

peated and begins with Beacon frame which is trans-

mitted periodically by a PAN coordinator. As shown

in Fig. 1, a superframe is subdivided into two parts:

Active and Inactive periods. During Active period,

data is exchanged between devices in a piconet while

nothing occurs during Inactive period. Inactive period

is required to save device’s energy. Therefore, even

though devices have pending frames, they have to

wait until upcoming active period in the next super-

frame. The durations of both periods can be varied.

The Active period is composed of 16 slots and basic

slot duration is 960us when using 2.4GHz-Direct Se-

quence Spread Spectrum (DSSS) mode (IEEE, 2011).

The durations of superframe, Active period, and In-

active period are decided by a PAN coordinator and

are informed to all member devices through Beacon

frame. In addition to inform superframe structure in-

formation to all participating member device, the

Beacon frame is also used to synchronize with partic-

ipating devices and to identify the WPAN. As shown

in Fig. 1, after Beacon frame, contention access pe-

riod (CAP) and Contention Free Period (CFP) are fol-

lowed in a row. CAP adopts the contention-based data

transmissions like carrier sense multiple access with

collision avoidance (CSMA/CA). CFP is composed

of multiple Guaranteed Time Slots (GTSs). The dura-

tions of GTSs are decided by PAN coordinator and

can be different in every superframe. The maximum

number of GTSs in CFP is 7 and a GTS can occupy

more than one slot. GTSs in CFP are also allocated by

the PAN coordinator by devices’ requests. The infor-

mation on the GTS allocation is included in the Bea-

con frame. During GTS, only designated device trans-

mits its packet without contention and collision.

Beacon Interval (BI) which is the length of the su-

perframe is defined as follows:

][2 symbolsionframeDurataBaseSuperBI



(1)

where aBaseSuperframeDuration is the number of

CAP

Superframe

# n

CFP

Superframe

# n+1

Superframe

# n-1

Active

Inactive

CAP

CFP

Active

Inactive

Beacon Frame

SENSORNETS 2016 - 5th International Conference on Sensor Networks

symbols forming a superframe when the superframe

order (SO) is equal to 0, and BO is Beacon Order that

means how often the beacon is to be transmitted.

aBaseSuperframeDuration is around 960 symbols

recommended in (IEEE, 2011), BO is one value from

0 to 15 and SO is one value from 0 to BO. The active

period, defined by SuperframeDuartion (SD), is cal-

culated by:

].[2 symbolsionframeDurataBaseSuperSD



(2)

2.2 Process When Beacon Frame Is

Lost

Even though we extensively research, no literature

describing the process when Beacon frame is not suc-

cessfully received has been found. Only both of

IEEE802.15.4 and IEEE802.15.4e standards (IEEE,

2011), (IEEE, 2012) describes process for the case in

which GTSs are allocated in the superframe. Based

on standards, if a device requesting GTSs fails to re-

ceive Beacon frame, it has to hold its transmission

during GTSs within the superframe. Even though it

was assigned with GTSs in the previous superframe,

it need to hold transmissions in the current super-

frame. Since a Beacon frame contains the information

on superframe structure like period of CAP, the allo-

cation of GTS, and so on, and the superframe struc-

ture can vary in every superframe, if a device fails to

receive a Beacon frame, it can be assumed that it

needs better to hold its transmissions during the su-

perframe to prevent from collisions with other sched-

uled transmissions. This is ensured for the cases that

the net-work parameters like the number of devices,

traffic loads, etc. are frequently fluctuated.

Moreover, based on the standards, if an

aMaxLostBeacons number of Beacon frames are not

successfully received at a device, the device declares

synchronization loss and starts orphan channel scan

after discarding all buffered packets. That is, the de-

vice restart to associate with new piconet and this pro-

cess waists lots of time.

Overall, the losses of Beacon frames cause hold-

ing devices’ transmissions as well as the synchroni-

zation loss, and as consequences it severely degrades

the network performances.

3 PROPOSED PROTOCOL

As mentioned in Section 2, the loss of Beacon frame

makes devices to hold their transmissions during

whole superframe period. In addition, aMaxLostBea-

cons time of Beacon frame losses causes re-associa-

tion process starting from scanning process. Both of

holding transmission and starting re-association pro-

cess degrades performances of LR-WPANs. Even

though many studies proposes methods to switch

channels to reduce beacon loss, Beacon frames are

still lost due to channel characteristics such as noise,

interference, fading and so on. While preventing from

losing Beacon frames has been studied a lot, there is

no study on enhancement for the process when a Bea-

con frame is lost. Therefore, in this paper, we try to

propose a backward-compatible and effective en-

hanced protocol.

The basic ides of the proposed protocol is to al-

lows a device not receiving a Beacon frame (herein-

after it is called ‘failed-device’) to transmit its queued

data not only in CAP, but also in Inactive period only

if the device cannot wait GTSs in the next upcoming

superframe because of delay constraints of the queued

data frames. Since every superframe is guaranteed to

have at least minimum CAP period which is around 7

slots, all failed-devices can safely transmit their data

during the minimum CAP period.

3.1 Protocol Operations

The detailed process of the proposed protocol for a

failed-device is as follows.

At the moment a device expects to receive a Bea-

con frame, if the frame is not received, the device de-

clares to fail to receive a Beacon frame. Then, it

checks if it has a data that was scheduled to be trans-

mitted in a GTS. If it has ones, the device checks if

the queued data can be held by the next upcoming su-

perframe. If the transmission can be held, the device

holds the data and wait for the next Beacon frame.

However, if the data is delay-constraint traffic, so that

it need to be transmitted in the current superframe, it

transmit the data during CAP. Before transmitting the

data, the device forms data frame by setting Frame

Type field to 100 in binary number. Binary number

100 is not used in IEEE802.15.4 standard and is used

to indicate that the data is transmitted by the rule of

the proposed protocol. Then, the data frames are

transmitted only during the possible minimum CAP

period.

Based on IEEE802.15.4 standard, the minimum

CAP period is defined as maximum number of slots

assigned for CFP minus the number of slots in a su-

perframe. Therefore, during CAP

min

, the failed-device

can safely transmit its data because CAP

min

is a guar-

anteed period.

At the end of 16 slots which is at the end of current

superframe period, if the device still has queued data

to be transmitted in the current superframe, it keeps

Performance Improvement in Beacon-enabled LR-WPAN-based Wireless Sensor Networks

sending data frames even in upcoming Inactive period

in the manner of CAP.

When transmitting a data in CAP, if a failed-de-

vice has more queued data to be transmitted, it sets

Frame Pending field to 1. By doing this, the destina-

tion device expects more data is coming. When the

device is not received next data frame in CAP, it

won’t go to sleep mode and waits even in Inactive pe-

riod to receive the data frames until receiving a data

frame with Frame Pending field set to 0.

After the current superframe period is completed,

normal operation will be proceeded.

Since our proposed method utilizes Inactive pe-

riod, it may incur additional energy consumption.

However, the use of inactive period is invoked only if

there are still pending data that has not ﬁnished trans-

mission in the CAP. In addition, even when listening

during the inactive period, devices can minimize en-

ergy loss by using low-power listening techniques

proposed in (Polastre et al., 2004). In addition, the

proposed protocol targets not only to battery-powered

WSNs, but also to many IEEE802.15.4 applications

such as the smart grid AMI, where each device can be

connected to a power source and low energy con-

sumption is not the upper most requirement (low cost

constraint is still valid, and low power is also valid

due to regulatory reasons). These applications need to

deal with high trac load since the network consists

of a large number of devices. Thus, we focus on

throughput rather than power consumption in our

evaluation.

4 PERFORMANCE

EVALUATIONS

4.1 Theoretical Analysis

The proposed protocol is compared with

IEEE802.15.4-based protocol. Even though, as men-

tioned in Section 2, many methods to avoid from in-

terferences are proposed, any protocol does not focus

the process for the case of Beacon frame loss. There-

fore, in terms of the process for beacon loss case,

there is no comparative protocol, but IEEE802.15.4-

based protocol.

The throughputs achieved by IEEE802.15.4 and

the proposed protocol can be derived as Eq. (3) and

(4), respectively.

)1()1()1(

PPDPPD

Throughput

DBLossDBSucc

IEEE





(3)

)1()1(

PPD

Throughput

DBSucc

proposed





(4)

where P

and P

are packet error rate of data and

Beacon frames, respectively, D

succ

is the amount of

data transmitted when Beacon frame is successfully

transmitted, D

Loss

is the amount of data transmitted

when it is failed to receive Beacon frame, and T is

superframe duration. Therefore, comparing to

IEEE802.14.5 standard-based protocol, the theoret-

ical throughput gain obtained by the proposed pro-

tocol is

)1(

Bsucc

BLoss

GainThroughput







(5)

4.2 Numerical Evaluations

4.2.1 Evaluations with Theoretical Analysis

Fig. 3 shows the throughput gains obtained from Eq.

(5) as functions of Beacon frame size, P

, and β. β is

defined as D

Loss

succ

, that is, β indicates how amount

of data can be transmitted when losing Beacon frame

comparing to that in normal case. When the bit errors

are independently and Identically Distributed (i.i.d),

the relationship between P

and P

is defined as fol-

lows (Rappaport, 2002), (Kim et al., 2010):

,)1(1

/ NM

PP --

(6)

where M and N are the number of bits in a Beacon

frame and a data frame, respectively. Varying the val-

ues of P

emulates erroneous channel environment

caused by interference, thermal noise, fading, colli-

sions, etc. We evaluate throughput improvements by

varying P

from 5% to 40%. 40% packet error rate

might be too high. However, as mentioned in Section

1, the packet error rate of LR-WPANs is widely dis-

tributed from 0% to 100%. Particularly, it is more se-

vere when LR-WPANs coexist with WLANs. Thus,

it is worthwhile to see the performances even in high

packet error rate such as 40%.

As shown in Fig. 2, the proposed protocol

achieves from 15% up to 67% performance improve-

ment. When β is high, the gain is also high because

the high value of β means more data transmitted dur-

ing CAP even though Beacon frame is lost. When

Beacon frame size is low, the proposed protocol

achieves relatively low gain because the P

is low.

Since the proposed protocol enhances IEEE802.15.4

SENSORNETS 2016 - 5th International Conference on Sensor Networks

Figure 2: Throughput gain as a function Beacon frame size,

β, and P

Table 1: Simulation Parameters.

Parameter Value

BO (Beacon Order)

Symbol

16us

aBaseSlotDuration

60 symbols

aMaxLostBeacons

Contention Window (CW)

macMaxFrameRetries 3

macMinBE

macMaxBE

macMaxFrameRetries

standard-based protocol when Beacon frame is lost,

low value of P

does not make big different in terms

of performances.

4.2.2 Evaluations with Simulations

Throughputs of IEEE802.15.4-based protocol and

proposed method are compared through simulations

using Network Simulator-2 (NS-2) version 2.34. For

the simulations, one piconet with a PAN coordinator

and member devices are considered and throughputs

between the PAN coordinator and the device are ob-

served. We intentionally set P

and change from 5%

to 40% to emulate the degree of interference environ-

ment. Parameters used in simulation are shown in Ta-

ble I. Data rate for the simulation is set to 125Kbps.

In the application layer, constant bit rate (CBR) traffic

is generated at the device, and the CBR packet size is

100bytes. The CBR packets are transmitted to the

PAN coordinator through UDP/IP layer. We evaluate

network performances in 0.01 and 0.001 packet inter-

arrival times. Each simulation runs 100seconds. As

shown in Fig. 3, throughput improvements are

achieved from 4.5% up to 35% and from 5.9% to

33.6% at 0.01 and 0.001 packet inter-arrival times, re-

spectively.

Figure 3: Throughput as a function of P

and packet inter

arrival time.

5 CONCLUSIONS

The reliability in the beacon transmissions is very

critical on the performance of Beacon-enabled LR-

WPANs because the loss of beacon causes for devices

to hold their transmissions during the superframe.

Unlike specification in the standard, the method pro-

posed in the paper allows devices to transmit its pend-

ing packet only during the minimum period of CAP

that is guaranteed in the superframe as well as Inac-

tive period without colliding with any transmission in

CFP. By using this protocol, it is proved that average

performance throughputs are improved up to 65% in

theoretical analysis and 35% in simulations over 40%

packet error rate channel and 100-bytes Beacon frame

size.

ACKNOWLEDGEMENTS

This research was supported in part by the Ministry

of Education (MOE) and National Research Founda-

tion of Korea (NRF) through the Human Resource

Training Project for Regional Innovation

(No.2014H1C1A1066943) and in part by Basic Sci-

ence Research Program through the National Re-

search Foundation of Korea (NRF) funded by the

Ministry of Education (2015R1D1A1A01059186).

REFERENCES

Chen, D., Khan, J. Y., Brown, J., 2015. An area packet

scheduler to mitigate coexistence issues in a

WPAN/WLAN based heterogeneous network. In

5 10 20 30 40

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

(%)

Throughput Gain

Beacon Size=100Bytes &



Beacon Size=100Bytes &



=0.5

Beacon Size=50Bytes &



Beacon Size=50Bytes &



=0.5

Beacon Size=100Bytes

Beacon Size=50Bytes

5 10 20 30 40

100

110

120

(%)

Throughput (Kbps)

IEEE802.15.4 w/ 0.01 Interval

Proposed w/ 0.01 Interval

IEEE802.15.4 w/ 0.001 Interval

Proposed w/ 0.001 Interval

Performance Improvement in Beacon-enabled LR-WPAN-based Wireless Sensor Networks

ICT’15, 22nd International Conference on Telecommu-

nications. IEEE.

Howitt, I., Gutierrez, J.A., 2003. IEEE 802.15.4 low rate-

wireless personal area network coexistence issues. In

WCNC’03, Wireless Communications and Networking

Conference. IEEE.

IEEE, 2011. IEEE Std. 802.15.4:2006, Part 15.4: Wireless

Medium Access Control (MAC) and Physical Layer

(PHY) Specification for Low Rate.

IEEE, 2012. IEEE Std 802.15.4e, Part 15.4: Low-Rate

Wireless Personal Area Networks (LR-WPANs).

Kim, J.W., Kim, J., Eom, D.S., 2008. Multi-dimensional

channel management scheme to avoid beacon collision

in LR-WPAN. IEEE Transactions on Con-sumer Elec-

tronics, 54(2):396-404.

Kim, B.-S., Kim, S.W., Fang, Y., Wong, T., 2010. Feed-

back-Assisted MAC Protocol for Real Time Traffics in

High Rate Wireless Personal Area Networks. Wireless

Networks, 16(4): 1109-1121.

Koubaa, A., Alves, M., Attia, M., Nieuwenhuyse, A. V.,

2007. Collision free beacon scheduling mechanisms for

IEEE 802.15.4/ZigBee cluster tree wireless sensor net-

works. In ASWN’07, 7th International Workshop Appli-

cations and Services in Wireless Networks.

Lau, S. Y., Lin, T. H., Huang, T. Y., Ng, I. H., Huang, P.,

2009. A measurement study of ZigBee-based indoor lo-

calization systems under RF interference. In

WINTECH'09, 4th ACM International Workshop on

Experimental Evaluation and Characterization, ACM.

Nam, S.-W., Hwang, K.-I., 2014. Enhanced Beacon Sched-

uling of IEEE802. 15.4 e DSME. Frontier and Innova-

tion in Future Computing and Communications.

Park, J. H., Kim, B. S., 2014. Reliable beacon transmission

based MAC protocol for LR-WPANs over WLAN in-

terferences. Journal of Zhejiang University SCIENCE

C, 15(6), 470-481.

Polastre, J., Hill, J., Culler, D., 2004. Versatile low power

media access for wireless sensor networks”, In Sen-

Sys’04, Proceedings of the 2nd ACM International

Conference on Embedded Networked Sensor Systems.

ACM.

Rappaport, T.S., 2002. Wireless Communications: Princi-

ples and Practices, Prentice Hall. New Jersey, 2

edi-

tion.

Sikora, A., Groza, V.F., 2005. Coexistence of

IEEE802.15.4 with other systems in the 2.4 GHz-ISM-

band. In IMTC’05, Instrumentation and Measurement

Technology Conference.

Shin, S.Y., Park, H.S., Kwon, W.H., 2007. Mutual interfer-

ence analysis of IEEE 802.15.4 and IEEE 802.11b.

Computer Networks, 51(12):3338-3353.

Stanciulescu, G., Farhangi, H., Palizban, A., Stanchev, N.,

2012. Communication technologies for BCIT smart mi-

crogrid. In PES’12, IEEE PES Innovative Smart Grid

Technologies. IEEE.

Yoon, D. G., Shin, S. Y., Kwon, W. H., Park, H. S., 2006.

Packet error rate analysis of IEEE 802.11 b under IEEE

802.15. 4 interference. In VTC’06, Vehicular Technol-

ogy Conference. IEEE.

Yuan, W., Wang, X., Linnartz, J.-P.M.G., 2007. A coexist-

ence model of IEEE 802.15.4 and IEEE 802.11b/g. 14th

IEEE Symposium on Communications and Vehicular

Technology. IEEE.

SENSORNETS 2016 - 5th International Conference on Sensor Networks